The Karma system is currently undergoing maintenance (Monday, January 29, 2018).
The maintenance period has been extended to 8PM EST.

Karma Credits will not be available for redeeming during maintenance.

The leading peer-reviewed journal for digital medicine, and health & healthcare in the Internet age

Latest Submissions Open for Peer Review

JMIR has been a leader in applying openness, participation, collaboration and other "2.0" ideas to scholarly publishing, and since December 2009 offers open peer review articles, allowing JMIR users to sign themselves up as peer reviewers for specific articles currently considered by the Journal (in addition to author- and editor-selected reviewers).

For a complete list of all submissions across all JMIR journals as well as partner journals, see JMIR Preprints

Note that this is a not a complete list of submissions as authors can opt-out. The list below shows recently submitted articles where submitting authors have not opted-out of open peer-review and where the editor has not made a decision yet. (Note that this feature is for reviewing specific articles - if you just want to sign up as reviewer (and wait for the editor to contact you if articles match your interests), please sign up as reviewer using your profile).

To assign yourself to an article as reviewer, you must have a user account on this site (if you don't have one, register for a free account here) and be logged in (please verify that your email address in your profile is correct).

Add yourself as a peer reviewer to any article by clicking the '+Peer-review Me!+' link under each article. Full instructions on how to complete your review will be sent to you via email shortly after. Do not sign up as peer-reviewer if you have any conflicts of interest (note that we will treat any attempts by authors to sign up as reviewer under a false identity as scientific misconduct and reserve the right to promptly reject the article and inform the host institution).

We now reward completed peer-reviews (all rounds must be completed) with 90 Karma points which can be used as credits towards your own submissions. In addition, you receive karma points at the time of self-assignment, and additional bonus points for nominating other reviewers as well as for excellent reviews. Conditions apply, see Karma Description for details. Note that assigning yourself as reviewer and not delivering a review will lead to negative karma points.

The standard turnaround time for reviews is currently 2 weeks, and the general aim is to give constructive feedback to the authors and/or to prevent publication of uninteresting or fatally flawed articles. Reviewers will be acknowledged by name if the article is published, but remain anonymous if the article is declined.

The abstracts on this page are unpublished studies - please do not cite them (yet). If you wish to cite them/wish to see them published, write your opinion in the form of a peer-review!

Tip: Include the RSS feed of the JMIR submissions on this page on your homepage, blog, or desktop RSS reader to stay informed about current submissions!

JMIR Submissions under Open Peer Review

↑ Grab this Headline Animator

If you follow us on Twitter, we will also announce new submissions under open peer-review there.

Titles/Abstracts of Articles Currently Open for Review:

  • Background: Randomized trials of web-based decision aids for prostate specific antigen (PSA) testing indicate that these interventions improve knowledge and reduce decisional conflict. However, we do not know about these tools’ impact on people who spontaneously use a PSA testing patient decision aid on the internet. Objective: 1) Determine the impact of publicly available web-based PSA Option Grid patient decision aids on preference shift, knowledge, and decisional conflict; 2) identify which frequently asked questions (FAQs) are associated with preference shift; 3) explore the possible relationships between these outcomes. Methods: Data were collected between January 1, 2016 and December 30, 2017. Users who accessed the online, interactive PSA Option Grid were provided with three options – have a PSA test, no PSA test, unsure. Users first declared their initial preference, completed five knowledge questions, and a four-item (yes or no) validated decisional conflict scale (SURE). Next, users were presented with ten FAQs and asked to identify their preference for each question based on the information provided. At the end, users declared their final preference and completed the same knowledge and decisional conflict questions. Paired sample t-tests were employed to compare before-and-after knowledge and decisional conflict scores. A multinomial regression analysis was conducted to determine which FAQs were associated with a shift in screening preference. Results: Of the 467 people who accessed the PSA Option Grid, 186 (40%) completed the interactive journey and associated surveys. After excluding 22 female users, we analyzed 164 responses. At completion, users shifted their preference to ‘not having the PSA test’ (26% vs 71%; P <.01), had higher levels of knowledge (68% vs 89%, P < .01), and lower decisional conflict (57% vs 11%, P < .01). Three FAQs were associated with preference shift: What does the test involve? If my PSA level is high, what are the chances that I have prostate cancer? What are the risks? No relationships were present between knowledge, decisional conflict, and preference shift. Conclusions: Unprompted use of the interactive PSA Option Grid leads to preference shift, increased knowledge, and reduced decisional conflict which confirms the ability of these tools to influence decision-making, even when used outside clinical encounters.

  • A novel approach to evaluating mobile smartphone screen time: feasibility and preliminary findings

    Date Submitted: May 20, 2018
    Open Peer Review Period: May 21, 2018 - Jul 16, 2018

    Background: Increasingly high levels of smartphone ownership and use pose the potential for addictive behaviors and negative health outcomes, particularly among younger populations. Previous methodologies to understand mobile screen time have relied on self-report survey or ecological momentary assessment (EMA). Self-report is subject to bias and unreliability, while EMA can be burdensome to participants. New methodology is needed to advance the understanding of mobile screen time. Objective: The objective of this study was to test the feasibility of a novel methodology to record and evaluate mobile smartphone screen time and use: Battery Use Screenshot (BUS). Methods: The Battery Use Screenshot (BUS) approach, defined for this study as uploading a mobile phone screenshot of a specific page within a smartphone, was utilized within an online cross-sectional survey of adolescents aged 12 to 15 years old through the survey platform Qualtrics. Participants were asked to provide a screenshot of their battery use page, a feature within smartphones, to upload within the online survey. Feasibility was assessed by smartphone ownership and response rate to BUS upload request. Data availability was evaluated as applications (apps) per BUS, completeness of data within screenshot, and five most used applications based on battery use percentage. Results: Among those surveyed, 309 (26.7%) indicated ownership of their smartphone. A total of 105 screenshots were evaluated. For data availability, screenshots contained an average of 10.2 (SD=2.0) apps per screenshot and over half (55%) had complete data available. The most common apps included safari and home/lock screen. Conclusions: Findings describe BUS as a novel approach for real-time data collection focused on mobile smartphone screen time among young adolescents. Though feasibility showed some challenges in upload capacity of young teens, data availability was generally strong across this large data set. This data from screenshots have the potential to provide key insights into precise mobile smartphone screen use and time spent per mobile application. Future studies could explore the use of the BUS methodology to correlate mobile smartphone screen time with health outcomes.

  • Background: In health-related, web-based information search, people should choose objectively correct information, but they are often misguided by confirmation bias–the tendency to select and evaluate information in line with their prior attitudes. They are also misguided by dubious information, not taking source credibility into account properly. Objective: We test whether people are prone to confirmation bias in mental health-related information search, particularly (1) if high confidence worsens confirmation bias, (2) if social tags are an appropriate interface to circumvent the influence of prior attitudes, and (3) if people successfully distinguish high and low source credibility. Moreover, we describe attitudes towards the efficacy of the treatment of depression with antidepressants and psychotherapy. Methods: 520 participants of a representative sample of the German population were recruited on an online platform of a panel company. 250 (48%) completed the fully automated, randomized, controlled web-based study, which was accessible online from November 14th to November 18th 2014, until at least 250 participants completed the survey. Participants provided prior attitudes about antidepressants and psychotherapy. We manipulated (1) confidence by having participants recall situations in which they were confident or doubtful. Next, participants searched for blog posts about the treatment of depression, with social tag clouds differing in (2) tag popularity–either psychotherapy or antidepressant tags were more popular. Finally, we manipulated (3) source credibility with banners indicating high or low expertise of the tagging community, and we measured tag- and blog post selection, and treatment efficacy ratings after navigation. Results: We observed a tendency to rate psychotherapy (mean = 5.24, SD = 1.10) as more effective than antidepressants (mean = 4.61, SD = 1.19; t(225) = 9.71, P < .001, d = .56.). Tag popularity predicted the proportion of selected antidepressant tags (beta = 0.44, SE = .11, P < .001), and blog posts (beta = 0.46, SE = .11, P < .001). We could not replicate the confidence manipulation (t(224) < 1, P = .78). Participants did not attend to source credibility on banners (t(224) = 1.67, P = .10). When confidence was low (-1 SD), participants selected more blog posts consistent with prior attitudes (beta = -0.26, SE = 0.05, P < .001). Moreover, when confidence was low (-1 SD) and source credibility was high (+1 SD), the efficacy ratings of attitude consistent treatments increased (beta = 0.34, SE = 0.13, P = .01). Conclusions: We found correlational support for defense motivation account underlying confirmation bias in the mental health-related search context. That is, participants did not tend to select objectively correct, but information that supported their prior attitudes.

  • Background: Patients with hypothyroidism report poor health-related quality of life despite having undergone thyroid hormone replacement therapy (THRT). Understanding patient concerns regarding levothyroxine can help improve the treatment outcomes of thyroid hormone replacement therapy. Objective: This study aimed to (1) identify the distinctive themes in patient concerns regarding THRT; to (2) determine whether patients have unique primary medication concerns specific to demographics, and to (3) determine the predictability of primary medication concerns on patient treatment satisfaction. Methods: We collected patient reviews from WebMD (1,037 reviews about generic levothyroxine and 1,075 reviews about the brand version) posted between September 1, 2007 and January 30, 2017. We used natural language processing (NLP) to identify the themes of medication concerns. Multiple regression analyses were conducted in order to examine the predictability of the primary medication concerns on patient treatment satisfaction. Results: NLP of the patient reviews of levothyroxine posted on a social networking site produced six distinctive themes of patient medication concerns related to levothyroxine treatment: ‘How to take the drug,’ ‘treatment initiation,’ ‘dose adjustment,’ ‘symptoms of pain,’ ‘generic substitutability’ and ‘appearance.’ Patients had different primary medication concerns unique to their gender, age, and treatment duration. Furthermore, treatment satisfaction on levothyroxine depended on what primary medication concerns the patient had. Conclusions: NLP of text content available on social networks could identify different themes of patient medication concerns that can be incorporated into tailored medication counseling to improve patient treatment satisfaction.

  • Background: Advanced lung cancer patients often have chronic lung disease with reduced exercise capacities and various symptoms leading to altered quality of life (QoL). No studies have assessed pulmonary rehabilitation (PR) employing a mobile application and an Internet of Things device in advanced lung cancer patients undergoing chemotherapy. Objective: We determined the feasibility and efficacy of a smartphone application-based PR on exercise capacity, symptom management, and QoL in these patients. Methods: A total of 100 patients were recruited in a prospective, single-arm intervention study using smartphone application-based PR program for 12 weeks. Exercise capacity (6-minute walking distance, 6MWD), QoL, symptom scale scores, and distress indexes were investigated. Results: Ninety patients completed the PR program. The most common cause of dropout was hospitalization due to cancer progression. After PR, there was significant improvement in the 6MWD; 380.1 ± 74.1 m at baseline, 429.1 ± 58.6 m at 6-weeks (P < .001), and 448.1 ± 50.0 m at 12-weeks (P < .001). However, the dyspnea scale score showed no significant improvement in the patients overall, but there was a trend for improvement in those with a stable tumor response (P = .065). Role (P = .02), emotional (P < .001), and social functioning (P = .002) scale scores showed significant improvement after PR. Symptom scale scores for fatigue (P < .001), anorexia (P = .047), and diarrhea (P = .01) also showed significant improvement. There was significant improvement in depression (P = .048) and anxiety (P = .01), while there was no significant change in QoL (P = .063) and severity of pain (P = .24). Conclusions: Smartphone application-based PR represents an effective and feasible program to improve exercise capacity, and to manage symptoms and distress in patients with advanced lung cancer, undergoing chemotherapy.

  • Person-Centered, Technology Enhanced Care Model For Managing Chronic Conditions: Development and Implementation

    Date Submitted: May 18, 2018
    Open Peer Review Period: May 19, 2018 - Jul 14, 2018

    Background: Caring for individuals with chronic conditions is labor intensive, requiring ongoing appointments, treatments, and support. The growing number of individuals with chronic conditions makes this current support model unsustainably burdensome on health care systems globally. Mobile health (mHealth) technologies are increasingly being used throughout health care to facilitate communication, track disease, and provide educational support to patients. Such technologies show promise, yet they are not being utilized to their full extent within US health care systems. Objective: The purpose of this study was to examine the utilization of staff and costs of a remote monitoring care model in persons with and without a chronic condition. Methods: At Dartmouth-Hitchcock Health, 2,894 employees volunteered to monitor their health, transmit data for analysis, and communicate digitally with a care team. Volunteers received Bluetooth-connected consumer-grade devices that were paired to a smartphone application that facilitated digital communication with nursing and health behavior change staff. Health data were collected, automatically analyzed, and generated behavioral support communications based on those analyses. Care support staff were automatically alerted according to purpose-developed algorithms. In a subgroup of participants and matched controls, we used difference-in-difference techniques to examine changes in per-capita expenditures. Results: Participants averaged 41 years of age; 73% (n = 2,104) were female and 13% (n = 376) had at least one chronic condition. On average, each month, participants submitted 23 vital sign measurements, engaged in 1.96 conversations, and received 0.25 automated messages. Persons with chronic conditions accounted for 40% of all staff conversations, with higher per-capita conversation rates for all shifts compared to those without chronic conditions (P<.001). Additionally, persons with chronic conditions engaged nursing staff more than those without chronic conditions (1.40 & 0.19 per-capita conversations, respectively, P<.001). When compared to the same period in the prior year, per-capita healthcare expenditures for persons with chronic conditions dropped by 15% (P=.06) more than did those for matched controls. Conclusions: The technology-based chronic condition management care model was frequently used and demonstrated the potential for cost savings among participants with chronic conditions. While further studies are necessary, this model appears to be a promising solution to efficiently provide patients with personalized care, when and where they need it.

  • Background: eHealth is the use of information and communication technology in the context of healthcare and health research. Recently, there has been a rise in the number of eHealth modalities and the frequency in which they are used to deliver technology-assisted self-management interventions for people living with chronic pain. However, there has been little or no research directly comparing these eHealth modalities. Objective: The aim of the current systematic review with a network meta-analysis is to directly compare the effectiveness of eHealth modalities in the context of chronic pain. Methods: Randomised controlled trials (N>20 per arm) that investigated technologically delivered interventions for adults with chronic pain were included. Data were extracted on pain severity, psychological distress and HRQoL, and the risk of bias was assessed. Studies were classified by their primary mode of delivery. Pair-wise meta-analyses were undertaken and a network meta-analysis was conducted to generate indirect comparisons of modalities for reducing pain severity. Results: The search returned 18,470 studies with 18,349 excluded (duplicates [2,310]; title and abstract [16,039]). Of the remaining papers, 30 studies with 4,595 randomised participants were included in the review. Rankings tentatively indicate that telephone supported interventions are the most effective, with a 46% chance that telephone intervention was the best modality, followed by studies delivered via interactive voice response, internet and virtual reality. Conclusions: This current systematic review with a network meta-analysis generated comparisons between previously un-compared technological modalities to determine which delivered the most effective interventions for the reduction of pain severity in chronic pain patients. There are limitations with this review; in particular, the underrepresented nature of some eHealth modalities included in the analysis. However, in the event that the review is regularly updated a clear ranking of eHealth modalities for the reduction of pain severity will emerge. Clinical Trial: PROSPERO: Registration database number: CRD42016035595

  • Social connection and online engagement: Insights from users of a mental health online forum.

    Date Submitted: May 18, 2018
    Open Peer Review Period: May 19, 2018 - Jul 14, 2018

    Background: Over the past two decades, online forums for mental health support have emerged as an important tool for improving mental health and wellbeing. There has been important research which analyses the content of forum posts, studies on how and why individuals engage with forums, and how extensively forums are used. However, we still lack insights into key questions on the effectiveness of medical and health Internet tools, how they are experienced by their users, and how they might be best configured and deployed – especially in rural and remote settings, outside major cities and towns. Objective: The aim of our study was to study the dynamics, benefits, and challenges of a peer-to-peer mental health online forum, from a user perspective. In particular, to better explore and understand user perspectives on connection, engagement, and support offered in such forums, information and advice they gained, and what issues they encountered. We studied both consumer and carer experiences of the forums. Methods: In order to understand the experience of forum users, we devised a qualitative study utilising semi-structured interviews with 17 participants (12 women, 5 men). Data were transcribed, coded, and analysed via key themes. Results: The study found that participants: experienced considerable social and geographical isolation which the forums helped to address; sought out the forums as a way to find social connection that was lacking in their everyday lives; used the forums to both find and provide information and practical advice. Conclusions: The study confirms that online peer support provides a critical, ongoing role in providing social connection for mental health consumers. In addition, the research shows forums also provide crucial information and advice for many in this study, especially those living in rural and remote areas. Key to the success of this online forum was careful configuration of moderation, deployment of professional support and advice.

  • Reminder use to improve service referral use among participants in a community engagement program

    Date Submitted: May 18, 2018
    Open Peer Review Period: May 19, 2018 - Jul 14, 2018

    Background: Recent research on health care delivery has shown that mobile based interventions are effective in bringing behavior change among clinical populations. Mobile phone-based interventions are considered to be convenient and economical to the provider and non-invasive and hence acceptable to the receiver. Studies have also shown that personalized text message reminders are more acceptable and effective than automated mass messages. However, despite the increasing literature on mobile phone-based interventions (mHealth) contributing to improvement in health care, there is a dearth of information on the use of reminders for social and medical service referrals provided to nonclinical populations by community-engagement programs. Objective: This pilot study aimed to apply mobile phone-based interventions to improve the utilization rates of medical and social service referrals provided to community members by Community Health Workers (CHW’s) through HealthStreet – a community engagement initiative at the University of Florida (UF). Methods: 300 eligible HealthStreet members were randomly assigned to CALLSONLY, TEXTONLY, CALLS+TEXT reminder intervention groups. Reminders were sent based on their assignment on the 15th and 45th day after the baseline assessment reminding them to utilize the medical and/or social referrals that were provided by the community health workers based on their expressed need during the baseline assessment. A telephonic follow up was done on the 30th and 60th day to assess the rate of utilization of the referrals. Further, a telephone based brief satisfaction survey was also administered to all participants on the 60th day follow up call. Results: Out of 290 participants, 201 (69.3%) successfully completed the 30-day HealthStreet follow up and 149 (51.4%) participants successfully completed the 60-day HealthStreet follow up. At the 30-day follow up, 52 of the 201 respondents reported the use of at least one referral and there were no differences for utilization by contact status. At the 60-day follow up, 69 of the 149 participants reported utilizing at least one referral; there were no differences for this group either. Conclusions: Findings from this study did not find any significant difference in the rate of service referral utilization between participants who received different types of reminders. The absence of a significant difference between the two groups may be because of the smaller sample size of the study.

  • Background: The prevalence of childhood obesity continues to increase, and clinic-based treatment options have failed to demonstrate effectiveness. One of the strongest predictors of child weight is parent weight. Parental treatment for weight loss may indirectly reduce obesity in the child. We have previously demonstrated the effectiveness among adults of a fully-automated, evidence-based digital weight loss intervention (Track). However, it is unknown if it is feasible to deliver such a treatment directly to parents with obesity who bring their child with obesity to a weight management clinic for treatment. Objective: To evaluate the feasibility of and engagement with a digital weight loss intervention to parents of children receiving treatment for obesity. Methods: We conducted a 6-month pre-post feasibility trial among parents/guardians and their children aged 4-16 years presenting for tertiary-care obesity treatment. Along with the standard family-based treatment protocol, parents received a 6-month digital weight loss intervention, which included weekly monitoring of personalized behavior change goals via mobile technologies. We examined levels of engagement by tracking completed weeks of self-monitoring and feasibility by assessing change in weight. Results: Participants (n=48) were on average 39 years old, mostly female (82%), non-Hispanic Black (51%) and had obesity (75%). Over a quarter had a yearly household income of less than $25,000 and about a third had the equivalent of a high school education. Children were on average 10 years old and had a BMI of 29.8. The median percent of weeks participants tracked their behaviors was 77% (IQR 6.3 to 100) (18.5/24 total weeks). The median number of attempts via phone or text required to complete one tracking week was 3.3 (IQR 2.6 to 4.9). Nearly half (48%) had high levels of engagement, completing 19 (80%) or more weeks of tracking. Of the 26 participants with weights at 6 months (81% self-reported) there was a median 2.44kg (IQR -6.5 to 1.0) decrease in weight. Conclusions: It is feasible to deliver an evidence-based digital weight loss intervention to parents/guardians whose children are enrolled in a weight management program. Given the feasibility of this approach, future studies should investigate the effectiveness of digital weight loss interventions for parents on child weight and health outcomes.

  • Background: Health behavior (HB) patterns reported through daily diary data are important to understand and intervene upon at the individual level for N-of-1 trials and related study designs. Analyses often utilize regressions that evaluate aggregate effects across individuals. Furthermore, standard analyses target single outcomes. There is often interest in relationships between multiple outcomes, such as stress and HB. Objective: This article illustrates how individuals’ daily reports of stress and HB (time series) can be explored using visualization tools. Methods: Secondary analysis was conducted on six months of daily diary reports of stress and HB (physical activity and diet quality) from mostly ethnic minority mothers who pilot tested a self-monitoring mobile health app. Time series with minimal missing data from 14 of 44 mothers were analyzed. Correlations between stress and HB within each time series were reported as a preliminary step. Stress and HB time series patterns were visualized by plotting moving averages and time points where mean shifts in the data occurred (changepoints). Results: Median correlation was small and negative for associations of stress with physical activity (r = -0.14) and with diet quality (r = -0.083). Moving averages and changepoints for stress and HB were aligned for some participants, but not others. A third subset of participants exhibited little variation in stress and HB reports. Conclusions: Median correlations corroborated prior findings. Time series visualizations highlighted variations in stress and HB across individuals and time that are difficult to capture through correlations and regression-based summary measures.

  • Background: An increasing number of people visit online health communities to esquire health information with doctors. In the online health community (OHC), patient crowds tended to label and vote the doctors’ specialties with encountered disease. Understanding how patients’ online labels can help us understand the service diversity for patients in online health communities and provide constructive suggestions for doctors serving more patients online. Objective: Our goal was to understand: (1) what kind of patterns are the labels of patient crowdvotes aggregated service diversity, including encountered disease labels and online votes, in a OHC? (2) wheather the patient crowdvotes aggregated service diversity make doctors’ service sales difference in OHC? (3) how can managers in OHC perform to improve doctors’ service sales with the feedback of crowdvotes aggregated service diversity? Methods: We designed a retrospective study with data collected from the largest OHC (Good Doctor website) in China. We first used descriptive statistics to investigate the patient crowdvotes aggregated service diversity. Then a multiple log-linear relationship was adapted to investigate the main and the interaction impact of service diversity on doctors’ service sales. Results: Our sample consists of 9,841 doctors from 1,255 different hospitals widely distributed in China. 18,997,018 patients had been serviced by these doctors since they became members of the study OHC. 704,467 votes of doctors’ clinical specialties were labeled by patient crowds in recent two years (Aug.26, 2015-Aug. 25, 2017). Gini coefficient of serviced patients is very high, 0.626, followed by the volume of votes (0.562). Based on the regression model, we found that the coefficients of the control variables, doctor review rating and clinic title, were 0.810(0.041), and 1.735 (0.027), respectively. For the breadth of voted specialties, volume of votes and degree of voted diversity, the standardized coefficient of the main effect were 0.309 (0.038), 0.745 (0.014) and 0.073 (0.018), respectively. All of the estimates are statistically significant at a 0.1% level. Conclusions: Our study provided empirical evidence that the patterns of both the labels of patient crowdvotes aggregated service diversity and doctors’ service sales were of inequality (as illustrated in Lorenz curves) in the distribution of its size of serviced patients in a OHC. Patient crowds’ online labels also leaded to differences in the doctors’ service sales online. The treads of the doctors’ service sales kept increasing as the patient crowdvotes aggregated service diversity increased. Finally, our findings suggested that the higher breadth of voted specialties and degree of voted diversity displayed a greater service sales with a higher review rating, deploying less inequality of Doctors’ service sales.

  • Background: Transitioning into parenthood can be stressful as parents struggle to cope with new parenting responsibilities. Although perinatal care in hospitals aims to improve parental outcomes, there is a general consensus that it is suboptimal and insufficient. Therefore, many studies have designed intervention methods to supplement support for parents during this stressful period. However, studies often focus on parental outcomes as indicators of intervention success and effectiveness. Studies evaluating participants’ experiences and feedback are limited. Objective: To examine the experiences and perceptions of participants who participated in the supportive education parenting program intervention study. Methods: A qualitative semi-structured interview was conducted with 16 mothers (6 control and 10 intervention) from a randomized controlled trial. The supportive education parenting programme included two phone-based perinatal educational sessions, a phone-based educational session after childbirth, and a one-month postpartum access to a mobile-health application. The interviews were approximately 30 to 60 minutes long, were audiotaped and transcribed verbatim, and analysed using thematic analysis. Study findings were reported according to the Consolidated Criteria for Reporting Qualitative Research checklist. Results: Three main themes evaluating mothers’ experiences and perceptions were generated: 1) changed perspective towards parenthood, 2) journey from pregnancy to after birth, and 3) a way forward. Mothers from the intervention group mostly had a good perinatal experience with sufficient support received, which alleviated their emotional wellbeing and increased parenting involvement. Mothers in the control group, although satisfied with the hospital care received, were more stressed and shared a need for professional advice and extra support. Apart from technical enhancements, mothers also requested extended social support during early pregnancy up to one year postpartum, taking into consideration Asian cultural practices. Conclusions: Mothers who received the intervention were overall satisfied with the support provided by the technology-based supportive educational parenting program. The success of the educational programme in this study highlights the need for supplementing standard care in hospitals with technology-based educational programmes. Future research should include fathers’ perceptions to attain an in-depth understanding of overall participants’ experiences and needs in the future development of supportive and educational programs. Clinical Trial: ISRCTN48536064

  • Improving Understanding of Test Results by Substituting (Not Adding) Goal Ranges: Experimental Study

    Date Submitted: May 15, 2018
    Open Peer Review Period: May 17, 2018 - Jul 12, 2018

    Background: Most displays of laboratory test results include a standard reference range. For some patients (e.g. those with chronic conditions), however, achieving a result within the standard range may be unachievable, inappropriate, or even harmful. Objective: To test the impact of including clinically-appropriate goal ranges outside the standard range in visual displays of laboratory test results. Methods: Participants (N=6776) from a demographically diverse online panel viewed hypothetical hemoglobin A1c test results (either A1c=6.2% or =8.2%) as part of a type 2 diabetes management scenario. Test result visual displays included either (1) standard range (4.5%-5.7%) only, (2) a goal range (6.5%-7.5%) added to standard range or (3) goal range only in one of three display formats: (a) table, (b) a simple, two-colored number line (simple line), or (c) a number line with diagnostic categories indicated via colored blocks (block line). Primary outcome measures were comprehension of and negative reactions to test results. Results: While goal range information did not influence understanding of A1c=8.2% results, goal range only displays produced higher levels of comprehension and decreased negative reactions to A1c=6.2% test results compared to the no goal range and goal range added conditions. Goal range information was less helpful in the block line condition versus other formats. Conclusions: Replacing the standard range with a clinically-appropriate goal range could help patients better understand how their test results relate to their personal targets.

  • The unmet needs of caregivers of stroke survivors: A review of the content of YouTube videos

    Date Submitted: May 15, 2018
    Open Peer Review Period: May 17, 2018 - Jul 12, 2018

    Background: Content produced by caregivers of stroke survivors on online video sharing platforms such as YouTube may be a good source of knowledge regarding caregivers’ unmet needs and caregiving experiences. Objective: We aimed to examine content, quantity and quality of YouTube videos that specifically target and discuss the needs and concerns of caregivers of stroke survivors. Methods: YouTube was systematically searched using six select search strings. The first 20 videos retrieved from each search string were screened against the eligibility criteria and included if they were: developed by either adult caregivers of stroke survivors and focused on caregivers’ needs and concerns; or professional organisations and depicted first-hand testimonials of the needs and concerns of caregivers of stroke survivors. A pre-determined coding schedule was used to report the rate of unmet needs in each video: (1) Impact of Caregiving on Daily Activities; (2) Relationship; (3) Emotional and Psychological; (4) Comprehensive Stroke care; (5) Information; and (6) Spirituality. General video characteristics were also reported. Results: Of the 132 videos screened, 26 individual videos were included in the analysis. Most videos were developed in the USA (61.5%) and featured spouses of stroke survivors (65.47%). The most common types of videos were video blogs (vlogs) in which the main content featured the caregiver(s) diarising their daily thoughts and feelings (n=16, 48.5%). In total, 291 unmet needs were reported by caregivers of stroke survivors, with an average of 11.2 unmet needs per video. The most common unmet needs domain was ‘Impact of Caregiving on Daily Activities’ which made up 44% of the unmet needs reported. The most frequently reported sub-category was ‘Financial Impact’, appearing 24 times in 26 videos (92.3%). Conclusions: Content produced by caregivers of stroke survivors on YouTube may be used as a tool for caregivers to provide and receive support through online communication. YouTube videos offer insight into the unmet needs of caregivers of stroke survivors across countries and healthcare systems, and may be used as an additional resource for stroke services to disseminate health information and support to caregivers. Further research is needed to investigate other online platforms to address the unmet needs of caregivers of stroke survivors.

  • Expertise modulates the student’s perception of pain from a self-perspective

    Date Submitted: May 16, 2018
    Open Peer Review Period: May 17, 2018 - Jul 12, 2018

    Background: FMRI studies show that medical doctors suppress the pain matrix when they view painful actions executed on others. But how do doctors perceive pain from a self-perspective? Objective: We hypothesize that virtual dental treatment from a self-perspective may induce brain activity in pain related areas in controls while this is less the case for dental students. We expect that dental students learn to control the motoric aspects of pain during their education because it is a prerequisite for manual treatment. Methods: In this fMRI study, neural correlates of pain perception from a self-perspective in a sample of 20 dental students and 20 age matched controls were investigated trough classic general linear model analysis and in house classification methods. All subjects viewed video clips presenting a dental treatment from the first‐person perspective. Dental students and naïve controls exhibited similar anxiety levels for invasive stimuli. Results: Invasive dentistry scenes evoked less affective component of pain in dental students compared to naïve controls. Reduced affective pain perception went along with suppressed brain activity in pain matrix areas including insula, anterior cingulate cortex and basal ganglia. Furthermore, a substantial reduction of brain activity was observed in motor related areas in particular the supplementary motor area, premotor cortex and basal ganglia (p<0.001, k=156). Within this context a classifier analysis based on neural activity in the nucleus lentiformis (p<0.0005 with k=125) could identify dental students and controls on the individual subject level in 85 percent of the cases (sensitivity = 90.0%; specificity = 80.0%). Conclusions: We speculate that dentistry students learn to control motoric aspects of pain during their education because it is a prerequisite for professional manual treatment of patients. We discuss that a specific set of learning mechanism might affect perceived self-efficacy of dental students, which in turn might reduce their affective component of pain perception.

  • Background: Design thinking and human-centered design approaches have become increasingly common in the healthcare literature, particularly in relation to health information technology (HIT), as a pathway toward the development of usable, diffusible tools and processes. There is a need in academic medical centers tasked with digital innovation for a comprehensive process model to guide development that incorporates current industry trends including design thinking and the lean and agile approaches to digital development. Objective: To describe the foundations and phases of our model for user centered HIT development. Methods: Based on our experience, we established an integrated approach and rigorous process for HIT development that leverages design thinking, lean and agile strategies in a pragmatic way while preserving methodological integrity in support of academic research goals. Results: A four phased pragmatic process model for user-centered digital development in HIT. Conclusions: The culmination of diverse innovation projects, this model for user-centered HIT development represents a multi-phased, high fidelity process for making more creative, flexible, efficient and effective tools. This model is a critical step in building a rigorous approach to HIT design that incorporates a multidisciplinary, pragmatic perspective, combined with academic research practices and state of the art approaches to digital product development to meet the unique needs of healthcare.

  • Using Passive Smartphone Sensing for improved Risk Stratification of patients with Depression and Diabetes

    Date Submitted: May 14, 2018
    Open Peer Review Period: May 15, 2018 - Jul 10, 2018

    Background: Research studies are establishing the use of smartphone-sensing to measure mental well-being. Smartphone sensor information captures behavioral patterns and its analysis helps reveal well-being changes. Depression in diabetes goes highly under-diagnosed and under-reported. The comorbidity has been associated with increased mortality and worse clinical outcomes; including poor glycemic control and poor self-management. Clinical only intervention has been found to have very modest effect on diabetes management among people with depression. Smartphone technologies could play a significant role in complementing co-morbid care. Objective: To present an approach to risk-stratify people with diabetes based on symptoms of depression detected using smartphone-sensing information. Methods: A cross-sectional observational study (Project SHADO- Analyzing Social and Health Attributes through Daily Digital Observation) was conducted on 47 participants with diabetes. The study smartphone-sensing app passively collected data regarding activity, mobility, sleep and communication from each participant. Self-reported symptoms of depression (using validated Patient Health Questionnaire- 9) was collected once every 2 weeks from all participants. A descriptive analysis was performed to understand the representation of the participants. A univariate analysis was performed on each derived sensing variable to compare behavioral changes between depression states- those with self-reported major depression (PHQ-9 > 9) and those with none (PHQ-9 <= 9). A classification predictive modeling, using supervised machine-learning methods, was explored using derived sensing variables as input to construct and compare classifiers that could risk-stratify people with diabetes based on symptoms of depression. Results: A noticeably high prevalence of self-reported depression (30 out of 47 participants, ~65%) was found among the participants. Low correlation was found between self-reported depression state and each of the 53 derived sensing variables. Between depression states, a significant difference was found for average activity rates (day time) among participant-day instances with symptoms of major depression (M=16.06, SD=14.90) and those with none (M=18.79, SD=16.72); P= .005. For average number of people called, a significant difference was found between participant-day instances with symptoms of major depression (M=5.08, SD=3.83) and those with none (M=8.59, SD=7.05); P < .001. These results suggest that participants with diabetes and symptoms of major depression exhibited lower activity through the day and maintained contact with fewer people. Using all the derived sensing variables, the XGBoost machine-learning classifier provided the best performance with an average cross-validation accuracy of 79.07% (95% CI: 74%, 84%) and test accuracy of 81.05% to classify symptoms of depression. Conclusions: Participants with diabetes and self-reported symptoms of major depression were observed to show lower levels of social contact and lower activity levels during the day. While findings must be reproduced in a broader RCT, the study shows promise in use of predictive modeling for early detection of symptoms of depression in people with diabetes using smartphone-sensing information.

  • Background: The demand for smoking cessation services has risen in Turkey, as those planning to quit, reached 35% in 2012. For quitting efforts, communication technologies are used widespread globally, yet service integration is rare. Objective: This study aims to assess the effect of a WhatsApp application embedded in the cessation service delivery, on success rate. Methods: Randomized controlled intervention study was conducted with 132 volunteers, followed up at Ege University Hospital, Smoking Cessation Clinic, March-July 2017. Intervention content based on Transtheoretic Model, was prepared and 60 WhatsApp messages were delivered to participants for 3 months, with 6 months follow up. Initial interview and follow up forms consisting sociodemographics, smoking status, medical condition and treatment, were used. The success rate at 1st and 3rd months were assessed by point-prevalence. Intention-to-treat analysis was used. As secondary outcomes; number of follow-ups, change in weight and continuity of medication were evaluated. Results: Success rate at 1st month was 65.9% for intervention group and 40.9% in control group; for the 3rd month it was 50.0% and 30.7%, consecutively. Being in the intervention group increased success rate by 3.50 (1.30-9.44) times in the 1st 2.50 (1.08-6.40) times in the 3rd and 2.31 (1.03- 5.16) times 6th month. The intervention was the only parameter effective in the two follow-up periods; all other factors were eliminated after adjustment for the intervention. As secondary outcomes, in the intervention group, number of follow-ups and face to face follow-ups increased for the 1st and 3rd months and continuity of medication was higher at 3rd month. Conclusions: WhatsApp support embedded in cessation service delivery increases the success rate and has favorable effects on follow-up. Moreover, the equalizing effect of the intervention in terms of socioeconomic differences, positively contributes to the elimination of inequalities for successful smoking cessation in disadvantaged groups.

  • Background: The ability to successfully recruit participants for eHealth clinical trials is largely dependent on the use of efficient and effective recruitment strategies. Determining which types of recruitment strategies to use presents a challenge for many researchers. Objective: This article presents an analysis of the time-efficiency and cost-effectiveness of recruitment strategies for eHealth clinical trials and describes a framework for cost-effective trial recruitment. Methods: Participants were recruited for one of 5 eHealth trials of interventions for common mental health conditions. A multi-pronged recruitment approach was used, including digital (e.g., social media, Craigslist), research registry-based, print (e.g. flyers, posters on public transportation), clinic-based (e.g., a general internal medicine clinic within an academic medical center, a large nonprofit healthcare organization), a market research recruitment firm, and traditional media strategies (e.g., newspaper and television coverage in response to press releases). The time costs and fees for each recruitment method were calculated and the participant yield on recruitment costs was calculated by dividing the number of enrolled participants by the total cost for each method. Results: A total of 777 participants were enrolled in one of the trials. Digital recruitment strategies yielded the largest number of participants across the 5 clinical trials and represented 34% of the total enrolled participants. Registry-based recruitment strategies were in second place by enrolling 28% of the total enrolled participants across trials. Research registry-based recruitment had a relatively high conversion rate from potential participants who contacted our center to being screened to being enrolled, and it was also the most cost-effective for enrolling participants in this set of clinical trials with a total cost per person enrolled at $8.99. Conclusions: Based on these results, a framework is proposed for participant recruitment. To make decisions on initiating and maintaining different types of recruitment strategies, the resources available and requirements of the research study (or studies) need to be carefully examined.

  • Background: The ability to identify, screen, and enroll potential research participants in an efficient and timely manner is crucial to the success of clinical trials. In the age of the internet, researchers can be confronted with large numbers of people contacting the program, overwhelming study staff and frustrating potential participants. Objective: This article describes a “do-it-yourself” recruitment support framework (DIY-RSF), using tools readily available in many academic research settings, to support remote participant recruitment, pre-screening, enrollment, and management across multiple concurrent eHealth clinical trials. Methods: This work was conducted in an academic research center focused on developing and evaluating behavioral intervention technologies. A needs assessment consisting of unstructured individual and group interviews was conducted to identify barriers to recruitment and important features for the new system. Results: We describe a practical and adaptable recruitment management architecture that used readily available software, such as REDCap and standard statistical software (e.g. SAS, R), to create an automated recruitment framework that supported prescreening potential participants, consent to join a research registry, triaging for management of multiple trials, capture of eligibility information for each phase of a recruitment pipeline, and staff management tools including monitoring of participant flow and task assignment/reassignment features. The DIY-RSF was launched in July 2015. As of July 2017, the DIY-RSF has supported the successful recruitment efforts for eight trials, producing 14,557 participant records in the Referral Tracking database and 5,337participants in the center research registry. The DIY-RSF has allowed for more efficient use of staff time and more rapid processing of potential applicants. Conclusions: Using tools already supported at many academic institutions, we describe the architecture and utilization of an adaptable referral management framework to support recruitment for multiple concurrent clinical trials. The DIY-RSF can serve as a guide for leveraging common technologies to improve clinical trial recruitment procedures.

  • Disrupted Care: eHealth engagement as a response to negative healthcare experiences

    Date Submitted: May 14, 2018
    Open Peer Review Period: May 15, 2018 - Jul 10, 2018

    Background: EHealth provides individuals with new means of accessing health information and communicating with providers through online channels. Prior evidence suggests that patients turn to eHealth to look for information online when they receive care that is low in patient centeredness. However, it is not clear how other problems with the healthcare delivery system motivate use of eHealth, how these problems relate to different kinds of eHealth activities, and who is most likely to turn to eHealth when they receive low quality care. Objective: We sought to determine (1) how two types of negative care experiences, low patient centeredness and care coordination problems, motivated use of different kinds of eHealth activities, and (2) whether highly educated individuals, who may find these tools easier to use, are more or less likely to make use of eHealth following negative experiences than less educated individuals. Methods: Using nationally representative data from the 2017 Health Information National Trends Survey (HINTS), we first used factor analysis to group 25 different eHealth activities into categories based on the correlation between respondents’ reports of their usage. We then used multivariate negative binomial generalized linear model regressions to determine whether negative healthcare experiences predicted greater use of these resulting categories. Finally, we stratified our sample based on educational attainment to test whether associations between healthcare experiences and eHealth use differed across groups. Results: Factor analysis suggested eHealth activities divide into two categories: provider-facing (e.g. facilitating communication with providers) and independent (e.g. patient-driven information seeking, communication with non-providers). Negative care experiences were not associated with provider-facing eHealth activity in the overall population or among highly educated respondents (P’s>.05), but respondents with lower education that had experienced problems with care coordination used somewhat more provider-facing eHealth (P=.07). Regarding independent eHealth activity, individuals engaged in more of these activities if they had experienced problems with either care coordination (P=.01) or patient centered communication (P=.01). Furthermore, while care coordination problems predicted independent eHealth activity across education levels (P=.01 for high education, P=.07 for low education), the relationship between low perceived patient centeredness and independent activity was constrained to individuals with lower levels of education (P=.02). Conclusions: Our findings indicate that individuals turn to eHealth activities--especially those that are independent of healthcare providers--when they experience problems with their healthcare. People with lower levels of education seem especially inclined to use eHealth in response to negative healthcare experiences. To maximize the potential for eHealth to meet the needs of all patients, and especially those traditionally underserved by the healthcare system, additional work should ensure eHealth resources are accessible and usable for all members of the population.

  • Electronic Health Record Portal Adoption- A New Integrated Model Approach

    Date Submitted: May 13, 2018
    Open Peer Review Period: May 15, 2018 - Jul 10, 2018

    Background: The future of health care delivery is becoming more patient focused and EHR portals are getting more attention from worldwide governments that consider this technology as valuable asset for the future sustainability of the national healthcare systems. Overall this makes the adoption of EHR Portals an important field to study. Objective: The goal of this study is to understand the factors that drive individuals to adopt EHR portals. Methods: We applied a new adoption model that combines three different theories, extended Unified Theory of Acceptance and Use of Technology (UTAUT2), Health Belief Model (HBM) and the Diffusion of Innovation (DOI), all of the three with relevant contributions for the understanding of EHR Portals. To test the research model, we used the partial least squares (PLS) causal modelling approach. We executed a national survey based on randomly generated mobile phone numbers. We collected 139 questionnaires. Results: Performance expectancy (β = .203; t = 2.699), compatibility (β = .530; t = 6.189) and habit (β =.251; t =2.660) have a statistically significant impact in behaviour intention (R2= 76.0%). Habit (β = .378; t =3.821), self-perception (β = .233; t = 2.971) and behaviour intention (β =.263; t =2.379) have a statistically significant impact in use behaviour (R2= 61.8%). Additionally, behaviour intention (β =.747; t =10.737) has a statistically significant impact in intention to recommend (R2= 69.0%), results demonstrability (β =.403 ; t =2.888 ) and compatibility (β =.337; t =2.243) have a statistically significant impact in effort expectancy (R2= 48.3%), and compatibility (β =.594 ; t = 6.141) in performance expectancy (R2= 42.7%). Conclusions: Our research model yields very good results, with relevant R2 in the most important dependent variables that help explain the adoption of EHR Portals, behaviour intention and use behaviour.

  • Background: Diabetes mellitus is a chronic metabolic disorder that results in abnormal blood glucose regulations. Blood glucose level (BG) is preferably maintained close to normality through self-management practices, which involves actively tracking BG levels and taking proper actions including adjusting diet and insulin medications. It is obvious that a tight BG control could sometimes results in out of range levels, i.e., hyper- or hypoglycemia episodes. Factors such as carbohydrate intake, stress and infections could result in hyperglycemia episodes. BG anomalies could be defined as any undesirable reading either due to a precisely known (normal cause variation) or unknown reason (special cause variation) to the patient. Even if the advancement in self-management applications and diabetes monitoring technologies have made things easier, the challenge of BG anomalies remains to be managed by the patient her-/himself. There are increasingly more technological developments in the direction of personalized decision systems and BG event alarms to provide an alert and decision support to the patient for these challenges. Techniques like detection of glycemic variability, hypoglycemia, and hyperglycemia in particular and blood glucose anomalies in general are central to the development of these diabetes technologies. The ubiquitous nature and widespread use of mobile health applications (mHealth apps), sensors and wearables and other point of care (POC) devices for self-monitoring and management purposes have made possible the generation of automated and continuous diabetes related data, which brought an opportunity for the introduction of machine learning applications for an intelligent and improved systems, capable of solving complex tasks within a dynamic knowledge and environment.Recently, machine learning applications have been widely introduced within the diabetes research in general and BG anomalies detection in particular. However, irrespective of their expanding and increasing popularity, there is lack of updated reviews that materialize the current trends in modelling options and strategies in BG anomalies detection within the context of personalized decision support systems and BG alarm events applications in type 1 diabetes. Objective: The objective of this review is to identify, assess and analyze the state of the art machine learning strategies and its hybrid systems focusing on blood glucose anomalies classification and detection including glycemic variability, hyperglycemia, and hypoglycemia in type 1 diabetes, which are important constituents for optimal diabetes self-management. The review covers machine learning approaches pertinent to personalized decision support systems and BG alarm events applications. Methods: A rigorous literature was conducted between September 1 and October 1, 2017, through various online databases including Google scholar, PubMed, ScienceDirect and others. Peer reviewed journals and articles were considered. Relevant articles were first identified by reviewing the title, keywords, and abstracts as a preliminary filter with our selection criteria, and then reviewed the full text articles that fulfilled the inclusion criteria. Information from the selected literature was extracted based on some predefined categories, which were based on previous research and further elaborated through brainstorming. Results: The initial hit was vetted using the title, abstract, and keywords, and retrieved a total of 467 papers (DBLP Computer Science (17), Diabetes Technology and Therapeutics (23), Google scholar (150), IEEE (211), Journal of Diabetes Science and Technology (21), PubMed Medlin (19), ScienceDirect (26)) (see Figure 3 below). After removing duplicates from the list, 389 records remained. Then, we did an independent assessment of the articles and screening based on the inclusion and exclusion criteria, which eliminated another 210 papers, leaving 179 relevant papers. After a full-text assessment, 45 articles were left (hyperglycemia = 5, glycemic variability = 3, hypoglycemia = 37), which were critically analyzed. The inter-rater agreement was measured using Cohen Kappa test, and disagreements were resolved through discussion. Conclusions: Despite the complexity of blood glucose dynamics, there are many attempts to capture hypoglycemia, hyperglycemia incidences and the extent of an individual’s glycemic variability using different approaches. Recently, due to the ubiquitous nature of self-management mHealth apps, sensors and wearables has paved the way for the continuous accumulation of self-collected health data, which in turn contributed for the widespread research of machine learning applications. The state of the art indicates that various classes of machine learning have been developed and tested in different BG pattern classification and anomalies detection tasks. These class includes feed forward artificial neural network, hybrid systems, support vector machine (SVM), decision tree, genetic algorithm, adaptive neural fuzzy inference system (ANFIS), nonlinear auto-regressive network with exogenous inputs (NARX), and nonlinear auto-regressive network (NAR), Gaussian process regression, deep belief network, and Bayesian neural network (BNN). These techniques have explored various kinds of input parameters such as blood glucose (BG), heart rate, QT interval, insulin, diet, physical activity, galvanic response, and skin impedance. Most of the identified studies used a theoretical threshold either suggest by physician or various concerned bodies like the American diabetes association. However, the problem with this kind of approaches is that the specified threshold may vary from patient to patient and also some patients might feel no symptoms. Therefore, such models should consider the difference among patients and also track its temporal change overtime. Moreover, the studies should also give more emphasis on the time lag (TL) and the various types of inputs used. Generally, we foresee these developments might encourage researchers to further develop and test these systems on a large-scale basis.

  • Detecting potential adverse drug reactions using a deep neural network model

    Date Submitted: May 10, 2018
    Open Peer Review Period: May 14, 2018 - Jul 9, 2018

    Background: Adverse drug reactions (ADRs) are common and they are the underlying cause of over a million serious injuries and deaths each year. The most familiar method to detect ADRs is relying on spontaneous reports. Unfortunately, the low reporting rate of spontaneous reports is a serious limitation of pharmacovigilance. Objective: Our objective was to identify a method to detect potential ADRs of drugs automatically using a deep neural network (DNN). Methods: We designed a DNN model that utilizes the chemical, biological, and biomedical information of drugs to detect ADRs. This model aimed to fulfil two main purposes, identifying the potential ADRs of drugs and predicting the possible ADRs of a new drug. For improving the detection performance, we distributed representations of the target drugs in a vector space to capture the drug relationships by using the word embedding approach to process amounts of biomedical literature. Moreover, we built a mapping function to address new drugs that do not appear in the dataset. Results: We predicted the ADRs of drugs recorded up to 2012, by using the drug information and the ADRs reported up to 2009. There were contained 746 drugs and 232 new drugs which only recorded in 2012 with 1,325 ADRs. The experimental results showed that the overall performance of our model with mean average precision (MAP) at top-10 is achieved 0.523 for ADR prediction on the dataset. Conclusions: Our model was effective in identifying the potential ADRs of a drug and the possible ADRs of a new drug. Most importantly, it can detect potential ADRs irrespective of whether they have been reported in the past.

  • Using the Google Maps to display the pattern of co-author collaborations in JMIR since 1999: A bibliometric study

    Date Submitted: Apr 13, 2018
    Open Peer Review Period: May 14, 2018 - Jul 9, 2018

    Background: The Google Maps have become increasingly attractive and the disease outbreak has been especially drawn on the Google Maps because of its easy-use characteristic to serve as a spatial representation with coordinates. The use of Google Maps in bibliometric studies is of particular interest in recent years due to the trend of the science evolution that can be known via the visual representations or a dashboard. Objective: The aim is to conduct a systematic evaluation of the literature on the feasibility and applicability of using Google Maps in bibliometrics and to assess the potential benefits and limitations of its application on the topic of co-author collaborations for an academic journal. Methods: The literature was searched for articles published in the Journal of Medical Internet Research (JMIR) since 1999. A total of 2,289 original research articles were extracted from the Medline library. The clusters of author nationalities, the productive authors, and the medical subject headings (MESH) terms were analyzed and distributed on the Google Maps. One-parameter Rasch model of continuous items was performed through cloud computation to examine the aberrant responses of count frequency in paper publications. We applied social network analysis (SNA), the Gini coefficient (GC) and the cluster coefficient(CC) to report study results with visual representations: (i) the trend of author collaboration in JMIR; (ii) the dominant nations and productive authors in JMIR; and (iii) the MESH terms to display the evolution of the journal features. Results: The trend of author collaboration in JMIR is increasing (=0.77) based on the number of authors per article. The mean number of individuals listed as authors in articles is 5.4. The most number of the 1st authors are from the U.S.( 819, 35.78 %), the European Netherlands (225, 9.83%), and the U.K. (199, 8.69%). The top two MESH terms are methods (1,141, 22.39%) and internet (844, 16.56%). The most productive author is Christensen, Helen (Australia) who published 27 articles in JMIR and had a significant network density (CC=0.45, t=7.14). The number of paper publications in nations has a close relation (=0.83) with the national GDP in 2016. Conclusions: There are promising feasibility and applicability of using Google Maps to serve as a spatial representation in bibliometrics or the disease outbreaks in the future. The social network analysis incorporated with the Google Maps provides wide and deep insights into the relationships among entities. The results can provide readers a concept map with the knowledge on the internet to know the JMIR characteristics.

  • Using social media to elicit support for arthritis-related health issues

    Date Submitted: May 13, 2018
    Open Peer Review Period: May 14, 2018 - May 24, 2018

    Background: People with arthritis are increasingly seeking support online, particularly for information about social role participation while experiencing symptoms of chronic arthritis. Social media enables support to be provided peer-to-peer on how hobbies and leisure activities can be adapted to allow participation. Research is needed to understand what type of peer support is provided online and how this support occurs. Objective: To describe how people support one another while experiencing arthritis-related issues that affect their ability to complete craft activities. Methods: Three vignettes were posted in a Facebook quilting group regarding arthritis-related symptoms or impairments that affect how people quilt. Facebook analytics were used to examine the groups’ demographics and responses were analysed thematically. Results: Members of the quilting Facebook group were mostly women (99.45%), aged 55 to 64 years and were located in the United States of America. As evident through the vignettes, participants predominantly offered emotional support and shared information. Participants shared their real-life experiences and creative means in adapting medical best practice to their crafting. More than half (55.6%) of the advice that was offered aligned with the OrthoInfo medical best practice guidelines relevant to the vignettes. Conclusions: Hobby-related social media groups can be important forums for facilitating emotional support and sharing information for arthritis-related issues. Health providers could support their patient’s recovery and lifestyles by stimulating and manoeuvring peer-to-peer conversations on similar hobby-related social media sites to ensure that shared advice aligns with medical best practice. Future research could explore the extent to which support provided online has the ability to adapt offline behaviours.

  • Basic Study on Automated Extraction of Symptom Related Words from Patient Complaints

    Date Submitted: May 12, 2018
    Open Peer Review Period: May 12, 2018 - May 24, 2018

    Background: Although methods of obtaining knowledge from texts written by healthcare professionals such as electronic medical records and discharge summaries have been studied, there are few reports analyzing free-text data on patients’ complaints in Japanese. Objective: This study aimed to establish a new method for extracting keywords from patients’ free descriptions accumulated in Japanese medical institutions. Methods: We developed a system that automatically annotates free-text data with the codes of the Tenth Revision of the International Statistical Classification of Diseases and Related Health Problems (ICD10) using electronic medication history data (target period: September 1, 2015 to August 31, 2016). The performance of the system was evaluated through comparisons with data manually annotated by healthcare workers. Results: The number of ICD10 codes extracted from 5,000 patient statements by healthcare workers was 2,348, while the system extracted 2,236 codes. Of those cases, 1,480 matched. Compared with manual extraction, the performance of the system was 0.66 in terms of precision, 0.63 in recall, and 0.65 for the F-measure. Conclusions: Our results suggested that the system was helpful for extracting and standardizing patient’s words related to symptoms from massive amounts of free-text data instead of manual work. After improving the extraction accuracy, we expect to utilize this system to detect the signals of adverse drug reactions from patients’ statements in the future.

  • The oncology patient and eHealth. How can new eHealth technologies be applied to cancer patient care?

    Date Submitted: May 9, 2018
    Open Peer Review Period: May 12, 2018 - Jul 7, 2018

    Background: Information and communication technologies (ICTs) in oncology can revolutionize the medical care of cancer patients. ICTs can promote patient’s empowerment and real-time disease monitoring. There is limited information about the impact of ICTs in cancer patients of ICTs in this patient group or indeed their level of interest in using these tools for greater management of their condition. Objective: We aimed to understand the ICT usage profile in hematology–oncology patients, to identify their needs, and to determine their level of interest in these technologies as a means of managing their disease. Methods: A 28-item questionnaire was drawn up by a multidisciplinary team including pharmacists and oncologists. The questions were organized into three blocks. A: socio-demographic characteristics; B: use of ICTs when searching for health-related information; and C: usage preferences for health apps. Hematology–oncology patients receiving treatment between May and July 2017 were included. A paper copy of the questionnaire was handed to patients in the Day Hospital or the Pharmaceutical Care Consultancy in Pharmacy Services. Results: A total of 650 questionnaires were handed out, with a participation of 94% (611/650). Patient socio-demographic characteristics: Mean age was 57.8 years-old [19–91]. Of participants, 61.9% (378/611) were women, 40.7% (249/611) had a university education, and 45.1% (276/611) of patients reported their overall state of health to be good. Use of ICTs when searching for health-related information: 87.1% (532/611) of subjects were interested in being informed about health-related matters. Of all subjects, 75.5% (532/611) sought information from health professionals and 61.3% (375/611) on the Internet. Of these, 71.2% (436/611) searched through Google and 21.9% (134/611) by means of social networks. Before going to their doctor’s appointment, 21.8% (133/611) of patients looked up information about their disease and/or treatment on the Internet. This access to the internet rose to 51% after their first medical appointment with their oncologist, up to 50.9% (311/611) of participants referred to the Internet. Usage preferences for health apps: 82.7% (505/611) had a smartphone, while 20.3% (124/611) had a health app installed. Some 81.5% (498/611) would use an app if their health professional recommended it to them, but 39.6% (242/611) were not willed to pay for it. Conclusions: The hematology–oncology patients showed a great deal of interest in searching for health-related information by means of ICTs, especially using smartphones and apps. The issues that drew the most interest in terms of apps were appointment management, advice on disease management, and communication with health professionals. Free access to these features and the recommendation by a health professional are important factors when it comes to their use. Therefore, the health care provider is a key element in the recommendation of ICTs, providing their knowledge and experience concerning their correct usage.

  • Background: Online and blended (face-to-face plus online) interventions for common mental health disorders are gaining in significance. However, many licensed psychotherapists still have guarded attitudes towards technology-aided formats, hindering dissemination efforts. Objective: This study provides a first therapist-centered evaluation of online and blended therapy and aims at identifying commonalities and differences in attitudes towards both formats. It further tests the impact of a short video clip on expressed attitudes. Methods: In total, 95 Austrian psychotherapists were contacted and surveyed via their listed occupational e-mail address. An eight-minute informational video clip was followed by two randomized survey blocks, assessing therapists’ attitudes towards online and blended therapy. Results: The sample resembled all assessed properties of Austrian psychotherapists (age, theoretical orientation, and region). Therapists did not hold a uniform overall preference. Instead, both intervention approaches’ advantages were perceived equally neutral (t(94)= 1.89, p= .063; d= 0.11), whereas online interventions were associated with more disadvantages and risks (t(94)= 9.86, p<.001; d= 0.81). The short informational clip did not excerpt any detectable effect on therapist views (r(95)= -.109 , p= 0.295). Personal use of modern technologies had a medium effect on given ratings and cognitive behavioral therapists tended to have more open-minded views than other therapists. Conclusions: This study is the first to directly compare attitudes towards online and blended therapy on a fine-grained level. Positive expectations play a pivotal role in the dissemination of new technologies, but unexperienced therapists seem to lack knowledge on how to benefit from technology-aided treatments. In order to speed up implementation, this aspects need to be addressed in the development of new interventions. Further, the preference of blended treatments over online interventions seems to relate to avoidance of risks. While this study is likely to represent therapist attitudes in countries with less advanced eHealth services, therapists’ attitudes in more developed countries might present differently.

  • Background: aaa Objective: to evaluate the effectiveness of a web-based personalized normative feedback (PNF) intervention in reducing alcohol use and negative consequences among college students with different motivation levels for receiving the intervention. Methods: : Parallel-group pragmatic randomized controlled trial among college drinkers aged 18 to 30 years from all Brazilian regions. Participants were randomized to a Control (assessment only) or PNF group - 1) drinking profile; 2) normative comparisons; 3) practical costs; 4) alcohol consequences; 5) strategies to decrease risks. Outcomes were: AUDIT scores (primary outcome), the maximum number of drinks, the number of consequences, the frequency of use and number of typical drinks. Motivation for receiving the intervention was assessed with a continuous scale (range=0-10) and students were categorized either as having low motivation (score <3) or being motivated (score ≥3). Students were followed after 1, 3 and 6 months (N=4,460). Generalized mixed models assessed intervention effects via two paradigms: observed cases and attrition models (multiple imputation and pattern-mixture). Results: Regardless of motivation, the PNF reduced typical drinks after 1 and 3 months and maximum drinks after one month (observed and attrition models), compared to the control. When considering motivation, motivated students (N=3,552) receiving PNF reduced the typical drinks after one (b=-0.51, 95CI:-0.76;-0.27, p<0.001), three (b=-0.60, 95CI:-0.91;-0.29, p<0.001) and six months (b=-0.58, 95CI:-0.93;-0.23, p=0.001), compared to control. Attrition models showed that effects were more robust after 1 and 3 months. In contrast, students with low motivation in the PNF group increased their AUDIT score after 6 months. There were no other significant effects. Conclusions: The web-based PNF for alcohol use among college students reduced only alcohol use among students, and motivation for receiving the intervention moderated the intervention effects. The intervention reduced the alcohol use only among motivated students, whereas low motivated students had an increased AUDIT score. Clinical Trial: NCT02058355. https://clinicaltrials.gov/ct2/show/NCT02058355

  • Background: Hypertension is a major cause of cardiovascular disease in older individuals. To ensure that blood pressure (BP) levels are in the optimal range, accurate BP monitoring is required. Contemporary hypertension clinical practice guidelines strongly endorse use of home BP measurement as a preferred method of BP monitoring for individuals with hypertension. The benefits of home BP monitoring may be optimized when measurements are telemonitored to care providers; however, this may be challenging for older individuals with less technological capabilities. Objective: The purpose of this qualitative study was to examine usability and acceptability of a home BP telemonitoring system with seniors. Methods: A qualitative descriptive study was conducted. Following a 1-week period of device use, an individual, semi-structured interview was conducted. Interview audio recordings were anonymized, de-identified, and transcribed verbatim. Thematic analysis was conducted on interview transcripts. Results: Seven seniors participated in usability testing of the home BP telemonitoring system. Participants were female (n=4) and male (n=3) with a mean age of 86 years (range 70 – 95 years). Overall, eight main themes were identified from the interviews: 1) positive features of the device; 2) difficulties or problems with device; 3) device was simple to use; 4) comments about wireless capability and components; 5) would recommend device to someone else; 6) would use device in future; 7) suggestions for improving the device; and, 8) assistance to use device. Additional sub-themes were also identified. Conclusions: Overall, the home blood pressure telemonitoring device had very good usability and acceptability with community dwelling seniors with hypertension. To enhance long-term use, a few improvements were noted that may mitigate some of the relatively minor challenges encountered by the target population.

  • Technology users’ experiences of remote rehabilitation as part of counseling planning: a qualitative study

    Date Submitted: May 8, 2018
    Open Peer Review Period: May 11, 2018 - Jul 6, 2018

    Background: Even though technology is becoming increasingly common in rehabilitation programs, insufficient data is available of rehabilitees’ perceptions and experiences as yet. Therefore, it is important to understand their abilities to use technology for remote rehabilitation. Objective: This is a qualitative study on technology experiences of persons affected by cardiovascular disease assessed before remote rehabilitation. The aim of the study was to explore rehabilitees’ experience with technology before 12 months remote rehabilitation. Methods: Qualitative interviews were conducted with 39 rehabilitees in four focus groups. The participants were aged 34−77 years (average 54.8 years) and male 74%, had coronary artery disease (CAD), and were undergoing treatment in a rehabilitation center. The interviews were conducted between September 2015 and November 2016. Data was analyzed using Glaser’s mode of the grounded theory approach. Results: The result of the study is an “identifying the e-usage” category, which relates to rehabilitees identifying their usage of information and communication technologies (ICT) (e-usage) in the process of behavior change. The category formed into four subcategories which define the rehabilitees’ technology experience. These categories are: feeling outsider, being uninterested, reflecting benefit, and enthusiastic using. All rehabilitees expected remote technology should to be simple, flexible, and easy to use and learn. The results reflecting their technology experience can be utilize in e-rehabilitation programs. Rehabilitees who feel outsiders and are not interested in technology need face-to-face communication for a major part of rehabilitation, while rehabilitees who reflect benefit and are enthusiastic about using technology need incrementally less face-to-face interaction and feel that web-based coaching provides sufficient support for rehabilitation. Conclusions: The findings show that persons affected by heart disease have had different experiences with technology and expectations of counseling, while all rehabilitees expected technology to be easy to use and their experiences to be smooth and problem-free. The results can also be used more widely in different contexts of social and health care for the planning of, and training in, remote rehabilitation counseling and education. Clinical Trial: Trial registration: ISRCTN Registry: ISRCTN61225589.

  • Correlation analysis among incidence and mortality of cancers and online public attention in China

    Date Submitted: May 8, 2018
    Open Peer Review Period: May 11, 2018 - Jul 6, 2018

    Background: Cancers pose a serious threat to health of Chinese people, resulting in a immense challenge to public health work. Nowadays, people could get related information not only from medical workers in hospitals, but also retrieve on Internet everywhere in real-time. Online search behaviors can reflect population’s awareness on cancers from a completely new perspective, which could be driven by underlying cancer epidemiology. Objective: The objective of our study was mainly to explore whether there was a correlation among incidence and mortaliy of cancers and online search behaviors. A secondary objective was to predict future incidence and mortality of theses cancers if the correlation was existed. Methods: We used standard Boolean operators to choose search terms of each cancer. We used Spearman’s rank correlation coefficient to validate the correlation among monthly search index values of each cancer type and their monthly incidence and mortality rates. ARIMA model and linear regression model were simultaneously used to predict incidence and mortality of these cancers in future. Results: Monthly search index values statistically significantly correlated with cancer incidence rates in 26 of 28 cancers in China, ( lung cancer: r=.80, P<.001; liver cancer: r=.28, P=.016; stomach cancer: r=.50, P<.001; esophageal cancer: r=.50, P<.001; colorectal cancer: r=.81, P<.001; pancreatic cancer: r=.86, P<.001; breast cancer: r=.56, P<.001; brain and nervous system cancer: r=.63, P<.001; leukemia: r=.75, P<.001; Non-Hodgkin lymphoma: r=.88, P<.001; Hodgkin lymphoma: r=.91, P<.001; cervical cancer: r=.64, P<.001; prostate cancer: r=.67, P<.001; bladder cancer: r=.62, P<.001; gallbladder and biliary tract cancer: r=.88, P<.001; lip and oral cavity cancer: r=.88, P<.001; ovarian cancer: r=.58, P<.001; larynx cancer: r=.82, P<.001; kidney cancer: r=.73, P<.001; squamous-cell carcinoma: r=.94, P<.001; multiple myeloma: r=.84, P<.001; thyroid cancer: r=.77, P<.001; malignant skin melanoma: r=.55, P<.001; mesothelioma: r=.79, P<.001; testicular cancer: r=.57, P<.001; basal-cell carcinoma: r=.83, P<.001). Meanwhile, monthly Baidu search index values statistically significantly correlated with cancer mortality rates in 24 of 27 cancers but not for stomach cancer, esophageal cancer and testicular cancer. Conclusions: Online search behaviors indeed reflect public awareness from a different angle. Research on Internet search behaviors may be an innovative and advance-time tool to monitor and estimate cancer incidence and mortality rates, especially for cancers not included in national registries.

  • Evidence-based evaluation of eHealth interventions: A systematic literature review

    Date Submitted: May 7, 2018
    Open Peer Review Period: May 11, 2018 - Jul 6, 2018

    Background: Until now, the use of technology in healthcare is driven mostly by the assumptions about the benefits of eHealth rather than the evidence of it. It is noticeable that the magnitude of evidence of effectiveness and efficiency of eHealth is not proportionate to the number of interventions that are regularly conducted. Reliable evidence generated through comprehensive evaluation of eHealth interventions may accelerate the growth of eHealth for long-term successful implementation and help to experience the benefits of eHealth in an enhanced way. Objective: The objective of this article was to understand how the evidence of effectiveness and efficiency of eHealth can be generated through evaluation. Hence, we aim to discern, (1) the phases of eHealth intervention as depicted by eHealth literature, (2) the aspects of eHealth intervention that are supposed to be evaluated according to eHealth literature, (3) how empirical studies about eHealth are evaluated, and (4) how evidence can be generated while evaluating eHealth interventions. Methods: A systematic literature review was conducted to explore the evaluation methods for eHealth interventions. PRISMA guidelines were followed. We searched Google Scholar and Scopus for the published articles that addressed the evaluation of eHealth or described an eHealth intervention study. A qualitative analysis of the selected articles was conducted in several steps. Results: We identified two distinct branches of conceptual framework that discussed the evaluation of eHealth interventions. The first branch identified evaluation as one of the phases of the intervention. The other phases identified through the literature review included design, pre-testing, pilot study, pragmatic trial, and post-intervention. The second branch provided the aspects of evaluation during intervention including clinical, human/social, organizational, technological, cost, ethical/legal, and transferability. The analysis of the empirical studies revealed that evaluation varied greatly and it may not be plausible to employ a standard evaluation procedure. Lastly, we proposed a conceptual model based on our findings that shows how the evidence of effectiveness and efficiency can gradually and reliably be built by evaluating specific aspects during specific intervention phases. Conclusions: The importance of evidence in eHealth has not been discussed as rigorously as have the diverse research approaches and evaluation frameworks. Further research directed towards evidence-based evaluation can not only improve the quality of intervention studies, but also facilitate successful long-term implementation of eHealth in general. We conclude that the development of more robust and comprehensive eHealth intervention studies or an improved validation of evaluation methods could ease the transferability of results among similar studies. Thus, the resources can be used to supplementary research in eHealth.

  • A Study of Web Page Understandability for Consumer Health Search

    Date Submitted: May 7, 2018
    Open Peer Review Period: May 11, 2018 - Jul 6, 2018

    Background: Understandability plays a key role in ensuring that people accessing health information are capable of gaining insights that can assist them with their health concerns and choices. The access to unclear or misleading information has been shown to negatively impact on the health decisions of the general public. Objective: We investigated methods to estimate the understandability of health Web pages and used these to improve the retrieval of information for people seeking health advice on the Web. Methods: Our investigation considered methods to automatically estimate the understandability of health information in Web pages, and it provided a thorough evaluation of these methods using human assessments as well as an analysis of pre-processing factors affecting understandability estimations, and associated pitfalls. Furthermore, lessons learnt for estimating Web page understandability were applied to the construction of retrieval methods with specific attention to retrieving information understandable by the general public. Results: We found that machine learning techniques were more suitable to estimate health Web page understandability than traditional readability formulae, which are often used as guidelines and benchmarking by health information providers on the Web (larger difference found for Pearson correlation of .602 using Gradient Boosting regressor compared to .438 using SMOG Index with CLEF 2015 collection). Learning to rank effectively exploited these estimates to provide the general public with more understandable search results (H_RBP^* reached 29.20, 22% higher than a BM25 baseline and 13% higher than the best system at CLEF 2016, both P≤.001). Conclusions: The findings reported in this article are important for specialised search services tailored to support the general public in seeking health advice on the Web, as they document and empirically validate state-of-the-art techniques and settings for this domain application. Clinical Trial: Not required.

  • Estimating Determinants of Attrition in Online Eating Disorder Community: An Instrumental Variables Approach

    Date Submitted: May 6, 2018
    Open Peer Review Period: May 9, 2018 - Jul 4, 2018

    Background: The use of social media as key health-information source has increased steadily among people affected by eating disorders. Intensive research has examined characteristics of individuals engaging in online communities, while little is known about discontinuation of engagement and the phenomenon of participants dropping out of these communities. Objective: This study aims to investigate characteristics of dropout behaviors among eating disordered individuals on Twitter and to estimate the causal effects of personal emotions and social networks on dropout behaviors. Methods: Using a snowball sampling method, we collected a set of individuals who self-identified with eating disorders in their Twitter profile descriptions, as well as their tweets and social networks, leading to 241,243,043 tweets from 208,063 users. Individuals’ emotions are measured from their language use in tweets using an automatic sentiment analysis tool, and network centralities are measured from users’ following networks. Dropout statuses of users are observed in a follow-up period 1.5 years later (from Feb. 11, 2016 to Aug. 17, 2017). Linear and survival regression instrumental variables models are used to estimate the effects of emotions and network centrality on dropout behaviors. An individual’s attributes are instrumented with the attributes of the individual’s followees (i.e., people who are followed by the individual). Results: Eating disordered users have relatively short periods of activity on Twitter, with one half of our sample dropping out at 6 months after account creation. Active users show more negative emotions and higher network centralities than dropped-out users. Active users tend to connect to other active users, while dropped-out users tend to cluster together. Estimation results suggest that users’ emotions and network centralities have causal effects on their dropout behaviors on Twitter. More specifically, users with positive emotions are more likely to drop out and have shorter-lasting periods of activity online than users with negative emotions, while central users in a social network have longer-lasting participation than peripheral users. Findings on users’ tweeting interests further show that users who attempt to recover from eating disorders are more likely to drop out than those who promote eating disorders as a lifestyle choice. Conclusions: Presence in online communities is strongly determined by individual’s emotions and social networks, suggesting that studies analyzing and trying to draw condition and population characteristics through online health communities are likely to be biased. Future research needs to examine in more detail the links between individual characteristics and participation patterns if better understanding of the entire population is to be achieved. At the same time, such attrition dynamics need to be acknowledged and controlled for when designing online interventions so as to accurately capture their intended populations.

  • Cyberchondria - Use of online health services in hypochondriacs compared to non-hypochondriacs.

    Date Submitted: May 5, 2018
    Open Peer Review Period: May 9, 2018 - Jul 4, 2018

    Background: An increasing number of people consult physicians because of worrying information found online. Cyberchondria refers to the phenomenon of health anxiety due to online health information. Objective: The present study aimed at examining Internet usage patterns of participants with and without symptoms of hypochondria and their impact on health anxiety as well as behavior. Methods: An online survey was conducted. Demographic data, health-related Internet use and general health behavior were assessed. The illness attitude scale was used to record symptoms of hypochondria. Results: The final sample consisted of N = 471 participants. More than 40% of participants showed at least some symptoms of hypochondria. As expected subjects with symptoms of hypochondria used the Internet more frequently for health related purposes and also frequented more various services than non-hypochondriacs. Most online health services were rated as more reliable by hypochondriacs. Reactions to online research such as doctor hopping or ordering non-prescribed medicine online were considered more likely by participants with hypochondria. Conclusions: Results contradict a substitution thesis, whereby users turn to online research as a result of a lack of alternatives, but rather show that intensive users of health-related services seem to consult them on- as well as offline.

  • Background: Evaluation of Massive Open Online Courses (MOOCs) is a necessary practice for appraising their application and analyzing their performance and benefits. Imperial College London in collaboration with Health iQ offered a 5-week MOOC Data Science Essentials: Real World Evidence. Objective: This study was conducted to evaluate the MOOC: Data Science Essentials: Real World Evidence, concerning its impact on learners knowledge, skills and attitudes and on the use of data science in healthcare. The aim of the evaluation was to investigate the success of the MOOC's objectives regarding ‘reach' about intended audience and social networks, ‘efficacy' about knowledge/skill gain skill and attrition, and adoption and sustainability of social networks for continual learning in this emerging field. Methods: The study design was a mixed-method evaluation drawing on semi-structured interviews of 2 learners transcribed and analysed using Braun and Clark's method for thematic coding. A Kirkpatrick evaluation was completed on two interview participants who could be assessed at all four levels of evaluation. Results: While the MOOC improved learners' knowledge of the topic, it could be improved regarding increasing communication and networking between learners. Participants reported increased knowledge but no change in skills and attitudes in the workplace after taking the MOOC. Lack of resources in the workplace and the change to a different workplace, which changed participant's responsibilities were barriers identified by not experiencing a change in skills and attitudes as a result of taking the MOOC. Conclusions: The are two reasons that may explain why the MOOC while improving participants’ learning, was not able to change participants’ attitudes and use of data science in the workplace. Since the MOOC offered was about a brand-new topic (Real World Evidence) that requires the availability of specific resources, participants were not able to apply the learning since these resources were still not available in the workplace. Also, having a job that requires a different set of skills than the ones taught in the MOOC was another reason that prevented a participant from applying the learning from the MOOC. Understanding the MOOC learners’ experience, learning level and influence on behaviour change can potentially improve future MOOCs to become more effective for learning and changing behaviour. Clinical Trial: The evaluation received ethical approval from the Imperial College Education Ethics Review Process (EERP).

  • Increasing Need for an eHealth Impact Framework for Africa

    Date Submitted: May 4, 2018
    Open Peer Review Period: May 9, 2018 - Jul 4, 2018

    Background: Efforts by African countries to improve healthcare and citizens’ health within challenging resource limitations put considerable pressure on investment decisions. eHealth is believed to have potential to strengthen health systems and is now seen by some entities as a determinant of health, yet its impact is not well estimated in advance or measured after implementation. This is an obstacle to the effective allocation of healthcare resources and limits national and regional abilities to develop eHealth and realise its potential in health systems transformation. Objective: A broad conceptual framework for appraising eHealth impact could help selection of the best course of action by informing individual steps, including: identification of promising initiatives, choosing between eHealth options, motivating for financing, managing implementation, preparing for obsolescence, analysing achievements, and promoting sustainability. Such insight could help strengthen and transform healthcare, provide benchmarks for oversight and Monitoring and Evaluation (M&E), and improve individual and population health. Such frameworks exist, but do not address the needs of African countries. Methods: Desktop literature review to explore African countries’ eHealth context and the need for an eHealth economic appraisal framework. Results: Unique aspects of Africa’s health context are elucidated, particularly relating to resource constraints and health strengthening priorities. The influence of these aspects on eHealth development, and how to address factors such as affordability, value for money and risk, are highlighted. Key eHealth appraisal terms are clarified. Conclusions: The paper presents the role of eHealth impact appraisal in appraising eHealth options for African countries and describes aspects critical for an eHealth impact framework for Africa.

  • Introduction: Trials of digital behavior change interventions (DBCIs) can yield extensive, in-depth usage data, yet usage analyses tend to focus on broad descriptive summaries of how an intervention has been used by the whole sample. This paper proposes a novel framework to guide systematic, fine-grained usage analyses that better enables understanding of how an intervention works, when and for whom. Framework description: The framework comprises three stages to assist: 1) familiarisation with the intervention and its relationship to the captured data; 2) identification of meaningful measures of usage and specifying research questions to guide systematic analyses of usage data; 3) preparation of datasheets, and consideration of available analytical methods with which to examine the data. Framework application: The framework can be applied to inform data capture during the development of a DBCI and/or in the analysis of data after the completion of an evaluation trial. We will demonstrate how the framework shaped preparation and aided efficient data capture for a DBCI to lower transmission of cold and flu viruses in the home, and informed a systematic in-depth analysis of usage data collected from a separate DBCI designed to promote self-management of colds and flu. Conclusions: The AMUsED framework guides systematic and efficient in-depth usage analyses that will support standardized reporting with transparent and replicable findings. These detailed findings will also enable examination of what constitutes effective engagement with particular interventions.

  • Background: Noncommunicable diseases (NCD) account for more than 18 million deaths in women annually, with many of these deaths being attributed to modifiable risk factors like physical inactivity. Women perceive a range of benefits and barriers to exercise, however, there is little evidence about the effect of different lifestyle intervention delivery modes on perceptions of exercise. Objective: This study compares the effect of a multiple health behavior change intervention called the Women’s Wellness Program or WWP. This intervention was delivered in three different modes on perceived exercise benefits, perceived exercise barriers, and actual physical activity and exercise in midlife women. Methods: Women aged 45 to 65 years were recruited via the study website. Women were assigned in blocks to three different treatment groups (A. online independent; B. face-to-face with nurse consultations, and; C. online with virtual nurse consultations). All participants received the 12 week intervention that utilises principles from social-cognitive theory to provide a structured guide to promote healthy lifestyle behaviours with an emphasis on regular exercise and healthy eating. Data were collected using self-report online questionnaire at baseline (T1) and post intervention (T2) including perceived exercise benefits and barriers (EBBS) and exercise and physical activity (SPA). Data analysis examined both within and between groups changes over time. Results: Participants in this study (n = 225) had a mean age of 50.9 years ± 5.9 and most were married or living with a partner (83.3%, n = 185). Attrition was 30.2% with 157 participants completing the final questionnaire. Women in all intervention groups reported a significant increase in positive perceptions of exercise (p < .05); a significant increase in exercise and overall physical activity (p > .01), with moderate to large effect sizes noted for overall physical activity (d = .5 - .87). Participants receiving support from registered nurses in the face to face and online groups, had a greater magnitude of change in benefits perceptions and physical activity compared to the online independent group. There was no significant change in exercise barriers perceptions within or between groups over time. Conclusions: Results of this study suggest the multiple health behaviour change intervention can be effective in increasing exercise benefits perceptions, overall physical activity and exercise in midlife women. While web-based programs are cost effective, flexible and can be delivered remotely, providing a range of options including face to face group delivery and personalised e-health coaching from registered nurses, has potential to enhance participant engagement and motivation. Clinical Trial: ANZCTR: ACTRN12616001189415

  • Background: OpenPrescribing is a freely accessible service that enables any user to view and analyse NHS primary care prescribing data at the level of individual practices. This tool is intended to improve the quality, safety, and cost-effectiveness of prescribing. Objective: We set out to measure the impact of OpenPrescribing being viewed on subsequent prescribing. Methods: Having pre-registered our protocol and code, we measured three different metrics of prescribing quality (mean percentile across 34 existing OpenPrescribing quality measures, available “price-per-unit” savings, and total “low-priority prescribing” spend) to see if they changed after CCG and practice pages were viewed. We also measured whether practices whose data were viewed on OpenPrescribing differed in prescribing, prior to viewing, to those who were not. We used fixed effects and between effects linear panel regression, to isolate change over time and differences between practices respectively. We adjusted for month of prescribing in the fixed effects model, to remove underlying trends in outcome measures. Results: We found a reduction in available price-per-unit savings for both practices and CCGs after their pages were viewed. The saving was greater at the practice level (-£40.42 per thousand patients per month, 95% confidence interval -54.04 to -26.01) than at CCG level (-£14.70 per thousand patients per month, 95% confidence interval -25.56 to -3.84). We estimate a total saving since launch of £243k at practice level and £1.47m at CCG level between the feature launch and end of follow-up (August to November 2017) among practices viewed. If the observed savings from practices viewed were extrapolated to all practices, this would generate £26.8m in annual savings for the NHS, approximately 20% of the total possible savings from this method. The other two measures were not different after CCGs/practices were viewed. Practices which were viewed had worse prescribing quality scores overall, prior to viewing. Conclusions: We found a clinically significant positive impact from use of OpenPrescribing, specifically for the class of savings opportunities that can only be identified by using this tool. We also show that it is possible to conduct a robust analysis of the impact of such an online service on clinical practice.

  • Economic evaluation of an internet-based stress-management intervention alongside an RCT

    Date Submitted: May 4, 2018
    Open Peer Review Period: May 9, 2018 - Jul 4, 2018

    Background: Work-related stress is widespread among employees and associated with high costs for the society. Internet-based stress management interventions (iSMI) are effective in reducing such stress. However, evidence for their cost-effectiveness is scant. Objective: The aim of the study was to evaluate the cost-effectiveness (CEA) and cost-utility (CUA) of a guided iSMI with mobile support for employees. Methods: A sample of 264 employees with elevated symptoms of perceived stress (Perceived Stress Scale, PSS-10 ≥ 22) was assigned to either the iSMI or to a waitlist control condition (WLC) with unrestricted access to treatment as usual. Participants were recruited in 2013, followed through 2014, and data analyzed in 2017. The iSMI consisted of 7 sessions plus one booster session and was based on problem-solving and emotion regulation techniques. Costs were measured from the societal perspective, in-cluding all direct and indirect medical and non-medical costs. Costs were related to symp-tomatic remission in the CEA and to gains in quality adjusted life years (QALYs) in the CUA. Statistical uncertainty was handled using bootstrapping (N=5000). Results: At a willingness-to-pay (WTP) ceiling of €0 for becoming free of self-perceived stress symptoms, there was a 70% probability of the intervention being more cost-effective than WLC. This rose to 85% and 93% when society is willing to pay €1000 and €2000 achieving symptomatic remission, respectively. The CUA yielded a 76% probability of the interven-tion being more cost-effective than WLC at a conservative WTP threshold of €20,000 ($25,800) per QALY gained. Conclusions: From the societal perspective, offering iSMI to stressed employees has an acceptable likeli-hood of being cost-effective compared to WLC. Clinical Trial: German Clinical Trials Register (DRKS): 00004749; http://drks-neu.uniklinik-freiburg.de/ drks_web/setLocale_EN.do (Archived by WebCite at http://www.webcitation.org/6e8rl98nl )

  • Background: Patients’ privacy is regarded as essential for the patient-doctor relationship. One example for a privacy-enhancing technology for user-controlled data minimization on content level is a redactable signature. It enables users to redact personal information from signed documents while preserving the validity of the signature and thus the authenticity of the document. In this paper, we present end users’ evaluations of a Selective Authentic EHR Exchange Service (SAE-service) in an eHealth use-case. In the use-case scenario, patients are given control to redact specified information fields in their electronic health records, which were signed by their doctors with a redactable signature and transferred to them into a cloud platform. They can then selectively disclose the remaining information in the EHR, which still bears the valid digital signature, to third parties of their choice. Objective: To explore the perceptions, attitudes and mental models concerning the SAE-service of both medical professionals and patients with different backgrounds. Another objective is to elicit usability requirements for this service based on our analysis. Methods: We chose empirical qualitative methods to address our research objective. Designs of mock-ups for the service were used as part of our User-centered Design approach in our studies with test participants from Germany and Sweden. Thirteen individual walk-throughs/interviews were conducted with medical staff to investigate the EHR signers’ perspectives. Moreover, five group walk-throughs in focus groups sessions with (N= 32) prospective patients with different technical knowledge to investigate redactor’s perspective of EHR data redaction control were used. Results: Our study participants showed to have correct mental models in regard to the redaction process. Users with some technical models lacked trust in the validity of the doctor’s signature on the redacted documents. Main results to be considered are the requirements concerning the accountability of the patients’ redactions and the design of redaction templates for guidance and control. Conclusions: For the SAE-service to be means for enhancing patient control and privacy, the diverse usability and trust factors of different user groups should be considered.

  • Background: Most chemotherapy is administered in the outpatient setting; therefore, patients are required to manage related side effects at home without direct support from health professionals. The Advanced Symptom Management System (ASyMS) has been developed to facilitate the remote monitoring and management of chemotherapy-related toxicity in patients with cancer, using standardized patient-reported outcomes (PRO) questionnaires. Objective: Our goal was to evaluate usability of this mobile phone-based technology from the perspective of Canadian cancer patients receiving chemotherapy where the objectives were to elicit their views, experiences and satisfaction with ASyMS and identify existing design, functionality, and usability issues. Methods: A mixed-method approach to data collection with user-based testing, think aloud technique, semi-structured interviews, and short answer questionnaires was used with 10 cancer patients. Patients attended usability testing sessions at the Centre for Global eHealth Innovation, UHN. Participants performed specific tasks on the ASyMS device. The test was video recorded and each task was timed during the test. After the usability sessions, patients completed the post-test questionnaire and participated in a semi-structured interview. Thematic analysis was used to code and categorize the identified issues according to the type and frequency of occurrence. Also, a set of variables related to the participant’s performance including the number of errors each participant made, requests for help, the time taken to complete the task, participant feedback, and reviewing the videos was used to identify a list of usability issues. Results: The thematic analysis generated three over-arching themes: (1) ASyMS user- friendliness, (2) Usefulness of the ASyMS (content quality & richness), and (3) Intention to use. Results from the post-test questionnaire indicate that 80% of participants (n=8) had great motivation to use the ASyMS, 70% (n=7) had positive perceptions of the successful use of the ASyMS, and all (n=10) had positive attitude towards using the ASyMS in future. Most identified design and functionality issues are related to the navigation of the ASyMS, using more advanced and attractive design with better functionality and features. Conclusions: The ASyMS has shown positive perceptions by patients; however, evaluation of the ASyMS through a trial is needed. Clinical Trial: N/A

  • Background: The effect electronic health record (EHR) implementation has on physician satisfaction and patient care remains unclear. A better understanding of physicians’ perceptions of EHR’s and factors that influence those perceptions is needed to improve the physician and patient experience when using EHR’s. Objective: The objective of this study was to determine provider and clinical practice factors associated with physicians’ EHR satisfaction and perception of patient impact. Methods: We surveyed a random sample of physicians, including residents and fellows, at a quaternary care academic hospital from February to March 2016. The survey assessed provider demographics, clinical practice factors, and overall EHR experience. The primary outcomes assessed were providers’ satisfaction and the providers’ perception of impact to patient care. Responses on the satisfaction and patient impact questions were recorded on a continuous scale initially anchored at neutral (scale range 0-100 with 0 defined as extremely negatively and 100 as extremely positively). Independent variables assessed included demographic and clinical practice factors including perceived efficiency in using the EHR. One-way analysis of variance or the Kruskal-Wallis Test was used for bivariate comparisons and linear regression was used for multivariable modeling. Results: 111 of 157 physicians (71%) completed the survey. 51% of the respondents were attending physicians, and of those, 72% reported a > 50% clinical FTE and half reported supervising residents >50% of the time. 36% of the respondents were primary care practitioners and previous EHR experience was evenly distributed. 13% of the total sample were EHR super users. Responses to how our current EHR affects satisfaction were rated above the neutral survey anchor point (mean=58, std. dev 22) as were their perceptions as to how the EHR impacts the patient (mean=61, std. dev 18). In bivariate comparisons, only physician age, clinical role (resident, fellow, or attending), and perceived efficiency were associated with EHR satisfaction. In the linear regression models, physicians with higher reported perceived efficiency reported higher overall satisfaction and patient impact after controlling for other variables in the model. Conclusions: Physician satisfaction with EHRs and their perception of its impact on clinical care were generally positive but physician characteristics, greater age and attending physician, were associated with worse EHR satisfaction. Perceived efficiency is the factor most associated with physicians’ satisfaction with EHRs when controlling for other factors. Understanding physician perceptions of EHRs may allow targeting of technology resources to ensure efficiency and satisfaction with EHR system use during clinical care.

  • ICUTogether a web based recovery program for intensive care survivors: a randomized controlled trial protocol

    Date Submitted: May 2, 2018
    Open Peer Review Period: May 5, 2018 - Jun 30, 2018

    Background: Those who experience a critical illness/event requiring admission to an intensive care unit frequently experience physical and psychological complications as a direct result of their critical illness/event and intensive care unit experience. Complications, if left untreated, can affect the quality of life of survivors and impact upon health care resources. Explorations of potential interventions to reduce the negative impact of an intensive care unit experience have failed to establish an evidence based intervention. Objective: The aim of this study is to evaluate the impact of a web based intensive care recovery program on the mental well-being of intensive care survivors and to determine if it is a cost effective approach. Methods: One hundred and sixty-two patients that survived an intensive care unit experience will be recruited and randomized into one of two groups. The intervention group will receive access to the web based intensive care recovery program, ICUTogether, two weeks after discharge (n=81) and the control group will receive usual care (n=81). Mental well-being will be measured using the Hospital Anxiety and Depression Scale, The Impact of Events Scale-Revised and the EQ-5D at three time points (2 weeks, 6 months and 12 months post discharge). Analysis will be conducted on an intention to treat basis using regression modelling. Covariates will include baseline outcome measures, study allocation (intervention or control), age, gender, length of intensive care unit stay, Apache III score, level of family support and hospital readmissions. Participants’ evaluation of the mobile website will be sought at 12 months post discharge. A cost utility analysis conducted at 12 months from a societal perspective will consider costs incurred by individuals as well as health care providers. Discussion: This study will evaluate a novel intervention in a group of intensive care unit survivors. The findings from this study will inform a larger study and the wider debate about an appropriate intervention in this population. Trial Registration: Australian New Zealand Clinical Trial Registry: ACTRN12618000252213

  • Background: Individuals diagnosed with chronic obstructive pulmonary disease (COPD) live with the burden of progressive life-threatening condition that is often accompanied by anxiety and depression. The severity of the condition is usually considered from a clinical perspective, and characterized according to the GOLD classification of severity that focuses on increasing symptoms and exacerbations (A through D) whereas information about perceived health or ability to manage the condition are rarely objectively included. Objective: To evaluate three patient-reported outcome measurements (PROMs) to understand how these can be used to report on individuals living with COPD supported by an innovative digitally assisted solution – the Epital, that aims to increase patient’s understanding and management of their condition to improve their wellbeing. Methods: A total of 93 individuals with COPD were enrolled. At entry and after six and twelve months, heiQ, HLQ and SF-36 PROMs were collected. The scores of the 19 PROM dimensions were related to COPD severity i.e. GOLD risk factor group, pulmonary function at entry and number of exacerbations over a three month period. Results: Only three dimensions showed significant differences between GOLD risk factor groups; heiQ3. Emotional distress (D and C, p<0.05), heiQ5. Constructive attitudes and approaches (D and C, p<0.01) and SF36 Physical health (D and A, p<0.001; D and B p<0.01), indicating more distress, poorer attitudes and poorer health status with increasing severity. heiQ3. Emotional distress was higher in group C than the other groups indicating less distress for this group. A similar pattern was seen in the SF36 Mental health score. As expected, pulmonary function (FEV1) was negatively associated with the severity of the condition (GOLD risk factor group) FEV1 was associated with increased heiQ2. Positive and active engagement in life (p<0.05), decreased heiQ3. Emotional distress (p<0.05), increased heiQ5. Constructive attitudes and approaches (p<0.001), increased heiQ8. Health service navigation (p<0.05), and also increased in SF36 Physical health (p<0.001). Furthermore, there was a negative association between FEV1 and HLQ 3. Actively managing my health (p<0.05). Only HLQ 8. Ability to find good health information was independently associated with reduced number of contacts over three months. After six months we observed an increase of heiQ 6. Skills and technical acquisition and a reduction in in heiQ 3. Emotional distress. The latter effect persisted after twelve months as did heiQ 4. Self-monitoring and insight had also increased. HLQ 3. Actively managing my health decreased after six and twelve months. Conclusions: The examined PROMs provides health professionals with new insights into individual’s ability to understand and manage their condition. In contrast to the Gold severity index, the PROMs did not predict number of contacts to a 24/7 available response and coordination center.

  • Background: Introduction : Needles are frequently required for routine medical procedures. Children with severe hemophilia can have hundreds of IV procedures per year. Fear of needle related procedures can cause avoidance of future health care. Virtual reality (VR) has shown promise as a distraction technique during procedures, but barriers to commercially available VR platforms have prevented widespread use. Objective: Objective : We hypothesized that we could create a VR platform that would be customized for pediatric hemophilia care, clinician orchestrated and feasible to use for distraction during intravenous (IV) procedures performed as a part of complex healthcare. Methods: Methods: We created a custom VR platform that consisted of a wireless, adjustable, disposable headset and a suite of VR games that could be orchestrated remotely. The platform was customized for a pediatric hemophilia population. Orchestration was performed by a hemophilia nurse observing the procedure. The primary endpoint of the trial was to assess feasibility of VR versus Standard-of-care (SOC) distraction during a planned IV procedure in the course of a comprehensive hemophilia clinic visit. Feasibility and usability of the platform was assessed in a single-center, randomized clinical trial from June through December 2016. Participants were children with hemophilia age 6 to 18 years. After informed consent 25 patients enrolled and were randomized. Each subject, 1 caregiver and 1 hemophilia nurse orchestrator assessed degree of pre-procedural nervousness/anxiety with an anchored, combined modified Visual Analog (VAS)/FACES scale prior to the IV procedure. Each participant then underwent a timed IV procedure with either VR or SOC distraction. Each rater assessed the distraction methods using a VAS/FACES scale at the completion of the IV procedure with questions targeting usability, engagement, impact on procedural anxiety, impact on procedural pain, and likability of the distraction technique. Participants, caregivers and nurses were also asked to rate how much they would like to use VR for future procedures. To compare the length of procedure time between the groups, the Mann-Whitney test was used. Results: Results: 25 children were enrolled and 24 were included in the primary analysis. The median procedure time was 10 minutes in the VR group (range 1-31 minutes) and was comparable to 9 minutes (range: 3-20 minutes) in the SOC arm (p=0. 764). Patients in both VR and SOC arms reported a positive influence of distraction on procedural anxiety and pain. Eighty percent of the children, caregivers and nurses reported that they would like to use VR for future procedures. No adverse events were experienced in either arm. Conclusions: Conclusion: We demonstrated that an orchestrated, VR environment could be integrated and used during pediatric hemophilia care during IV interventions without adversely disrupting clinic flow. This platform has the potential to improve the patient experience during medical procedures. Clinical Trial: Clinical Trials.gov NCT03507582

  • Crossing the digital divide in online self-management support: analysis of usage data from HeLP-Diabetes.

    Date Submitted: May 1, 2018
    Open Peer Review Period: May 2, 2018 - Jun 27, 2018

    Background: Digital health is increasingly recognised as a cost-effective means to support patient self-care. However there are concerns about whether the “digital divide”, defined as the gap between those who do and do not make regular use of digital technologies will lead to increased health inequalities. Access to the internet, computer literacy, motivation to use digital health interventions, and fears about internet security are barriers to use of digital health interventions. Some of these barriers disproportionately affect people of older age, black or minority ethnic background, and low socioeconomic status. HeLP-Diabetes, a theoretically informed online self-management programme for adults with type 2 diabetes, was developed to meet the needs of people from a wide demographic background. Objective: The aim of the study was to determine whether there was evidence of a digital divide when HeLP-Diabetes was integrated into routine care, by comparing the characteristics of people who registered for the programme against the target population (people with type 2 diabetes in inner London); comparing the characteristics of people who registered for the programme and used it with those who did not use it; and comparing sections of the website visited by different demographic groups. Methods: A retrospective analysis of data on the use of HeLP-Diabetes in routine clinical practice in four inner London CCGs was undertaken. Data were collected from patients who registered for the programme as an NHS service. Data on gender, age, ethnicity and educational attainment were collected at registration, and data on webpage visits (user identification number, and date, time and page visited) were collected automatically by software on the server side of the website. Results: The characteristics of people who registered for the programme were found to reflect those of the target population. The mean age was 58 years (SD=30), over 50.0% were from black and minority ethnic backgrounds, and nearly a third (29.8%) had no qualifications beyond school leaving age. There was no association between demographic characteristics and use of the programme, apart from weak evidence of less use by the mixed ethnicity group. There was no evidence of differential use of the programme by any demographic group, apart from weak evidence for people with degrees and school leavers being more likely to use the “Living and working with diabetes” (p=0.03) and “Treating diabetes” (p=0.04) sections of the website. Conclusions: This study is one of the first to provide evidence that a digital health intervention can be integrated into routine health services without widening health inequalities. The relative success of the intervention may be attributed to integration into routine health care, and careful design with extensive user input and consideration of literacy levels. Developers of digital health interventions need to acknowledge barriers to access and use, and collect data on the demographic profile of users, in order to address inequalities. Clinical Trial: N/A

  • Background: Internet-based cognitive behavioural therapy (iCBT) is a promising new treatment format for depression and anxiety. However, it is important to establish whether results can be transferred into routine care before implementing it on a large scale. While ample studies have demonstrated its efficacy under controlled conditions, fewer studies have investigated its effectiveness in routine care. Furthermore, the picture is blurred, with treatment effects in routine care ranging from none to very large. Objective: The objective of the present study was to evaluate the clinical effectiveness of iCBT for depression and anxiety in routine secondary care. Methods: A retrospective cohort design was used analysing N=203 patients treated for either depression or anxiety in a dedicated iCBT clinic in secondary care in Denmark. Patients were measured before commencing treatment and subsequently weekly. Baseline and last observation were used as pre- and post-intervention time points. Outcome measures were the Patient Health Questionnaire-9 and the Generalised Anxiety Disorder-7 scale for the two diagnoses respectively. Primary analyses were conducted using a linear mixed-effects model. Secondary analyses were conducted using baseline characteristics as predictors (gender, age, highest level of education, occupational status, marital status, psychotropic medication use, consumption of alcohol and leisure drugs). Additionally, logistic regressions were performed to predict non-completion of treatment. Results: The primary analyses revealed large and significant reductions in symptomatic levels of depression (ß=-6.27, SE=0.83, p<.001, d=1.0) and anxiety (ß=-3.78, SE=0.43, p<.001, d=1.1). In all, 62% and 40%, respectively, completed depression and anxiety treatment. Baseline severity of primary disorder was associated with higher treatment gains, which for anxiety patients also predicted higher risk of non-completion. Higher baseline severity of comorbid disorder increased risk of non-completion for both disorders. Being unemployed predicted higher risk of non-completion for depression OR=7.50, CI=1.40 – 59.25, p=.028. Time spend with the program predicted a slight increase in improvement for depressed patients ß=-0.03 SE=0.01, p=.012, but not for anxiety patients. Conclusions: The results support the hypothesis that iCBT can be an effective treatment when implemented in routine secondary care, in line with previous cohort studies conducted by similar clinics. The results are even comparable to within-group results found in efficacy and effectiveness studies. Non-completion rates of 38% for depressed patients and 60% for anxiety patients resemble those observed in psychotherapy in general but were higher than those seen in similar clinics. Multiple factors predicted outcome and non-completion. However, all predictor effects were statistically relatively weak. The main results of the present investigation are important findings, since depression and anxiety are costly and debilitating disorders that are vastly under-treated. These findings support that ICBT may be able to help bridge this gap between the need for treatment and its provision.

  • Background: Ecological Momentary Assessment (EMA) can be a useful tool for collecting real time behavioral data in studies of health and health behavior. However, EMA administered through mobile technology can be burdensome and it tends to suffer from suboptimal user engagement, particularly in low health literacy populations. Objective: To report a case study involving the design and evaluation of a mobile EMA tool that supports context-sensitive EMA reporting of location and social situations accompanying eating and sedentary behavior. Methods: An iterative, user-centered design process with obese, middle-aged women seeking care in a safety-net health system was used to identify the preferred format of self-report measures, and the look, feel and interaction of the mobile EMA tool. A single arm feasibility field trial with 21 participants receiving 12 prompts each day for momentary self-reports over a 4-week period (336 total prompts per participant) was used to determine user satisfaction with interface quality and user engagement, operationalized as response rate. A second trial among 38 different participants randomized to receive or not a feature designed to improve engagement was conducted. Results: The feasibility trial results showed high interface satisfaction and engagement, with an average response rate of 50% over 4 weeks. Qualitative feedback pointed to the need for auditory alerts. We settled on three alerts at 10-minute intervals to accompany each EMA reporting prompt. The second trial testing this feature showed a statistically significant increase in response rate among participants randomized to receive repeat auditory alerts versus those not (60% vs 40%). Conclusions: This paper reviews the design research and a set of design constraints that may be considered in the creation of mobile EMA interfaces personalized to users’ preferences. Novel aspects of the study include the involvement of low health literacy adults in design research, the capture of data on time, place, and social context of eating and sedentary behavior, and reporting prompts tailored to an individual’s location and schedule. Clinical Trial: NCT03083964

  • Creating engaging health promotion campaigns on social media: observations and lessons from Fitbit and Garmin

    Date Submitted: Apr 30, 2018
    Open Peer Review Period: Apr 30, 2018 - Jun 25, 2018

    Background: The popularity and reach of social media make it an ideal delivery platform for interventions targeting health behaviours, such as physical inactivity. Research has identified a dose-response relationship whereby greater engagement and exposure are positively associated with intervention effects, hence enhancing engagement will maximise the potential of these interventions. Objective: This study examined the social media activity of successful commercial activity tracker brands to understand which creative devices they use in their communication to engage current and potential users, which platforms attract the most engagement, and which creative devices prompted the most engagement. Methods: Posts (n=509) made by Fitbit and Garmin on Facebook, Twitter and Instagram over a 3-month period were coded for the presence of creative devices. User engagement in terms of the total number of likes, comments or shares per post was recorded. Results: Posts (n=509) made by Fitbit and Garmin on Facebook, Twitter and Instagram over a 3-month period were coded for the presence of creative devices. User engagement in terms of the total number of likes, comments or shares per post was recorded. Conclusions: Future work should prioritise Instagram as a delivery platform and incorporate principles of market segmentation, or tailoring. Health messages on social media should be clear, refreshed regularly, incorporate inspirational messages and imagery and be focussed on tangible end products of health to maximise engagement and therefore the potential of this approach for positive behaviour change.

  • Transition education for young adults with type 1 diabetes: Feasibility of a group telehealth intervention

    Date Submitted: Apr 30, 2018
    Open Peer Review Period: Apr 30, 2018 - Jun 25, 2018

    Background: Young adults with type 1 diabetes (T1D) experience a decline in glycemic outcomes and gaps in clinical care. A diabetes education and support program designed for young adults was delivered via group videoconference and mobile web. Objective: To assess the feasibility, acceptability, and preliminary efficacy of the program as measured by attendance and web page views, satisfaction, and pre-/post intervention psychosocial outcomes respectively. Methods: Young adults ages 18-25 years were recruited to attend five 30-minute group diabetes education videoconferences during an 8-week time period. Videoconferences included an expert presentation followed by a moderated group discussion. Within 48 hours of each videoconference, participants were sent a link to more information on the study website. Feasibility was assessed using data on videoconference attendance and web page views. Acceptability was assessed via satisfaction survey completed at the conclusion of the study. Descriptive statistics were generated. Preliminary efficacy was assessed via survey as changes in diabetes-specific self-efficacy and diabetes distress. Pre- and post-intervention data were compared using paired samples t-tests. Results: Twenty young adults (mean age=19.2 years, SD=1.1), attended an average of 5.1 (SD=1.0) videoconferences equivalent to 153 minutes (SD=30.6) of diabetes education per participant during an 8-week period. Average participant satisfaction scores were 62.2 (SD=2.6) out of a possible 65 points. A total of 102 links sent via text/email resulted in 504 web page views. There was no statistically significant difference in pre- and post-intervention diabetes-specific self-efficacy or diabetes related distress. Conclusions: Delivery of diabetes education via group videoconference with mobile web follow up is feasible and acceptable to young adults with T1D. This model of care delivery has the potential to improve attendance, social support, and patient-reported satisfaction. Further research is required to establish effect on long-term psychosocial and glycemic outcomes.

  • Background: Smartphones have integrated into our society and are expected to serve as tools to improve health outcomes. In the summer of 2016, Pokémon Go, a location-based augmented reality game for smartphones, was released and attracted attention from the perspective of health, especially with its potential to increase physical activity (PA). A few studies have compared objectively-measured step counts before and after the release of the game; however, they were conducted over a short study period and evaluated only young people. Objective: Forty-eight players and 192 non-players aged ≥38 years were matched for sex, age group, and PA level; they were respondents to a questionnaire randomly sent to citizens who were given free pedometers by Yokohama city. Their play status was identified through the questionnaire. To investigate the change in step counts before and after the release of Pokémon Go according to play status, two-way repeated measures ANOVA was performed. Step counts one month before the release of the game were compared with those eight months after the release. In addition, subgroup analysis according to sex, age group, PA level, and subjective health status was conducted. Methods: Among citizens who was given a free pedometer by Yokohama City, 48 players and 192 non-players, matched for sex, age group and PA level, were selected. They were aged 38 or older and their play statuses were identified through a questionnaire randomly sent to those citizens. To investigate the change in step counts before and after the release of Pokémon GO by play status, two way repeated measures ANOVA was used. Step counts before the release of 1 month were compared with those after the release of 8 months. In addition to the analysis of all samples, subgroup analysis by sex, age group, PA level and subjective health status were conducted. Results: The mean ages of players and non-players were 56.6 (±9.7) years and 57.3 (±9.6) years, respectively, and the mean of baseline step counts of players and non-players were 7776.5 (±2773.3) and 7947.1 (±2849.2), respectively. There was no significant difference in the age and baseline step counts according to a t-test. In the analysis of all samples, the interaction between play status and time effect was significant for three of eight months after release. In the subgroup analyses, the interaction was significant for five months in men, eight months in the 55–64-year-old group, three months in workers, two months in the active group in PA level, and four months in participants with subjectively good health. The interaction was significant for only one month at most in other subgroups. Conclusions: The present study confirmed a difference in step counts between players and non-players before and after the release of Pokémon GO. According to our analysis, step counts were higher until six months after the release. There was no decrease in the step counts of players even in cold months, despite the decrease in step counts of non-players. In subgroup analyses, players were more likely to be men, aged <55 years, workers, active, and subjectively in good health.

  • Mobile Health Interventions for Self-control of Unhealthy Alcohol Use: A Systematic Review

    Date Submitted: Apr 29, 2018
    Open Peer Review Period: Apr 30, 2018 - Jun 25, 2018

    Background: Unhealthy alcohol use (UAU) is one of the major causes of preventable morbidity, mortality and associated behavioral risks worldwide. Although mobile health (mHealth) interventions can provide consumers with an effective means for self-control of UAU in a timely, ubiquitous, and cost-effective manner, to date, there is a lack of evidence about the health outcomes of these interventions. The components of these interventions are also unclear. Objective: To systematically review and synthesize the research evidence about the health outcomes of mHealth interventions for self-control of UAU and to explore the common components of these interventions. Methods: We systematically searched four electronic interdisciplinary databases: Scopus, PubMed (PMC), CINAHL Plus with full text and MEDLINE with full text. Search terms and MeSH headings, “mHealth”, “text message”, “self-control”, “self-regulation”, “alcohol*” were used individually or in combination to identify peer-reviewed publications in English from 2007 to 2017. We screened titles and abstracts, and assessed full-text articles as per inclusion and exclusion criteria. Data were extracted from the included articles according to CONSORT-EHEALTH checklist (V.1.6.1) by two authors independently. Data quality was assessed by the Mixed Methods Appraisal Tool. Data synthesis and analyses were conducted following the procedures for qualitative content analysis. Statistical testing was also conducted to test differences among groups of studies. Results: Twenty publications were included in the review. In fourteen studies (70%), mHealth interventions brought significant positive outcomes in improving participants’ health as measured by behavioral indicators (n=13), physiological indicators (n=1) and cognitive indicators (n=1). No significant health outcome was reported in the other five studies (25%). Surprisingly, a significant negative outcome was reported for the male participants in the intervention arm in one study (5%), but no change was found for the female participants. Five common components reported in mHealth interventions for consumer self-control of UAU were context, theoretical base, delivery mode, content and implementation procedure. The health outcomes were similar regardless of types of UAU, deployment setting, with or without non-mobile co-intervention and with or without theory. Conclusions: The major change brought by mHealth interventions for consumer self-control of UAU appeared to be improving behavior. There is still a lack of sound evidence on the effects of these interventions in improving the physiological and cognitive outcomes. More robust trials are needed to validate the short- and long-term health benefits and the contributing factors of mHealth interventions.

  • Background: Patient portals are becoming more common, and with them the ability of patients to access their personal Electronic Health Records (EHRs). EHRs, in particular the free-text EHR notes, often contain medical jargon and terms that are difficult for laypersons to understand. There are many online resources for learning more about particular diseases or conditions, including systems that directly link to lay definitions or educational materials for medical concepts. Objective: Our goal is to determine whether use of one such tool, NoteAid, leads to higher EHR note comprehension ability. Methods: In this work we compare a passive, self-service educational resource (MedlinePlus) with an active resource (NoteAid) where definitions are provided to the user for medical concepts that the system identifies. We use Amazon Mechanical Turk (AMT) to recruit individuals to complete ComprehENotes, a new test of EHR note comprehension. Results: Mean scores for individuals with access to NoteAid are significantly higher than the mean baseline scores, both for raw scores (p=0.008) and estimated ability (p=0.017). Conclusions: In our experiments we show that the active intervention leads to significantly higher scores on the comprehension test as compared to a baseline group with no resources provided. In contrast, there is no significant difference between the group that was provided with the passive intervention and the baseline group. Finally, we analyze the demographics of the individuals who participated in our AMT task and show differences between groups that align with current understanding of health literacy between populations.

  • Context-aware systems for chronic disease patients: A scoping review

    Date Submitted: Apr 27, 2018
    Open Peer Review Period: Apr 27, 2018 - Jun 22, 2018

    Background: Context-aware systems, also known as context-sensitive systems, are computing applications designed to capture, interpret, and use contextual information and provide adaptive services according to the current context of use. Context-aware systems have the potential to support patients with chronic conditions, yet little is known about how such systems have been utilized to facilitate patient work. Objective: To characterize the different tasks and contexts in which context-aware systems for patient work were used, as well as to assess any existing evidence about the impact of such systems on health-related process or outcome measures. Methods: Six databases (Medline, Embase, CINAHL, ACM Digital, Web of Science and Scopus) were scanned using a predefined search strategy. Studies were included in the review if they: focused on patients with chronic conditions; involved the use of a context-aware system to support patients’ health-related activities; and reported the evaluation of the systems by the users. Studies were screened by independent reviewers, and a narrative synthesis of included studies was conducted. Results: The database search retrieved 1478 citations; 6 papers were included, all published from 2009 onwards. The majority of the papers were quasi-experimental and involved pilot and usability testing with a small number of users; there were no randomized controlled trials (RCTs) to evaluate the efficacy of a context-aware system. In the included studies, context was captured using sensors or self-reports, sometimes involving both. Most studies used a combination of sensor technology and mobile applications to deliver personalized feedback. Three studies examined the impact of interventions on health-related measures, showing positive results. Conclusions: The use of context-aware systems to support patient work is an emerging area of research. RCTs are needed to evaluate the effectiveness of effectiveness of context-aware systems in improving patient work, self-management practices, and health outcomes in chronic disease patients. Clinical Trial: N/A

  • Background: Digital mobile technology presents a promising medium for reaching young adults with smoking cessation interventions because they are the heaviest users of this technology. Objective: The primary aim of this trial was to determine the efficacy of an evidence-informed smartphone app for smoking cessation, Crush the Crave (CTC), on reducing smoking prevalence among young adult smokers in comparison to an evidence-informed self-help guide On the Road to Quitting (OnRQ). Methods: A parallel, double-blind randomized controlled trial (RCT) with two arms was conducted in Canada to evaluate CTC. In total, 1599 young adult smokers (19 to 29 years old) intending to quit smoking in the next 30 days were randomized to receive CTC or the control condition OnRQ for a period of 6 months. The primary outcome measure was self-reported continuous abstinence at the 6 month follow-up. Results: Overall follow-up rates were 55.5% and 58.3% at 3 and 6 months respectively. Forty-five percent of participants (n=725) completed baseline, 3 and 6 month follow-up. Intention-to-treat (last observation carried forward) continuous abstinence (n=1,599) at 6 months was not significantly different at 7.8% for CTC vs. 9.2% for OnRQ (OR = 0.83, 95% CI = 0.59-1.18). Similarly, 30-day point prevalence abstinence at 6 months was not significantly different at 14.4% and 16.9% for CTC and OnRQ respectively (OR = 0.82, 95% CI = 0.63-1.08); however, these rates of abstinence are favourable compared to unassisted 30-day quit rates of 11.5% among young adults. Secondary measures of quit attempts and the number of cigarettes smoked per day at 6 month follow-up did not reveal any significant differences between groups. For those who completed 6 month follow-up, 85.1% of young adult smokers downloaded CTC as compared to 81.8% of OnRQ, χ2 (1,N = 845) = 1.64, p =.23. Further, OnRQ participants reported significantly higher levels of overall satisfaction (M=3.3, SD=1.1 versus M=2.6, SD=1.3, t(644)=6.87, p < .0001) and perceived helpfulness (M=5.8, SD = 2.4 versus M=4.3, SD = 2.6, t(657)=8.0, p < .0001) as compared to CTC participants. Conclusions: CTC was feasible for delivering cessation support but was not superior to a self-help guide in helping motivated young adults to quit smoking. CTC will benefit from further formative research. Digital mobile technology smoking cessation interventions may serve as useful alternatives to traditional self-care guides due to the wide-spread availability of mobile technology. Clinical Trial: ClinicalTrials.gov NCT01983150; http://clinicaltrials.gov/ct2/show/NCT01983150

  • Background: Gestational trophoblastic disease (GTD) encompasses complete and partial moles, uncommon pregnancies with the potential to evolve to a malignant disease, the gestational trophoblastic neoplasia (GTN). While highly curable, the pre-malignant nature of molar pregnancies worries patients, leading them to seek information on the internet. A Facebook page headed by Brazilian specialized physicians in GTD was created in 2013 to provide online support for GTD patients. Objective: To describe the netnography of Brazilian patients with gestational trophoblastic disease (GTD) on Facebook® (FBGTD) and to evaluate whether experiences differed whether they received care in a Brazilian trophoblastic disease reference center (BRC) or elsewhere. Methods: This was a cross-sectional study using a survey of G Suit Google Platform and a netnographic analysis of interactions among patients with GTD members of FBGTD, invited to answer a survey from March 6th to October 5th, 2017. Results: The survey was answered by 356 Brazilian GTD patients: 176 reference center patients (RCP) treated at a BRC and 180 non-reference center patients (NRCP) treated elsewhere. Comparing the groups, RCP felt safer and more confident at the time of diagnosis of GTD (P = 0.001), and they utilized FBGTD through referral by health assistants (P < 0.001), whereas NRCP discovered FBGTD on search sites and pages (P < 0.001). NRCP had higher educational level (P = 0.009), were on FBGTD ≥ 6 months (P = 0.026), reported more frequently (P = 0.007) that what most negatively affected them at the time of diagnosis was the fact that doctors did not explain about GTD and had more doubts about GTD treatment (P = 0.013), and were unlikely to use hormonal contraception (P < 0.001). Most patients accessed the internet preferentially from home and using cell phones (89%). Almost all patients (98%) declared that they felt safe reading the guidelines posted by FBGTD physicians. Conclusions: This netnographic analysis of GTD patients on FBGTD shows that an online doctor-patient relationship can supplement the care for women with GTD. This resource is particularly valuable for women being cared for outside of established reference centers.

  • Background: Background: The inter-pregnancy period is an important window of opportunity to prevent gestational weight retention and associated risks in the next pregnancy. Despite an overwhelming number of available health apps, targeted and validated applications supporting a healthy lifestyle in women during both pregnancy and postpartum are lacking. Objective: Objective: The aim of this paper is to describe the process of development and pilot testing of the INTER-ACT app. The INTER-ACT app is part of an inter-pregnancy coaching module targeting women with excessive gestational weight gain. These women are followed up to the end of the next pregnancy in order to reduce the risk for maternal and childbirth complications. Methods: Methods: a mixed methods design was used to identify needs of healthcare providers and end users (pregnant and postnatal women) before designing the app. This was based on 15 semi-structured interviews with midwives (n=5), GP’s (n=4), gynecologists (n=3) and dieticians (n=3), two focus groups with midwives, a survey in 43 healthcare providers and 50 pregnant women as well as on iterative discussions within our research group and two professional organizations. The app user interface was designed by user experience researchers and evaluated in a pilot study including 9 women. Results: Results: From the healthcare providers’ perspective, a mobile app can enhance a healthy lifestyle in pregnant women provided that the app can be combined with face-to-face coaching without increasing administrative workload. Pregnant women preferred a graphic display of weight evolution in the app and weekly notifications and support messages, based on their own preset goals. In the user evaluation study, mothers reported that the INTER-ACT app increased their awareness about their own behavior. In comparison with commercially available health apps mothers explicitly valued the combination of the INTER-ACT app with face-to-face contact. This finding aligns with healthcare providers’ perspectives. Conclusions: Conclusions: We report on the development and evaluation process of an application to promote a healthy lifestyle in a very specific female population. These pregnant and postpartum women specifically valued the INTER-ACT app because it was offered in combination with face-to-face contact with a caregiver.

  • The perceptions of Midwives, Obstetricians, and recently delivered Mothers to remote monitoring for prenatal care

    Date Submitted: Apr 26, 2018
    Open Peer Review Period: Apr 27, 2018 - Jun 22, 2018

    Background: There have been few studies on remote monitoring (RM) in midwifery. These studies were mostly performed several decades ago, and no recent studies have investigated the perceptions to or experiences of new technologies. The Pregnancy Remote Monitoring (PREMOM) study, which started in January 2015 in Ziekenhuis Oost-Limburg (Genk, Belgium), enrolled pregnant women at increased risk of developing gestational hypertensive disorders (GHD). Women enrolled in PREMOM underwent conventional prenatal follow-up, which was complemented with RM. Objective: We sought to investigate the perceptions and experiences of mothers, midwives, and obstetricians to the RM approach used in the PREMOM study. Methods: We developed specific questionnaires for the mothers, midwives, and obstetricians. The questionnaires comprised five domains: ‘prior knowledge and experience of RM’, ‘reactions to abnormal values’, ‘privacy’, ‘quality and patient safety’, and ‘financial aspects’. The caregivers were also questioned about which issues they consider important when implementing RM. A five-point Likert scale was used to provide objective scores. Results: Ninety-one participants completed the questionnaires, including 47/92 (51.08%) mothers, 35/52 (67.30%) midwives, and 9/14 (64.29%) obstetricians. The mothers, midwives, and obstetricians reported positive experiences and perceptions to RM. Overall, 29/35 (82.85%) midwives and 7/9 (77.78%) obstetricians had no or little prior experience with this technology. After working for 1 year with RM, 28/35 (80.00%) midwives and 6/9 (66.67%) obstetricians felt that this technology is an important component in the prenatal monitoring of high-risk pregnancies and that it had a positive contribution to the care of pregnant women. They support a further roll-out of RM in Belgium, but caregivers need additional training on RM devices and the pathological aspects of GHD. Nearly three-quarters of the mothers who participated in the PREMOM study (34/47, 72.34%) did not report any problems with taking the measurements at the required times. Almost half of the mothers (19/47, 40.43%) wanted to be contacted within 3–12 hours after abnormal values, preferably by telephone. Nearly all of the mothers (41/47, 87.24%) did not have any problems with regularly sharing their health data with their gynaecologist. Finally, most of the mothers (39/47, 82.97%) reported that RM gave them a feeling of security throughout their pregnancy. Conclusions: Although the majority of midwives and obstetricians had no or very little experience with RM before enrolling in the PREMOM study after one year, they reported that RM is an important component in the follow-up of high-risk pregnancies and would recommend it to their colleagues and pregnant patients.

  • Background: Non-professional caregivers who provide support to an individual with a psychiatric or neurological disorder will often themselves experience symptoms of stress, anxiety, and low mood, and perceive that they receive little support. Internet-based interventions have previously been found to be effective in the treatment and prevention of a range of mental health difficulties. Objective: The current review seeks to investigate the status of existing internet-based interventions for informal carers of people with psychiatric and neurological disorders or traumatic brain injury, establishing the degree to which they improve carer mental health. Methods: A systematic literature search was conducted in October 2017 using the Embase (1974 – present), Ovid MEDLINE (1946 – present), PsychARTICLES, PsychINFO (1806 – present) and Global Health (1973 – present) databases, via the Ovid Technologies database. Search terms included carer, caregiver, online, technology, web-based, interactive, intervention, and evaluation. Studies selected for inclusion in this review met the following predetermined criteria: (i) delivering an intervention aimed primarily at informal (non-professional) caregivers; of (ii) individuals with psychiatric disorders, stroke, dementia, or brain injury; (iii) the intervention delivered to the caregivers was primarily internet-based; (iv) the study reported a pre- and post- quantitative measure of caregiver depression, anxiety, stress, burden, or quality of life; (v) appearing in a peer-reviewed journal; and (vi) accessible in English. Results: A total of 39 studies were identified for inclusion through the detailed search strategy. The search was conducted and data information extracted independently by two researchers. Studies were found to vary greatly with regards to both methodological quality, and reported efficacy/effectiveness of the intervention on measures relating to caregiver mental health, however, the majority of studies did report that one or more measures relating to carer mental health was found to improve following receipt of a relevant intervention. Conclusions: Studies investigating internet-based interventions for carers of individuals with psychiatric and neurological difficulties and brain injury show some evidence in support of the effectiveness of internet-based interventions, and additionally, that such interventions are likely to be found acceptable by the caring population. The evidence currently available within this field is of varying quality, and has some methodological shortcomings. Further research in this area is required to establish how specific components, such as the structure and interactivity of the intervention, may contribute to the overall efficacy of the intervention with regards to carer mental health, and how this frequently underserved population can best be supported.

  • Background: Patient-reported outcomes (PROs) are increasingly utilized in routine orthopedic clinical care. Computer adaptive tests (CATs) from the Patient-Reported Outcomes Measurement Information System (PROMIS) offer brief and precise assessment that is well suited for collection within busy clinical environments. However, software applications that support the administration and scoring of CATs, immediate access to PRO scores, and minimize clinician burden are not widely available. Objective: Our objective was to design, implement, and test the feasibility and usability of a web-based system for collecting CATs in orthopedic clinics. Methods: AO Patient Outcomes Center (AOPOC) was subjected to two rounds of testing. Alpha testing was conducted in 3 orthopedic clinics to evaluate ease of use and feasibility of integration in clinics. Patients completed an assessment of PROMIS CATs and a usability survey. Clinicians participated in a brief semi-structured interview. Beta phase testing evaluated system performance through load testing and usability of the updated version of AOPOC. In both rounds of testing, user satisfaction, bugs, change requests, and performance of PROMIS CATs were captured. Results: Patient feedback supported the ease of use in completing an assessment in AOPOC. Across both phases of testing, clinicians rated AOPOC as easy to use, but noted difficulties in integrating a web-based software application within their clinics. PROMIS CATs performed well; the default assessment of 2 CATs was completed quickly (mean=9.5 items) with a satisfactory range of measurement. Conclusions: AOPOC was demonstrated to be an easy to learn and easy to use software application for patients and clinicians that can be integrated into orthopedic clinical care. The workflow disruption in integrating any type of PRO collection must be addressed if patients’ voices are to be better integrated in clinical care.

  • A systematic review of online systems for patients to report and manage the side effects of cancer treatment

    Date Submitted: Apr 25, 2018
    Open Peer Review Period: Apr 26, 2018 - Jun 21, 2018

    Background: There has been a dramatic increase in the development of electronic systems to support cancer patients to report and manage side effects of treatment from home. Systems vary in features that they offer to patients which may impact on how patients engage with them, and subsequently on the benefit that may be derived in terms of patient centred outcomes. Objective: This review aimed to 1) Describe the features and functions of existing electronic symptom reporting systems (e.g. symptom monitoring, tailored self-management advice etc.) 2) Explore which features may be associated with patient engagement and patient-centred outcomes. Methods: The review was registered with PROSPERO and followed guidelines from the Centre for Reviews and Dissemination. Primary searches were undertaken of MEDLINE, EMBASE, PsychInfo, Web of Science, Cochrane Central Register of Controlled Trials and the Health Technology Assessment databases. Secondary searches were undertaken by screening reference lists and citations. Two researchers applied broad inclusion criteria to identify and select relevant records. Data was extracted and summarised using Microsoft Excel. In order to meet the aims, study selection, data extraction and data synthesis evolved into two stages. Stage 1: Identifying and characterising available systems and Stage 2: Summarising data on patient engagement and patient centred outcomes. Results: 77 publications were identified relating to 41 distinct systems. Stage 1: All publications were included (n=77). Features identified which supported clinicians and care were: facility for health professionals to remotely access and monitor patient reported data (58%) and function to send alerts to health professionals for severe symptoms (41%). Features which supported patients were: facility for patients to monitor/review their symptom reports over time (e.g. graphs) (46%), general patient information about cancer treatment and side-effects (41%), tailored automated patient advice on symptom management (29%), feature for patients to communicate with the healthcare team (15%) and a forum for patients to communicate with one another (10%). Stage 2: Only publications which included some data on patient engagement and/or patient-centred outcomes were included (n=29). A lack of consistency between studies in how engagement was defined, measured or reported and a wide range of methods chosen to evaluate systems meant we couldn’t compare across studies or make conclusions on relationships with system features. Conclusions: Electronic systems have the potential to support patients to manage side effects of cancer treatment with some evidence to suggest a positive effect on patient centred outcomes. However, comparison across studies is difficult due to the wide range of assessment tools used. There is a need to develop guidelines for assessing and reporting engagement with systems, and a set of core outcomes for evaluation. We hope that this review will contribute to the field by the introduction of a taxonomy for characterising system features.

  • Television advertising and Marketplace consumer engagement: the case of Kentucky

    Date Submitted: Apr 24, 2018
    Open Peer Review Period: Apr 25, 2018 - Jun 20, 2018

    Background: Reductions in health insurance enrollment outreach could have negative effects on the individual health insurance market. Specifically, consumers may not be informed about the availability of coverage and if some healthier consumers fail to enroll, there could be a worse risk pool for insurers. Kentucky created its own Marketplace, known as kynect, and adopted Medicaid expansion under the ACA, which yielded the largest decline in adult uninsured rate in the country from 2013 to 2016. The state sponsored an award-winning media campaign yet declined to renew the advertising contract for kynect and canceled all pending television ads with over a month remaining in the 2016 open enrollment period after the election of a new governor in 2015. Objective: The objective of this study is to examine the stark variation in advertising across multiple open enrollment periods in Kentucky and to use this variation to estimate the dose-response effect of state-sponsored advertising on consumer engagement with the Marketplace. In addition, we assess to what extent private insurers can potentially help fill the void when governments reduce or eliminate television advertising. Methods: We obtained advertising (Kantar Media/CMAG) and Marketplace data (Kentucky Health Benefit Exchange) for the period of October 1, 2013 through January 31, 2016 for Kentucky. Advertising data at the spot level were collapsed to state-week counts by sponsor type. Similarly, a state-week series of Marketplace engagement and enrollment measures were derived from state reports to CMS. We use linear regression models to estimate associations between health insurance advertising volume and measures of information seeking (calls to call center; page views, visits, and unique visitors to website) and enrollment (online and total applications, Marketplace enrollment). Results: We find significant dose-response effects of weekly state-sponsored advertising volume during open enrollment on information-seeking behavior (page views: 7,972.9, visits: 390.2, and unique visitors: 387.5) and enrollment activity (applications, online: 61.4 and total: 55.9). Conclusions: State-sponsored television advertising was associated with nearly 40% of unique visitors and online applications. Insurance company advertising was not a significant driver of engagement, an important consideration if cuts to government-sponsored advertising persist.

  • A Novel Insight into the Challenges of Diagnosing Cervical Spondylotic Myelopathy using Online Symptom Checkers

    Date Submitted: Apr 25, 2018
    Open Peer Review Period: Apr 25, 2018 - Jun 20, 2018

    Background: Cervical Spondylotic Myelopathy (CSM) is a common debilitating condition resulting from degeneration of the cervical spine. Whilst decompressive surgery can halt disease progression, existing spinal cord damage is often permanent, leaving patients with lifelong disability. Treatment within 6 months improves the likelihood of recovery, yet most patients wait over 2 years for a diagnosis. The majority of delays occur initially, before and within primary care, mainly due to a lack of recognition. Symptom checkers are widely used by patients before medical consultation and can be useful for preliminary triage and diagnosis. Lack of recognition of CSM by symptom checkers may contribute to the delay in diagnosis. Objective: The present study sought to investigate whether online symptom checkers are able to recognise relevant symptoms of CSM, to characterise the CSM differential returned, and to evaluate the diagnostic performance of its recognised symptoms. Methods: Textbook CSM symptoms were pooled from leading review articles. These symptoms were entered into the algorithms used by the top 20 symptom checker websites (N=4) (Google Search). The most widely cited symptom checker, WebMD, was used to characterise the differential diagnosis for CSM symptoms. Results: 31 textbook CSM symptoms were identified, of which 14 (45%) listed CSM as a differential and 3 (10%) placed CSM in the top third of the differential. The mean differential rank for motor symptoms was significantly better than for arthritic symptoms (P = .0093) and the average differential rank for all symptoms (P = .048). The symptom checker WebMD performed best at recognising CSM, placing it nearer the top of the differential list (average rank of 5.6) than both Healthline (rank of 12.9, P = .015) and Healthtools.AARP (rank of 15.5, P = .0014). On WebMD, only one combination of symptoms resulted in CSM as the primary differential: neck, shoulder and arm pain with hand weakness. 151 differential diagnoses for CSM symptoms were recorded on WebMD. Multiple sclerosis and peripheral neuropathy were the most common differentials, shortlisted for 52% and 32% of CSM symptoms respectively. Conclusions: CSM symptoms perform poorly in online symptom checkers and lead to a large differential of many other common conditions. Whilst a diagnosis becomes more likely as the number of symptoms increases, this will not support much needed earlier diagnosis. Symptom checkers remain an attractive concept with potential. Further research is required to support their optimisation.

  • Quantifying Impact of Validation and Testing on FHIR Standards Compliance

    Date Submitted: Apr 24, 2018
    Open Peer Review Period: Apr 25, 2018 - Jun 20, 2018

    Background: There is wide recognition that the lack of health data interoperability has significant impacts. Traditionally health data standards are complex, and test-driven methods played important roles in achieving interoperability. HL7 FHIR may be a technical solution that aligns with policy, but systems need to be validated and tested. Objective: Our objective is to explore the question of whether or not the regular use of validation and testing tools improves server compliance with the HL7 FHIR specification. Methods: We used two independent validation and testing tools, Crucible and Touchstone, and analyzed the usage and result data to determine their impact on server compliance with the HL7 FHIR specification. Results: The use of validation and testing tools such as Crucible and Touchstone are strongly correlated with increased compliance, and “practice makes perfect.” Additional data analysis reveals trends over time with respect to vendors, use-cases, and FHIR versions. Conclusions: Validation and testing tools can aid in the transition to an interoperable health care infrastructure. Developers that use testing and validation tools tend to produce more compliant FHIR implementations. When it comes to health data interoperability, “practice makes perfect.”

  • Background: The quality and quantity of families’ support systems during pregnancy can affect maternal and fetal outcomes. The support systems of expecting families can include many elements, such as family members, friends, and work or community groups. Emerging health information technologies such as social media, Internet websites, and mobile apps provide new resources for pregnant families to augment their support systems and to fill information gaps. Objective: This study sought to determine the number and nature of the components of the support systems of pregnant women and their caregivers, and the role of health information technologies in these support systems. We examined the differences between pregnant women’s support systems and those of their caregivers and the associations between support system composition and stress levels. Methods: We enrolled pregnant women and caregivers from an advanced maternal fetal and group prenatal care clinics. Participants completed surveys assessing sociodemographic characteristics, health literacy, numeracy, and stress levels, and they were asked to draw a picture of their support system. Support system elements were extracted from drawings, categorized by type (i.e., individual persons, groups, technologies, and other), and summarized for pregnant women and caregivers. Participant characteristics and support system elements were compared using Pearson Chi-square test for categorical variables and Wilcoxon ranked sum test for continuous variables. Associations between support system characteristics and stress levels were measured with Spearman correlation coefficient. Results: The study enrolled 100 participants: 71 pregnant women and 29 caregivers. The support systems of pregnant women were significantly larger than those of caregivers, with an average of 7.4 components for pregnant women and 5.4 components for caregivers (P=0.003). For all participants, the most commonly reported support system elements were individual persons (60.3% of support systems), followed by people groups (19.3%), technologies (16.2%), and other resources (4.2%). Pregnant women’s and caregivers’ technology preferences within their support systems differed: pregnant women more often identified informational websites, apps, and social media as parts of their support systems, whereas caregivers more frequently reported general Internet search engines. The size and components of these support systems were not associated with levels of stress. Conclusions: This study is one of the first demonstrating that technologies comprise a substantial portion of the support systems of pregnant women and their caregivers. Pregnant women more frequently reported specific medical information websites as part of their support system, whereas caregivers more often reported general Internet search engines. Although social support is important for maternal and fetal health outcomes, no associations among stress, support system size, and support system components were found in this study. As health information technologies continue to evolve and their adoption increases, their role in patient and caregiver support systems and their effects should be further explored.

  • Background: Randomized controlled trials conducted in Mediterranean countries have shown that the Mediterranean diet lowers adverse cardiovascular events. In the American population, diet remains the biggest uncontrolled risk factor for cardiovascular disease, and patients with cardiovascular disease require more extensive dietary counseling to reduce their risk of myocardial infarction, stroke and hypertension. Objective: Our study tested the hypothesis that asynchronous dietary counseling supplied through a custom smartphone application results in better adherence to a Mediterranean diet in a non-Mediterranean population compared with traditional standard-of-care counseling. Methods: 100 patients presenting to the cardiology clinic of an academic medical center in Washington D.C. were randomized to either the standard-of-care (SOC) or smartphone app-based (EXP) Mediterranean diet intervention after informed consent and 1 hour of individual face-to-face dietary counseling with a registered dietician. Participants in EXP received a custom smartphone application that reinforced the Mediterranean diet while participants in SOC received two additional sessions of in-person dietary counseling with the registered dietician – 30 minutes at 1 month and 30 minutes at 3 months. Pre-existing knowledge of a Mediterranean diet was measured by the validated Mediterranean Diet Score instrument (MDS). Baseline height, weight, blood pressure and laboratory biomarkers were collected. At 1, 3 and 6 months, participants presented for a follow up appointment to assess compliance to the Mediterranean diet using the MDS, as well as a patient satisfaction survey, blood pressure and weight. Repeat laboratory biomarkers were performed at 3 and 6 months. Results: Enrolled participants had a mean age with standard error of 56.6 ± 1.7 for SOC and 57.2 ± 1.8 for EXP; 65.3% of SOC and 56.9% of EXP were male, and 20.4% of SOC and 35.3% of EXP had coronary artery disease. There were no significant differences between EXP and SOC with respect to blood pressure, lipid parameters, HgbA1C, or CRP. Participants in EXP achieved a significantly greater weight loss on average of 3.3 lbs vs. 3.1 lbs for participants in SOC, p=0.04. Adherence to the Mediterranean diet increased significantly over time for both groups (p<0.001), but there was no significant difference between groups (p=0.69). Similarly, there was no significant difference in diet satisfaction between EXP and SOC, although diet satisfaction increased significantly over time for both groups. The proportion of participants with high Mediterranean diet compliance (defined as MDS ≥ 9) increased significantly over time (p<0.001) – from 18.4 to 57.1% for SOC and 27.5 to 64.7% for EXP – however, there was no significant difference between the groups. Conclusions: Both traditional standard-of-care counseling and smartphone-based counseling were effective in getting participants to adhere to a Mediterranean diet in this US-based study of cardiology patients, and these dietary changes persisted even after counseling had ended. However, neither method was more effective than the other. This study demonstrates that the American cardiology patient can change to and maintain a Mediterranean diet with either traditional or smartphone app-based nutrition counseling.

  • Technology-based alcohol interventions in primary care: A systematic review

    Date Submitted: Apr 23, 2018
    Open Peer Review Period: Apr 24, 2018 - Jun 19, 2018

    Background: Primary care settings are uniquely positioned to reach individuals at-risk of alcohol use disorder through technology-delivered behavioral health interventions. Despite emerging effectiveness data, few efforts have been made to summarize the collective findings from these delivery approaches. Objective: The purpose of this study was to review recent literature on the use of technology to deliver, enhance, or support the implementation of alcohol-related interventions in primary care. We focused on addressing questions related to (1) categorization and target, (2) descriptive characteristics and context of delivery, (3) reported efficacy, and (4) factors influencing efficacy. Methods: We conducted a comprehensive search and systematic review of completed studies at the intersection of primary care, technology, and alcohol-related problems published from January 2000 to May 2015 within EBSCO databases, ProQuest Dissertations, and Cochrane Reviews. Of 1498 initial records, 35 were included and coded independently by two investigators. Results: Studies on technology-based alcohol interventions in primary care have increased four-fold between the years of 2000-2005 and 2010-2015. Of the 35 studies, 23 (66%) were randomized controlled trials. Further, as a proxy for study quality, studies were rated on risk of bias and found to be low risk/high quality (n=15), moderate risk (n=13), and high risk/low quality (n=7). Approximately two-thirds (68%) of studies reported reduced drinking or harm in all primary and secondary efficacy outcomes. Furthermore, 38% of studies reported that at least half of the primary and secondary efficacy outcomes of the technology-based interventions were superior to treatment as usual. High efficacy was associated with provider involvement and the reported use of an implementation strategy to deliver the technology-based intervention. Conclusions: Our systematic review highlights a pattern of rapid growth in the use and evaluation of technology-based alcohol interventions in primary care. While these interventions appear largely beneficial in primary care, outcomes may be enhanced by provider involvement and implementation strategy use. This review enables better understanding of the typologies and efficacy of these interventions and informs recommendations for those developing and implementing technology-based alcohol interventions in primary care settings.

  • Daily Smartphone Overdependence Screening Model using Support Vector Machine

    Date Submitted: Apr 23, 2018
    Open Peer Review Period: Apr 23, 2018 - Jun 18, 2018

    Background: Smartphone overdependence has caused many social problems. To overcome these problems, it is necessary to screen and identify smartphone overdependence before it becomes a serious issue. Objective: We aimed to developed a daily smartphone overdependence screening model using a Support Vector Machine (SVM). Methods: We used smartphone application usage time and frequency data from 224 participants whose ages ranged from their 20s to their 40s. We classified the participants into two groups the smartphone usage control group (SUC) and the smartphone usage disorder addiction group (SUD) using the Korean Smartphone Addiction Proneness Scale (K-SAPS) for Adults. We built a 3-dimensional tensor as the input of machine learning training. This study used the SVM to develop a daily smartphone overdependence screening model. We compared the model performance between the SVM, the Artificial Neural Network(ANN) and the Logistic Regression. Results: We identified the frequency of smartphone application usage, age, and marital status as the dominant features of screening smartphone overdependence. Using these features as the inputs of the SVM machine learning model showed a 90% of accuracy for the smartphone overdependence screening. Conclusions: We developed a SVM model, which is a tool for self-control of smartphone daily usage. As a pre-testing tool before visiting a mental health clinic. The SVM model is a powerful analysis method for smartphone overdependence screening. Notably, psychiatry studies have used the SVM when identifying a psychiatric disease. We suggest using the SVM model for smartphone overdependence screening as a smartphone application or intervention system for smartphone dependency management.

  • Background: Online health information seeking by patients was shown to affect medical consultations and patient-doctor relationships. An up-to-date picture of patients’ online health information seeking behaviors can inform and prepare frontline healthcare professionals to collaborate, facilitate or empower patients to access and manage health information online. Objective: This study explores the prevalence, patterns, and predictors of online health information seeking behaviors among primary care patients in Hong Kong, and the associations between online health information seeking and eHealth literacy. Methods: Patients attending a university primary care clinic in Hong Kong were asked to complete a questionnaire survey on their demographic backgrounds; health status; frequency and pattern of online health information seeking; contents, sources, and reasons of online health information seeking, and their eHealth literacy. eHealth literacy was measured by the validated eHealth Literacy Scale (eHEALS). Independent-samples t-tests were performed to examine the association between eHealth literacy and respondent characteristics, and between eHealth literacy and source of online health information. A hierarchical multiple regression analysis explored various demographics and behavioral predictors to online health information seeking. Results: 97.3% (1162/1194) respondents used the Internet, of which 87.4% (1016/1162) had used the Internet to find health information. Most respondents (66.0%, 665/1008) searched once monthly or more. Few (26.7%, 271/1016) asked their doctor about health information found online, but most doctors (56.1%, 152/271) showed little or no interest at all. The most sought topic was symptom (81.6%, 829/1016), the top reason was noticing new symptoms or change in health (70.1%, 712/1016), the most popular source is online encyclopedia (70.0%, 711/1016), and the top reason for choosing a source was convenience (55.4%, 563/1016). Hierarchical multiple regression identified eHEALS score, fair or poor self-rated health, senior job rank, having chronic medical condition, and using the Internet daily or more as significant predictors to online health information seeking. Respondents choosing authoritative websites for health information were associated with higher eHEALS score. Conclusions: OHIS is prevalent among primary care patients in Hong Kong, but only a minority shared the information with doctors. Higher eHealth literacy was associated with using authoritative websites. Doctors should recognize, facilitate and empower patients for OHIS.

  • A multi-centre pragmatic randomized controlled trial of a web-based mental health platform in individuals seeking specialized mental health care services

    Date Submitted: Apr 20, 2018
    Open Peer Review Period: Apr 23, 2018 - Jun 18, 2018

    Background: Web-based self-directed mental health applications are rapidly emerging as a solution to health service gaps and unmet needs for information and support. Objective: The aim of this study was to determine if a multi-component web-based moderated mental health application could benefit individuals with mental health symptoms severe enough to warrant specialized mental health care. Methods: A multi-centre, pragmatic randomized controlled trial was conducted across several outpatient mental health programs affiliated with 3 hospital programs in Ontario, Canada. Individuals referred to or receiving treatment, 16 years of age or older, with access to the internet and an email address, and having the ability to navigate a web-based mental health application were eligible. 812 participants were randomized 2:1 to receive immediate (ITG) or delayed (DTG) access for 3 months to the Big White WallTM, a web-based multi-component mental health intervention based in the United Kingdom and New Zealand. The primary outcome was total score on the Recovery Assessment Scale, revised (RAS-r) measuring mental health recovery. Secondary outcomes were total scores on the Patient Health Questionnaire-9 item (PHQ-9), the Generalized Anxiety Disorder Questionnaire-7 item (GAD-7), the EQ-5D-5L, and the Community Integration Questionnaire (CIQ). An exploratory analysis examined the association between actual BWW use (categorized into quartiles) and outcomes among study completers. Results: Intervention participants achieved small, statistically significant increases in adjusted RAS-r score (4.97 points, 95% CI 2.90 to 7.05), and decreases in PHQ-9 score (-1.83 points, 95% CI -2.85 to -0.82) and GAD-7 score (-1.55 points, 95% CI -2.42 to -0.70). Follow-up was achieved for 446 (55%) at 3 months; 48% of ITG participants, and 69% of DTG participants. Only 58% of ITG participants logged on more than once. Some higher BWW user groups had significantly greater improvements in PHQ-9 and GAD-7 relative to the lowest use group. Conclusions: The web-based application may be beneficial, however, many participants did not engage in an ongoing way. This has implications for patient selection and engagement as well as delivery and funding structures for similar web-based interventions. Clinical Trial: Clinicaltrials.gov NCT02896894. Registered on 31 August 2016 (retrospectively registered). https://clinicaltrials.gov/ct2/show/NCT02896894

  • Background: Research has shown that some individuals can develop problematic patterns of online gaming, leading to significant psychological and interpersonal problems. Psychiatric distress and impulsivity has been suggested to contribute to problematic online gaming (POG). Objective: The underlying mediating and/or moderating mechanism of impulsivity as well as gender differences in possible associations between psychiatric distress and POG are largely unknown, which guides this current study Methods: To address this gap, the current study examined relationships between self-reported impulsivity, psychiatric distress, and POG in a sample of matched Hungarian female and male online gamers (n=596, 50% males). Results: Results showed that psychiatric distress directly predicted POG, and impulsivity partially mediated the relationship between psychiatric distress and POG. However, this mediation effect was found only for the impatience subscale. Impulsivity did not moderate the relationship between psychiatric distress and POG. A moderating effect of gender was not found in the direct relationship between psychiatric distress and POG. However, a moderated mediation analysis revealed that impatience mediated the association between psychiatric distress and POG in males, whereas the indirect effect of impatience was not significant in females. Conclusions: Future implications are discussed in light of these results. Clinical Trial: N/a

  • Characterizing tweet volume and content about common health conditions across Pennsylvania

    Date Submitted: Apr 20, 2018
    Open Peer Review Period: Apr 23, 2018 - Jun 18, 2018

    Background: Tweets can provide broad, real time perspectives about health and medical diagnoses that can inform disease surveillance in geographic regions. Less is known however about how much individuals post about common health conditions or what they post about. Objective: We sought to collect and analyze tweets from one state about high prevalence health conditions and characterize tweet volume and content. Methods: We collected 408,296,620 tweets originating in Pennsylvania from 2012-2015 and compared the prevalence of 14 common diseases to the frequency of disease mentions on Twitter. We identified and corrected bias induced due to variance in disease term specificity and used the machine learning approach of differential language analysis to determine the content (words and themes) most highly correlated with each disease. Results: Common disease terms were included in 226,802 tweets. Posts about breast cancer (22.5% messages, 2.4% prevalence) and diabetes (23.1% messages, 17.2% prevalence) were overrepresented on Twitter relative to disease prevalence, while hypertension (9.9% messages, 36.3% prevalence), COPD (0.9% messages, 8.5% prevalence), and heart disease (7.8% messages, 19.4% prevalence) were underrepresented. The content of messages also varied by disease. Personal experience messages accounted for 12% of prostate cancer tweets and 24% of asthma tweets. Awareness themed tweets were more often about breast cancer (23%) than asthma (6%). Tweets about risk factors were more often about heart disease (10%) than lymphoma (2%). Conclusions: Twitter provides a window into the online visibility of diseases and how the volume of online content about diseases varies by condition. Further, the potential value in tweets is in the rich content they provide about individuals’ perspective about diseases (e.g. personal experiences, awareness, risk factors) that are not otherwise easily captured through traditional surveys or administrative data.

  • Effects of a Mindfulness-Meditation App on Subjective Well-Being: An Active Randomized Controlled Trial and Experience Sampling Study

    Date Submitted: Apr 20, 2018
    Open Peer Review Period: Apr 23, 2018 - Jun 18, 2018

    Background: Mindfulness training (MT) includes a variety of contemplative practices aimed at promoting intentional awareness of experience, coupled with attitudes of non-judgment and curiosity. Following the success of 8-week, manualized, group interventions, MT has been implemented in a variety of modalities, including smartphone applications that seek to replicate the success of group interventions. However, while smartphone apps are scalable and accessible to a wider swath of population, their benefits remain largely untested. Objective: The present study investigated a newly developed MT application called Wildflowers, which was co-developed with the lab for use in mindfulness research. It was hypothesized that MT through this application would improve subjective well-being, attentional control, and interoceptive integration, albeit with weaker effects than those published in the group intervention literature. Methods: Undergraduate students completed 3-weeks of MT with Wildflowers (n = 45), or 3-weeks of mathematical problem-solving training with a game called 2048 (n = 41). State training effects were assessed through pre- and post-session ratings of current mood, stress level, and heart rate. Trait training effects were assessed through pre- and post-intervention questionnaires canvassing subjective well-being, and behavioural task measures of attentional control and interoceptive integration. State and trait training data were analyzed in a multilevel model, using emergent latent factors (acceptance, awareness, openness) to summarize the trait questionnaire battery. Results: Analyses revealed both state and trait effects specific to MT; participants engaging in MT demonstrated improved mood (r = 0.14) and a reduction of stress (r = -0.13) immediately after each training session compared to before the training session, and decreased post-session stress over 3-weeks (r = -0.08). Additionally, MT relative to math training resulted in greater improvements in attentional control (r = -0.24). Interestingly, both groups demonstrated increased subjective ratings of awareness (r = 0.28) and acceptance (r = 0.23) from pre- to post-intervention, with greater changes in acceptance for the MT group trending (r = 0.21). Conclusions: MT using a smartphone app may provide immediate effects on mood and stress while also providing long-term benefits for attentional control. Although further investigation is warranted, there is evidence that with continued usage, MT via a smartphone app may provide long-term benefits in changing how one relates to their inner and outer experiences. Clinical Trial: This study was not registered at ClinicalTrials.gov, but the study protocol matches that of the REB approval acquired prior to the study trial.

  • Background: There is growing evidence that physical activity is related to a better prognosis after a breast cancer diagnosis, whereas sedentary behavior is associated with worse outcomes. It is therefore important to stimulate physical activity and reduce sedentary time among breast cancer patients. Activity trackers offer a new opportunity for interventions directed at stimulating physical activity behavior change. Objective: This study explored the experience of breast cancer patients who used an activity tracker in addition to a supervised exercise intervention in the randomized UMBRELLA Fit trial. Methods: Ten patients with breast cancer who completed cancer treatment participated in semi-structured in-depth interviews about their experience with and suggestions for improvements for the Jawbone UP2 activity tracker. Results: The activity tracker motivated women to be physically active and created more awareness of their (sedentary) lifestyles. The women indicated that the automatically generated advice (received via the UP app) lacked individualization and was not applicable to their personal situations (i.e., having been treated for cancer). Furthermore, women felt that the daily step goal was one-dimensional and they preferred to incorporate other physical activity goals. The activity tracker’s inability to measure strength exercises was a noted shortcoming. Finally, women valued personal feedback about the activity tracker from the physiotherapist. Conclusions: Wearing an activity tracker raised lifestyle awareness in breast cancer patients. The women also reported additional needs not addressed by the system. Potential improvements include a more realistic total daily physical activity representation, personalized advice, and personalized goals.

  • “It's not just technology, it's people”: A Conceptual Model of Shared Health Informatics for chronic illness

    Date Submitted: Apr 19, 2018
    Open Peer Review Period: Apr 21, 2018 - Jun 16, 2018

    Background: Chronic illness is a substantial public health burden and the day-to-day work of a person with chronic illness is central to illness management. However, although management programs are available to train people to manage their chronic illnesses, those programs rarely provide people with sufficient guidance on implementing fine-grained management strategies. This lack of guidance transfers much of the burden of day-to-day management practices to people with chronic illness. Objective: To decrease the management burden many people have created technological solutions, but we claim that to be maximally effective and tailored to people’s needs, those solutions need to be informed by a model that integrate patient work and self-management frameworks with personal health informatics models. We see an opportunity to improve personal health informatics models to (1) incorporate the ongoing nature of management work, and (2) append a social dimension to more accurately depict the context of illness management. Methods: We used qualitative methods to analyze the chronic illness management practices of 63 people through the lenses of self-management and informatics frameworks. Results: Analysis results are distilled into a new Conceptual Model of Shared Health Informatics. We describe the people and work involved in chronic illness management and contribute a Conceptual Model of Shared Health Informatics depicting the process of chronic illness management in a social context. Conclusions: Through insight into management work provided by our Conceptual Model of Shared Health Informatics, technology designers and implementers can improve the quality of chronic illness management tools to ensure confident and capable management of chronic illness.

  • Background: Wrist-worn smart watches and fitness monitors (i.e., wearables) have become widely adopted by consumers and are gaining increased attention from researchers for their potential contribution to digital measurement of health in a scalable, and ecologically valid way. Various studies have begun to examine the accuracy of these devices in controlled laboratory settings (e.g., treadmill, stationary bike), yet no studies to date have investigated the accuracy of wearables during a continuous and ecologically valid 24-hour period that approximates actual consumer device use conditions. Objective: The goal of the present study was to determine the accuracy of two popular wearable devices, the Apple Watch 3 and Fitbit Charge 2, to assess heart rate accuracy as compared to the gold-standard reference method for real-world settings, an ambulatory electrocardiogram (ECG). Data was collected across five daily conditions, including sitting, walking, running, daily activities (e.g., cleaning, chores, work transitions), and sleeping. Methods: One participant, (BN; 29-year-old Caucasian male) completed a 24-hour ecologically valid protocol by wearing two of the most popular wrist wearable devices (Apple Watch 3 and Fitbit Charge 2). In addition, an ambulatory ECG (Vrije Universiteit Ambulatory Monitoring System) was used as the reference method during a normal day, which resulted in the collection of 102,740 individual heart beats. A single subject design was used to keep all variables constant except for wearable devices. Accuracy of these devices compared to the gold-standard ECG was assessed using percent error, Bland-Altman analyses, and concordance class correlation to assess agreement between devices. Results: The Apple Watch 3 and Fitbit Charge 2 were generally highly accurate across the overall 24-hour condition. Specifically, the Apple Watch 3 had a mean difference of -1.80 bpm, a mean error percent of -2.25%, and a mean agreement of 95% when compared to the ECG across 24 hours. The Fitbit Charge 2 had a mean difference of -3.47 bpm, a mean error of -4.25%, and a mean agreement of 91% when compared to the ECG across 24 hours. These findings varied by condition. Conclusions: The Apple Watch 3 and the Fitbit Charge 2 provided acceptable heart rate accuracy overall across a 24 hour continuous and ecologically valid period as compared to the gold-standard ECG, although during some conditions and devices error rose past the 5% acceptable error rate and some individual measurements were significantly erratic. Overall, these devices appear to be useful for implementing as ambulatory measures of cardiac activity in research studies, especially those where the specific advantages of these methods (e.g., scalability, low participant burden) are particularly suited to the population or research question.

  • Background: Transition to parenthood can be stressful to new parents in Singapore, especially with the lack of continuity of care from healthcare professionals during the postpartum period. Short hospital stays limit the availability of support and time parents need to be well-equipped with parenting and infant care skills. Poor parental adjustment may in turn lead to negative parental outcomes and adversely affect the child’s development. Therefore, for the family’s future wellbeing and in order to facilitate smoother transition to parenthood, there is a need for easily accessible technology-based educational programs that can support parents during this crucial perinatal period. Objective: To examine the effectiveness of a technology-based supportive educational parenting program (SEPP) on parenting outcomes during the perinatal period. Methods: A randomized controlled pretest-posttest experimental study design was used. The study recruited 236 patients (118 couples) from an antenatal ward of a tertiary hospital. Eligible parents were randomly assigned to the intervention group (n = 118) or the control group (n = 118). The SEPP is based on Bandura’s self-efficacy theory and Bowlby’s theory of attachment. Components of the intervention include two telephone-based education sessions and a mobile-health application follow-up for one month. Outcome measures including parental bonding, parental self-efficacy, perceived social support, parenting satisfaction, postnatal depression, and anxiety were measured using reliable and valid instruments. Data were collected over 12 months (December 2016 to December 2017) at four timepoints: during pregnancy (third trimester), immediately postpartum, one month postpartum, and three months postpartum. A linear mixed methods analysis was used to compare the percentage change of all outcome variables. Results: The intervention group had statistically significant improvements for parental bonding [P < 0.001, 95%CI -4.78 – -3.90], parental self-efficacy [P < 0.001, 95% CI: 1.94 – 2.79], social support [P < 0.001, 95% CI: 2.78 – 3.47], parenting satisfaction [P < 0.001, 95% CI: 3.14 – 3.83], postnatal depression [P < 0.001, 95% CI: -3.93 – -3.15], and postnatal anxiety [P < 0.001, 95% CI: -3.65 – -2.85] at one month and three months postpartum compared to the control group. A cessation of the intervention at one month postpartum showed a decline in all parental outcomes. Conclusions: The technology-based SEPP is effective in improving parental bonding, parental self-efficacy, perceived social support, parental satisfaction, and parental bonding. Hence, nurses and midwives should incorporate it with existing hands-on parent-craft classes and routine care. Further studies should investigate the long-term effectiveness (beyond three months) of this program on parental outcomes. Clinical Trial: ISRCTN48536064

  • Perceptions of expertise in food hypersensitivity on social media: An email interview study

    Date Submitted: Apr 18, 2018
    Open Peer Review Period: Apr 20, 2018 - Jun 15, 2018

    Background: Seeking and sharing information are primary uses of the internet and social media. It is therefore vital to understand the processes individuals go through when engaging with information on these diverse platforms; especially in areas such as health and risk-related information. One important element of such engagement is evaluating and attributing expertise to others. Objective: The study aimed to explore how meanings around expertise in relation to food allergy/intolerance (food hypersensitivity) were constructed by two groups of social media users; 1) those who use platforms for reasons relating to food hypersensitivity, and 2) those seen as experts by this community. Methods: Email interviews were conducted with food hypersensitive adults (n = 4), parents of food hypersensitive children (n = 4), and perceived experts in food hypersensitivity on social media (n = 5). Data were analysed thematically using Braun and Clarke’s approach. Results: The thematic analysis demonstrated that judging expertise on social media is a complex and multi-faceted process. Users might be judged as expert through their professional background, or their experience living with food hypersensitivities. How users behave on social media, and the traces of their online activity can influence how others will see them. Such considerations are both measured and moderated through the social media community itself. Findings highlighted how social media often acts as a supportive information tool following a diagnosis, but this also raised concerns if patients cannot access suitable vetted information. Conclusions: This work has implications for understanding how users perceive expertise on social media in relation to a health concern, and how information assessments are made during management of risks. Findings will prove beneficial to both medical and organisational stakeholders involved in the support of those living with life-changing conditions, such as food hypersensitivities.

  • Background: There has been a dearth of collaborative learning across tertiary institutions due to challenges in scheduling and geographical locations. Three-dimensional virtual environments are a viable and innovative tool to bring diverse healthcare students to learn together. Objective: The purpose of this study is to describe the development of a multiuser virtual environment and to evaluate healthcare students’ experiences of their collaborative learning in the environment. Methods: A mixed method study design was employed. Participants from six healthcare courses (Medicine, Nursing, Pharmacy, Physiotherapy, Occupational Therapy, and Medical Social Work) were recruited from three institutions to form six interprofessional healthcare teams who participated in interdisciplinary team care via a virtual environment. Four focus groups were conducted with 27 healthcare students after they completed questionnaires to evaluate their perceived usability, the sociability of computer-supported collaborative learning, and senses of presence. Interview transcripts were analyzed using thematic analysis. Results: Four themes emerged from the students’ experiences: “Bringing everyone together” to learn in the virtual platform was perceived as a valuable experience; “Feeling real”, whereby the students felt immersed in their own healthcare profession’s roles; learning in the virtual environment was perceived as “less threatening” compared to face-to-face interactions; and there were some “technical hiccups” related to sound quality and navigation. The participants reported positively on the usability (mean 3.48, SD 0.64) and feasibility (mean 3.39, SD 0.60) of the virtual environment in supporting collaborative learning. With a maximum score of 168, they also perceived a moderately strong sense of presence (mean 107.24, SD 17.78) in the environment. Conclusions: This study provides evidence for the acceptance of a virtual platform for collaborative learning in team care delivery. Given its flexibility, practicality, and scalability, this virtual platform serves as a promising tool for collaborative learning across different healthcare courses and institutions.

  • Continuous Subcutaneous Insulin Infusion and Its Characteristics in Young People with Type 1 Diabetes with Optimal Glucose Control

    Date Submitted: Apr 17, 2018
    Open Peer Review Period: Apr 19, 2018 - Jun 14, 2018

    Background: Continuous subcutaneous insulin infusion (CSII) is established model for mimicking physiological insulin, which can overcome some of these barriers. CSII therapy is becoming more popular treatment modality in young people with T1D.Several studies have demonstrated the superiority of CSII over MDI therapy in T1D patients in terms of improved glucose control, the rate of hypoglycemia, dawn phenomenon and glucose variability Objective: To describe Continuous Subcutaneous Insulin Infusion (CSII) characteristics in young people with Type 1 Diabetes (T1D) with optimal glucose control and to describe possible simple CSII settings for fine tuning. Methods: This retrospective study was conducted at University Clinic of Endocrinology, Diabetes and Metabolic Disorders in Skopje. The study enrolled CSII patients with T1D (age 12-25 years) visited our center from January to December 2016. Patient characteristics were collected through the electronic medical record system and CSII characteristics were obtained from 8 weeks reports prior to Hba1c, generated by Carelink Therapy Management Software (Medtronic, Northridge, USA). Results: Ninety-three patients were grouped according age: 12-18 years and 19-25 years. More than 70% of patients achieved HbA1c <7.5% (<58 mmol/mol). Significant difference in basal insulin was found between two age groups. Patients aged 12-18 years had five basal segments, less basal rate in early morning (03-07h) and slight decrease of afternoon basal rate (13-19h), comparing with patients aged 19-25 years with four basal segments, more basal rate in early morning (03-07h) and no decrease of afternoon basal rate (13-19h). Conclusions: Optimal glucose control is achievable in real life conditions among T1D patients on CSII therapy. Bolus wizard, frequent bolusing, multiple basal segments, and close follow up can be determinants for better control. Simple CSII settings as a tool, derived from our data may help clinicians to fine tune T1D patients and achieve optimal glucose control. Clinical Trial: EN16-12

  • Background: Vital signs monitoring is a universal tool for the detection of postoperative complications, but unwell patients can be missed in between traditional observation rounds. New remote monitoring technologies promise to convey the benefits of continuous monitoring to patients on general wards. Objective: The aim of this study was to evaluate whether continuous remote vital signs monitoring is a practical, acceptable and effective way of monitoring surgical patients. Methods: A prospective, cluster-randomised, parallel-group, controlled study was performed. Patients admitted to two surgical wards at a large tertiary hospital received either continuous and intermittent vital signs monitoring, or intermittent monitoring alone. The primary outcome measure was time to administration of antibiotics in sepsis. Secondary outcome measures included length of hospital stay, 30-day readmission rate, mortality and patient acceptability. Results: 350 patients were recruited between January and June 2017. 140 patients received continuous remote monitoring and 210 received intermittent monitoring alone. On average, patients receiving continuous monitoring were administered antibiotics faster after evidence of sepsis, had a shorter average length of hospital stay and were less likely to require readmission within 30 days of discharge. Wide confidence intervals suggest these differences are not statistically significant. Patients found the monitoring device to be acceptable in terms of comfort and perceived an enhanced sense of safety. Conclusions: Remote continuous vital signs monitoring on surgical wards is practical and acceptable to patients. Large, well-controlled studies in high-risk populations are required to determine if the observed trends translate into a significant benefit for continuous over intermittent monitoring. Clinical Trial: The study was prospectively registered on the ISRCTN registry on 19th December 2016 under registry number ISRCTN60999823; URL: https://doi.org/10.1186/ISRCTN60999823.

  • Predicting Current Hemoglobin A1c Values Among Adults with No Previous Evidence of Hyperglycemia

    Date Submitted: Apr 16, 2018
    Open Peer Review Period: Apr 18, 2018 - Jun 13, 2018

    Background: Electronic, personalized clinical decision support tools to optimize diabetes screening are lacking. Current screening guidelines are based on simple, categorical rules developed for populations of patients. Although personalized diabetes risk calculators have been created, none are designed to predict current glycemic levels using structured data commonly available in electronic health records (EHRs). Objective: The goal of this project was to create a mathematical equation for predicting the probability of hemoglobin A1c >= 5.7% among patients with no history of hyperglycemia using readily available variables that will allow integration with EHR systems. Methods: This study was conducted on all adult patients who have undergone a glycated hemoglobin test prior to evidence of hyperglycemia or any diabetes related diagnostic code in the EHR. Data was extracted from the Epicare EHR at Wake Forest Baptist Health in Winston-Salem, NC, which included dates between September 2012 and September 2016. Candidate predictor variables were chosen based on theoretical associations with hyperglycemia. Multiple logistic regression was used to model the association between the independent variables and the outcome of hemoglobin A1c >=5.7%. Continuous variables were fit using restricted cubic splines with 3-knots. The model was reduced using Harrell's model approximation method. The reduced model was compared head-to-head with calculators created by Baan (1999) and Griffin (2000). Ten-fold cross-validation was used to calculate the bias-adjusted prediction accuracy of the new model. Statistical analyses were performed in R 3.2.5 using the rms package. Results: The final model contained the following variables in order from most to least importance according to their impact on the discriminating accuracy of the model: age, body mass index, race, random glucose, serum non-High Density Lipoprotein, serum total cholesterol, smoking status, estimated glomerular filtration rate, serum triglycerides, and diagnosis of obesity. The new model achieved a Concordance Statistic of 0.75 which was statistically significantly better than prior models. The model appeared to be well calibrated according to a plot of the predicted probabilities versus the prevalence of the outcome at different probabilities. Conclusions: The personalized prediction model presented in this paper could improve the efficiency of screening initiatives aimed at detecting prediabetes and diabetes.

  • Bleeding Event Detection in EHR Notes Using CNN Models Enhanced with RNN Autoencoders

    Date Submitted: Apr 15, 2018
    Open Peer Review Period: Apr 17, 2018 - Jun 12, 2018

    Background: Bleeding events are common and critical which may cause significant morbidity and mortality. Studies show that high incidences of bleeding events are associated with cardiovascular disease (CVD) patients on anticoagulant therapy. Prompt and accurate detection of bleeding events are essential for preventing serious consequences. As bleeding events are often described in clinical notes, automatic detection of bleeding events from Electronic Health Record (EHR) narratives has the potential to improve drug safety surveillance and pharmacovigilance. Objective: We developed a natural language processing (NLP) system to automatically classify whether an EHR note sentence contains a bleeding event. Methods: We expert-annotated 878 EHR notes (76,577 sentences and 562,630 word tokens) for identifying bleeding events at the sentence-level. This annotated corpus was then used to train and validate our NLP systems. We developed an innovative hybrid CNN and LSTM Autoencoder model (HCLA), which integrates a convolutional neural network architecture (CNN) with a bidirectional Long-short term memory (BiLSTM) autoencoder model to leverage large unlabeled EHR data. Results: HCLA achieved an F-score of 93.79% for identifying whether a sentence contains a bleeding event, surpassing the strong baseline SVM and other CNN models. Conclusions: By incorporating a supervised CNN model with a pre-trained unsupervised BiLSTM Autoencoder, HCLA achieved a high performance in detecting bleeding events.

  • Background: While health literacy is important for people to maintain good health and manage disease, medical educational texts are often written beyond the reading level of the average individual. To mitigate this disconnect, text simplification research provides methods to increase readability and therefore comprehension. One method of text simplification is to isolate particularly difficult terms within a document and replace them with easier synonyms (lexical simplification) or an explanation in plain language (semantic simplification). Unfortunately, existing dictionaries are seldom complete and consequently resources for many difficult terms are unavailable. This is the case in English and Spanish resources. Objective: Our objective is to automatically generate explanations for difficult terms in both English and Spanish when they are not covered by existing resources. Our system combines existing resources for explanation generation with a novel algorithm (SubSimplify) to create additional explanations. Methods: SubSimplify uses word-level parsing techniques and specialized medical affix dictionaries to identify the morphological units of a term and then source their definitions. While the underlying resources are different, Subsimplify applies the same principles in both languages. To evaluate our approach, we used term familiarity to identify difficult terms in English and Spanish and then generated explanations for them. For each language, we extracted 400 difficult terms from two different article types (General and Medical topics) balanced for frequency. For English terms, we compared SubSimplify’s explanation with the explanations from the Consumer Health Vocabulary , WordNet Synonyms and Summaries, as well as Word Embedding Vector synonyms. For Spanish terms, we compared it to WordNet Summaries and Word Embedding Vector Embedding synonyms. We evaluated quality, coverage and usefulness for the simplification provided for each term. Quality is the average score from two subject experts on a 1-4 Likert scale (two per language) for the synonyms or explanations provided by the source. Coverage is the number of terms for which a source could provide an explanation. Usefulness is the same expert score however with a zero assigned when no explanation or synonyms was available for a term. Results: SubSimplify results in quality scores of 1.64 English (P<.001) and 1.49 Spanish ( P<.001), which is lower than that of existing resources (CHV=2.81). However, in Coverage SubSimplify outperforms all existing written resources; increasing the coverage from 53.0%-80.5% in English, and 20.8%-90.8% in Spanish (P<.001). This result means that the usefulness score of SubSimplify (1.32) (P<.001) is greater than most existing resources at (CHV=0.169). Conclusions: Our approach is intended as an additional resource to existing, manually created resources. It greatly increases the number of difficult terms for which an easier alternative can be made available resulting in greater actual usefulness.

  • Adverse Effects of eHealth Applications Replacing or Complementing Healthcare Contacts: A Scoping Review

    Date Submitted: Apr 14, 2018
    Open Peer Review Period: Apr 15, 2018 - Jun 10, 2018

    Background: The use of eHealth has increased tremendously in recent years. eHealth is generally considered to have a positive effect on healthcare quality and to be a promising alternative for face-to-face healthcare contacts. Surprisingly little is known about possible adverse effects of eHealth applications. Objective: We conducted a scoping review on empirical research into adverse effects of eHealth applications that aim to deliver healthcare on a distance. We investigated if adverse effects were reported and the nature and quality of research into these possible adverse effects. Methods: For this scoping review, we followed the five steps of Arksey and O’Malley’s scoping review methodology. We examined the literature between December 2012 and August 2017 in the following databases: PubMed, CINAHL, Web of Science and the Cochrane Library. Each paper was independently screened by at least two authors; differences were resolved through consensus development. Study characteristics were extracted. The methodological quality of the studies was assessed using the Critical Appraisal Skills Programme (CASP) lists. Results related to the research question were described and categorized. Results: Our search identified 79 papers that were potentially relevant; 11 studies met our inclusion criteria after screening. These studies differed in many ways and the majority was characterized by small research populations and low study quality. Despite the shortage of solid research, we found some indications of possible adverse effects Conclusions: There is a lack of high-quality empirical research on adverse effects of eHealth applications replacing or complementing face-to-face care. While the development of eHealth applications is ongoing, the knowledge with regard to possible adverse effects is limited. The available research focuses often on efficacy, added value, implementation issues, use and satisfaction, whereas adverse effects are underexposed. A better understanding of possible adverse effects might be a starting point to improve the positive impact of eHealth-based health care delivery.

  • Background: Chronic obstructive Pulmonary Disease (COPD) is a common chronic disease. Exacerbations of COPD (eCOPD) contribute to worsening of the disease and patient´s evolution. There are some clinical prediction rules that may help to stratify patients with eCOPD by their risk of poor evolution or adverse events. The translation of these clinical prediction rules into computer applications would allow their implementation in clinical practice. Objective: The goal of the study was to create a computer application to predict various outcomes related to adverse events of short-term evolution in eCOPD patients attending an emergency department (ED) based on valid and reliable clinical prediction rules. Methods: A computer application, PrEveCOPD (Prediction of Evolution of patients with eCOPD), was created for prediction of two outcomes related to adverse events: 1) mortality during hospital admission or within a week after the ED visit; and 2) admission to an intensive care unit (ICU) or an intermediate respiratory care unit (IRCU) during the eCOPD episode. The algorithms included in the computer tool were based on clinical prediction rules previously developed and validated within the IRYSS-COPD Study. The app was developed for Windows and Android systems, using Visual Studio 2008 and Eclipse, respectively. Results: The computer application PrEveCOPD implements the prediction models previously developed and validated for two relevant adverse events in the short-term evolution of patients with eCOPD. The application runs into Windows and Android systems and it can be used locally or remotely as a web application. Full description of the clinical prediction rules, as well as the original references, is included in the screen. Input of the predictive variables is controlled for out of range and missing values. Language can be switched between English and Spanish. The application is available for downloading and installing in a computer, as a mobile app or to be used remotely via internet. Conclusions: The PrEveCOPD app shows how clinical prediction rules can be summarized into simple and easy to use tools that allow the estimation of the risk of short-term mortality and ICU or IRCU admission for patients with eCOPD. The app can be used in any computer device, including smartphones or tablets, and it can guide the clinicians to a valid stratification of patients attending the ED with eCOPD.

  • Factors affecting the reach of messages in a dental twitter network: user popularity, communication pattern and network structure: a cohort study

    Date Submitted: Apr 13, 2018
    Open Peer Review Period: Apr 15, 2018 - Jun 10, 2018

    Background: Increasing the reach of messages disseminated through twitter promotes the success of twitter-based health education campaigns. Objective: This study aimed to identify factors associated with reach in a dental twitter network (1) initially and (2) sustainably, at individual and network levels. Methods: We used instructors’ and students’ twitter usernames from a Saudi dental school in 2016-2017, Gephi, social network analysis and social media analytics to calculate user and network metrics. We used content analysis to identify users disseminating oral health information (OHI). The study outcomes were reach at baseline and sustainably over 1.5 years. The explanatory variables were indicators of popularity (number of followers, likes, tweets retweeted by others), communication pattern (number of tweets, retweets, replies, tweeting/ retweeting OHI or not). Multiple logistic regression models were used to investigate associations. Results: Among dental users, 31.8% and 62.9% had reach at baseline and at the end of the study, reaching 749,923 and 37,169 users. At individual level, reach at baseline and sustainably over 1.5 years was associated with the number of followers (OR= 1.003, 95%CI= 1.001, 1.005 and OR= 1.002, 95%CI= 1.0001, 1.003), likes (OR= 1.001, 95%CI= 1.0001, 1.002 and OR= 1.0031, 95%CI= 1.0003, 1.002) and replies (OR= 1.02, 95%CI= 1.005, 1.04 and OR= 1.02, 95%CI= 1.004, 1.03). At network level, users with the least followers, tweets, retweets and replies had the greatest reach. Conclusions: Reach was reduced by time. Factors increasing reach at the user level had different impact at the network level. More than one strategy is needed to maximize reach.

  • Quantifying the relationship between diseases and symptoms using big data

    Date Submitted: Apr 12, 2018
    Open Peer Review Period: Apr 15, 2018 - Jun 10, 2018

    Background: Crises in endemic transmitted diseases affect humans worldwide, and the symptoms these diseases cause may provide firsthand information about these disorders. Objective: We suggest that massive new data sources resulting from human interaction with the Internet may offer a unique perspective on the relationship between illness and symptoms. Methods: By analyzing changes in Google query volumes for search terms related to disease, we find a pattern that may define the relationship between symptoms and disorders. We first retrieved pattern data from Google Trend using the common cold as the primary disease, and sore throat, stuffy nose, sneeze, fever, cough, and headache as symptoms. Pearson’s correlation coefficient was calculated using SPSS to determine the relationship between the symptoms and the disease. Results: Data created since 2013/1/13 was retrieved from Google Trend on a weekly basis. A total of 261 sets of data were calculated to create a high correlation coefficient of 0.925 between the common cold and the stuffy nose symptom. The cough symptom has the second highest correlation coefficient of 0.925, sore throat has a correlation coefficient of 0.853, and fever has a correlation coefficient of 0.626, which was significant at the 0.01 level in a two-tailed test. Conclusions: Data on the relationship between diseases and symptoms often comes from facilities such as government, hospitals, and clinics, where the data is collected through the documentation of physicians and nurses. A conventional study can be limited by the region, the number of patients and the interpretation of the specialist. However, with access to Google Trend’s big data, millions or even billions of data points are accumulated directly from the patient. Another contribution of this study is that the quantified relationship between symptoms and diseases can be used to educate future physicians or even artificial intelligence.

  • Background: Modern healthcare focuses on Shared Decision Making (SDM) because of its positive effects on patient satisfaction, therapy compliance, and outcomes. Patients’ knowledge about their illness and available treatment options, gained through medical education, is one of the key drivers for SDM. Current patient education relies heavily on medical consultation and is known to be ineffective. Objective: To determine whether providing patients with information in a subdivided, categorized, and interactive manner via an educational app for smartphone or tablet may increase the knowledge of their illness. Methods: A surgeon blinded randomized controlled trial was conducted with 213 patients who were referred to one of the six Dutch hospitals by their general practitioner owing to knee complaints that were indicative of knee osteoarthritis (OA). An interactive app that, in addition to standard care, actively sends informative and pertinent content to patients about their illness on a daily basis by means of push notifications in the week prior to their consultation. The primary outcome was the level of perceived and actual knowledge that patients had about their knee complaints and the relevant treatment options after the intervention. Results: In total 122 patients were enrolled in the control group and 91 in the intervention group. After the intervention, the level of actual knowledge (measured on a 0-36 scale) was 52% higher in the app group (26.4 versus 17.4, P < 0.001). Moreover, within the app group, the level of perceived knowledge (measured on a 0-25 scale) increased by 22% during the week within the app group (from 13.5 to 16.5, P < 0.001), compared to no gain in the control group. Conclusions: Actively offering patients information in a subdivided (per day), categorized (per theme), and interactive (video and quiz questions) manner significantly increases their levels of knowledge (both perceived and actual), compared to standard care educational practices. Clinical Trial: The trial was approved by the regional Medical Ethical Board (Maxima Medisch Centrum, Eindhoven, The Netherlands), reference number N16.130.

  • Online Reputation Management in Radiation Oncology

    Date Submitted: Apr 12, 2018
    Open Peer Review Period: Apr 13, 2018 - Jun 8, 2018

    Background: Online Reputation Management (ORM) is an emerging practice strategy that emphasizes the systematic and proactive monitoring of online reviews relating to one’s professional reputation. Objective: We developed this survey project to assess whether radiation oncologists are aware of ORM, and how it is utilized in their practices. We hypothesized that ORM is largely unknown by most practicing radiation oncologists, and that little time is spent actively managing their reputations. Methods: An online survey was submitted to 1,222 radiation oncologists using the Qualtrics research platform. Physician emails were gathered from the American Society for Radiation Oncology (ASTRO) member directory. 85 physicians initiated the survey while 76 physicians completed ≥94% (15/16) of the survey questions and were subsequently used in our analyses. The survey consisted of 15 questions querying practice demographics, patient satisfaction determination, ORM understanding, and activities to address ORM and one question for physicians to opt-in to a $50 Amazon card raffle. The survey data was summarized using a frequency table and data was analyzed using Chi-Square and Fisher’s Exact test. Results: We calculated a 7% (85/1,222) response rate for our survey with a completion rate of 89% (76/85). A majority of respondents (n = 68/76, 89%) endorsed being somewhat or strongly concerned about patient satisfaction (P < .001). However, 58% (n = 44/76) reported spending zero hours per week reviewing or managing their online reputation and 39% (n = 30/76) reported spending <1 hour per week (P < .001). A majority of physicians (n = 43/76, 57%) endorsed no familiarity with ORM (P < .001), and 70% (n = 53/76) did not actively manage their online reputation (P < .001). Although 83% (n = 63/76) of respondents strongly or somewhat believed that patients read online reviews (P < .001), 57% (n = 43/76) of respondents did not check their online reviews (P = .25) and 80% (n = 61/76) endorsed never responding to online reviews (P < .001). 58% (n = 44/76) strongly or somewhat supported the idea of managing their online reputation going forward (P = .0012). Conclusions: ORM is presently under recognized in radiation oncology. Although most practitioners are concerned about patient satisfaction, little effort is directed towards the internet on this matter. ORM offers an area of practice improvement for many practicing radiation oncologists.

  • Background: Videoconferencing-based treatments have shown great potential in increasing engagement and compliance by decreasing the barriers of time and distance. In general, employees tend to experience a lot of stress, but find it difficult to visit a clinic during office hours. Objective: The purpose of this study was to investigate the effectiveness of a mobile videoconference-based intervention for stress reduction and resilience enhancement in employees. Methods: Eighty-one participants were randomly allocated to one of three conditions: mobile videoconferencing, in-person, and self-care; among them, 72 completed the study. All participants underwent assessment via self-reported questionnaires before, immediately after, and 1-month after the intervention. Intervention lasted for 4 weeks and consisted of elements of cognitive behavior therapy, positive psychology, and meditation. Changes in clinical variables regarding stress and resilience across time were compared between treatment conditions. Results: There were significant condition × time effects on variables measuring perceived stress, resilience, emotional labor, and sleep. Moreover, there were significant effects of condition on perceived stress and occupational stress. Conclusions: Results indicate that both mobile videoconferencing and in-person interventions were comparably effective in decreasing stress and enhancing resilience. Further studies with a larger sample size and a longer follow-up period are warranted to investigate the long-term effect of mobile videoconferencing intervention. Clinical Trial: ClinicalTrials.gov identifier: NCT03256682

  • Background: Numerous published articles show that physicians do not follow clinical practice guidelines. However, few studies explore what physicians consider evidence and how they use different forms of evidence in their care decisions. Many of the existing studies on how physicians use evidence occurred before the advent of smart phones and advanced online information retrieval technologies. Objective: It is important to understand how these new technologies influenced the ways that physicians use evidence in their practice. Methods: In this paper, we draw on ethnographic data collected through shadowing internal medicine teams at two teaching hospitals to understand the roles that scientific evidence plays for attending physicians and trainees when caring for patients. Results: Clinical practice guidelines represent just one of several sources of scientific evidence that are used when making care decisions. The majority of scientific evidence was accessed online, often through smart phones. Forms of evidence were used differently depending on the level of experience of the person drawing on the evidence and were often blended together to arrive at shared understandings and approaches to patient care. In applying evidence to care, internal medicine team members are cognizant that scientific evidence is evolving, occasionally of low quality, can make competing claims, and does not cover all clinical problems. In moving from incorporating summaries of scientific evidence to primary sources of scientific evidence into their care decisions, trainees and attending clinicians increasingly add scientific uncertainty to the medical uncertainty that is inherent in their practice. Conclusions: This paper outlines one way that the ethos of evidence-based medicine has been incorporated into the daily work of care. Here multiple online forms of evidence were mixed with other information. This is different from the pathway that is often articulated by health administrators and policy makers whereby clinical practice guideline adherence is equated with practicing evidence-based medicine.

  • Change in Waist Circumference with Continuous Use of a Smart Belt: An Observational Study

    Date Submitted: Apr 12, 2018
    Open Peer Review Period: Apr 13, 2018 - Jun 8, 2018

    Background: Several insurers and health policymakers have exerted effort to prevent and reduce cardiovascular diseases due to obesity. A new concept for conquering obesity has been introduced, and policymakers have started adopting new strategies. Objective: This preliminary study aimed at evaluating whether a waist circumference of smart belt users improved. Methods: Preexisting data on waist circumference and self-reported demographics were obtained from the WELT corporation . In the database, 451 male subjects were registered initially; 24 were excluded because of missing data. Demographic and waist circumference data of the subjects were collected. Paired t-test and repeated measurements analysis of variance (ANOVA) were used to identify the change in waist circumference by considering time intervals. In addition, a linear mixed model was used to incorporate all subjects’ waist circumference data at each time point. Results: Compared with the baseline waist circumferences, the waist circumference (inches) decreased significantly at all time points (weeks) (week 4, β=-0.106; week 8, β=-0.300; week 12, β=-0.776; P<.01 ). Although each paired t-test had a different sample size because of loss to follow-up, the differences between baseline and each subsequent week increased. Equal continuous reduction in waist circumference was observed with ANOVA and mixed model analysis (β=-0.056, every week). Conclusions: A smart belt is a newly developed, wearable device that measures real-time steps, sedentary time, and waist circumference. In this study, the smart belt helps users decrease waist circumference after 12 weeks This direct-to-consumer smart health device may contribute toward reducing the risk of non-communicable diseases and controlling the increasing health cost for health insurers. Clinical Trial: None

  • Background: Children necessarily rely on their parents with respect to oral health routines but parental behavior in this area is not always adequate. Web-based interventions are a promising way of improving parental behavior in the area of oral health. Objective: To evaluate the effectiveness of a web-based film about oral health, oral hygiene and dental care targeting parental knowledge and behavior with respect to oral health, directly after exposure and six months later. Methods: We performed a non-blinded quasi-experimental study of the effects of a 8.5 minute web-based film in Dutch, English, Turkish and Moroccan/Berber about the oral health of children. Parents attending well-child clinics in mixed urban and rural areas in the Netherlands were assigned to an intervention (n=88) or control group (n=41). The control group received care as usual. We measured parental knowledge and behavior with respect to oral health before and directly after the intervention and six months later, and assessed differences between the intervention and the control group. Results: Parental knowledge about oral health improved: the mean knowledge score was 93% correct after watching the film in the intervention group by comparison with 59% correct in the control group, Cohen’s d=2.64, and it was still better in the intervention group six months after watching the film. Oral health behavior did not differ between the intervention group and the control group after the intervention. Conclusions: A web-based educational film for parents is an effective way to address specific health topics like oral health and to improve parental knowledge.

  • Comparison of developers’ and end-users’ perspectives about smoking cessation support through the Crush the Crave app.

    Date Submitted: Apr 10, 2018
    Open Peer Review Period: Apr 12, 2018 - Jun 7, 2018

    Background: High smoking rates among end-users, combined with their high rates of app use, render this age group as a particularly captive audience for quit smoking apps. There is emerging evidence that apps are an effective way to support smoking cessation among end-users. How the expectations behind the design of apps align with the needs and preferences of users, and if this differs by gender, is poorly understood, limiting the ability to evaluate and scale these interventions. Objective: The objective of this qualitative case study was to detail how the overall design approach of Crush the Crave (CTC), a quit smoking app that targets end-users, compares with young adult women’s and men’s perspectives and experiences, with consideration for the influence of gender. Methods: Semi-structured interviews were conducted with 15 developers involved in the development of CTC, and 31 young adult CTC users. Data were analyzed inductively to derive thematic findings in relation to the perceived pros and cons of CTC by both developers and end-users. Findings were grouped under four categories 1) technology and platforms utilized for the app, 2) foundation of app content, 3) underlying focus of the app, and 4) look, feel and functionality of the app. Results: 1) Technology and platforms utilized for the app: both developers and end-users agreed that apps aligned with the needs and preferences of young adult smokers. Major limitations with the technology identified by end-users were the frequent “glitches” and requirement for WiFi or data. 2) Foundation of app content: while developers agreed that the strength of CTC was in its strong evidence-base, what mattered to end-users was that the content was packaged in a positive way, focusing on the benefits of quitting versus the consequences of smoking. 3) Underlying focus of the app: that the app entailed an individually-led focus resonated with both developers and end-users, especially young men. 4) Look, feel and functionality of the app: while developers were more positive about the look and feel of the app, end-users did not like the aesthetics of the app because it incited a negative affect. Also, while end-users found it easy to use, they did not find the app intuitive. Finally, end-users thought that, because the app functions were largely based on a user’s quit date versus their ongoing efforts, this often lent to unmeaningful data. Conclusions: The current study findings highlight the importance of understanding multiple perspectives of stakeholders involved in a mobile-based intervention. By gathering the viewpoints of developers and end users, both problematic and effective approaches that underlie development goals were revealed as a means to informing end-users the development, implementation and evaluation of future eHealth interventions.

  • Measuring the quality of mobile applications for the management of pain: a standardised review using the mobile app rating scale (MARS)

    Date Submitted: Apr 9, 2018
    Open Peer Review Period: Apr 11, 2018 - Jun 6, 2018

    Background: Chronic pain is a major health issue requiring an approach that not only considers medication, but also many other factors included in the Bio-Psycho-Social model of pain. New technologies, such as mobile applications, are tools to address these factors, although in many cases they lack proven quality or are not based on scientific evidence, so it is necessary to review and measure their quality. Objective: The aim is to review and measure the quality of mobile applications for the management of pain using the mobile app rating scale (MARS). Methods: This review included 18 pain related mobile apps from App Store and Android Market. The MARS was administered to measure their quality. We list the scores (of each section and the final score) of every app and we report the mean score (and standard deviation) for an overall vision of the quality of the pain related apps. We compare the section scores between the groups defined according to the tertiles via ANOVA analyses, after checking for normality (Kolmogorov-Smirnov test). Results: The global quality ranged from 1.74 (worst app) to 4.35 (best app). The 18 apps, as a whole, obtained an average score of 3.17 (SD=0.75). The best rated sections were “Functionality” (Mean=3.92;SD=0.72), “Aesthetics” (Mean=3.3;SD=1.05), and “Engagement” (Mean=2.87;SD=1.14), while the worst rated were “App specific” (Mean=2.48;SD=1), “Information” (Mean=2.52;SD=0.82), and “App subjective quality” (Mean=2.68;SD=1.22). The main differences between tertiles were found on “app subjective quality”, “Engagement”, “Aesthetics”, and “App specific”. Conclusions: Current pain related apps are of a certain quality, mainly regarding their technical aspects, although they fail to offer information and have an impact on the user. Most apps are not based on scientific evidence, have not been rigorously tested, and the confidentiality of the information collected is not guaranteed. Future apps would need to improve these aspects and exploit the capabilities of current devices.

  • What the health? Information sources and maternal lifestyle behaviours

    Date Submitted: Apr 9, 2018
    Open Peer Review Period: Apr 11, 2018 - Jun 6, 2018

    Background: Regular physical activity (PA), adequate gestational weight gain (GWG) and healthy eating are important for the long-term health of both mother and baby. Hence, it is important that women receive current and updated advice on these topics, and are encouraged to adopt a healthy lifestyle during pregnancy. Objective: The aim of this study was to investigate the main information sources among pregnant women regarding PA, GWG and nutrition, as well as evaluate how these information sources may affect their health behaviours. Methods: A cross-sectional study design, comprising an electronic questionnaire, was distributed to two antenatal clinics, as well as pregnancy-related online chat forums and social media. Inclusion criteria were ≥ 18 years, ≥ 20 weeks gestation and able to read/write Norwegian. In total, 150 pregnant women answered the questionnaire, which was a mix of 11-point Likert scales, close-ended questions and semi close-ended questions. The relationship between information sources and selected variables, including health behaviours and descriptive variables, were assessed by logistic regression, linear regression or X2 as appropriate (p < 0.05). Results: Mean age (years), gestation week and pre-pregnancy BMI (kg/ m2) were 31.1 (± 4.3), 30.6 (± 5.9) and 24.2 (± 4.2), respectively. More than eight out of ten had received or retrieved information about nutrition (89%) and PA (80%), whereas 54% reported information on GWG. When combining all three lifestyle factors, 38.5% had retrieved information from blogs and Internet forums, and 26.6% from their midwife or family physician. Reporting Internet and media as primary source increased the odds of gaining below the IOM guidelines (P = 0.02), but was also associated with higher compliance with nutritional guidelines (P = 0.03). Family and friends’ advice was significantly associated with gaining above the IOM guidelines (P = 0.03). Otherwise, no other associations were found between information sources and health behaviours. Conclusions: The small number of health professionals giving information, and that about 40% use blogs and Internet forums, highlight the need to address the quality of advice and guide women towards trustworthy sources of information during pregnancy. The association between information sources and PA, GWG and nutrition requires further research.

  • Background: Non-adherence to antidepressants is a major obstacle to antidepressants therapeutic benefits, resulting in a significant burden on individuals and the healthcare system. Several studies showed that non-adherence is weakly associated with personal and clinical variables, but strongly associated with patients’ beliefs and attitudes towards medications. Patients’ drug review posts in online healthcare communities may provide a significant insight into patients’ attitudes towards antidepressants and address the challenges of self-report methods, such as patient recruitment. Objective: The primary objective of this study is to use consumer health posts to identify factors affecting the patient’s attitude towards antidepressant drugs, which in turn, is a strong determinant of treatment non-adherence. Methods: We randomly selected 892 patient drug reviews from the healthcare forum “askapatient.com”. We used the Framework Method to build the analytical framework containing the themes for developing structured data from the qualitative drug reviews. In the next step, four annotators coded the drug reviews at the sentence level using the analytical framework. To test the hypotheses, we first managed missing values using different imputation methods. Then, we used chi-square and ordinal logistic regression to test and model the association between variables and attitude. Results: Analysis of the sample include drug reviews that were posted from February 2001 to February 2016. The majority of the patients (76%; n= 680) were female. Approximately two-thirds of the patients (n=540) were less than 40 years of age. Testing the association between variables and attitude showed that “experience of adverse drug reactions (ADRs)” (x^2= 31.11, P= 2.91e-06), “perceived distress from ADRs” (x^2= 231.6, P= 2.2e-16), “drug effectiveness” (x^2= 548.52, P= 2.2e-16), “complaint about the lack of knowledge” (x^2= 59.36, P= 3.96e-12), “experience of withdrawal” (x^2= 55.63, P= 2.4e-11), and “duration of usage” (F-value= 43.66, P = 6.76e-11) were significantly associated with patients’ attitudes towards antidepressants. While variables “age” (F-value= 0.72, P = 0.4) and “gender” (x^2= 2.7, P= 0.21) were not associated with the patient attitudes. Moreover, modeling the relationship between variables and attitudes showed that “drug effectiveness” and “perceived distress from adverse drug reactions” were the two most significant factors affecting patient attitudes toward antidepressants. Conclusions: Patients self-report experiences of medications in online healthcare communities, can provide a unique insight for identifying underlying factors associated with patients’ perceptions and attitudes towards antidepressants. However, it cannot be used as a replacement for self-report methods due to the lack of information for some of the variables, colloquial language, and the unstructured format of the reports.

  • Listening to the HysterSisters: Online Conversations about Hysterectomy Recovery

    Date Submitted: Apr 7, 2018
    Open Peer Review Period: Apr 9, 2018 - Jun 4, 2018

    Background: Despite increased focus on patient-centered care, physicians remain pressured by time constraints. Physicians must optimize perioperative counseling during this limited time. Although patients experience major complications infrequently, minor complications are more common and can negatively impact quality of life. These complications may be inadequately addressed during clinical counseling prior to surgery. As a result, less time may be spent discussing common but non-life threatening symptoms. However, these still significantly impact patient quality of life. This creates an information gap that patients attempt to bridge, often through phone calls during the recovery period. Increasingly, patients are turning to social media such as online health communities to seek answers and reassurance. One such community is HysterSisters, which is “dedicated to medical and emotional issues surrounding the hysterectomy experience […], supporting women from diagnosis, to treatment, to recovery.” It offers a unique view into patients’ conversations surrounding hysterectomy and other gynecologic care. Objective: To characterize conversation about hysterectomy recovery as experienced by members of the HysterSisters online community. Methods: A retrospective keyword frequency and survival analysis of the HysterSisters Hysterectomy Recovery forum. Results: Over 1.78 million posts in total were captured from the HysterSisters forum. Among the users who created these posts were 69,950 patients who declared their hysterectomy type and date. Among these posts were also 80,704 threads from the Hysterectomy Recovery board that were posted since 2005 and during the author’s first 12 weeks post-op by 33,311 unique patients who declared hysterectomy type and date. A symptom taxonomy of 8 primary symptom groups was created using a seed list of symptoms generated from a three-gram keyword frequency analysis of these threads. Patient post volumes and distribution of symptom mentions over time reveal key times for potential beneficial symptom-specific patient engagement during recovery. Pain, bleeding and hormones/emotions were the three most mentioned symptom groups and account for over half of all symptom mentions. Minimally invasive hysterectomies were compared against abdominal hysterectomy. There is no significant difference in the proportion of patients who mention pain- and hormone/emotion-related keywords during their recovery across all procedures. As compared to the abdominal hysterectomy group, mentions of bleeding-related keyword mentions were more common in the laparoscopic group, while urinary and intimacy-related keywords were more common in the vaginal hysterectomy group. Nevertheless, the interquartile difference in days for the final mention of these keywords is significantly earlier across all keyword groups for the minimally invasive procedures. Conclusions: Analysis of the HysterSisters Hysterectomy Recovery forum reveals a rich source of information about patient experience and provides novel actionable insights that can improve patient-centered care during the postoperative period.

  • Background: Timely and effective dissemination of the latest clinical evidence to healthcare providers is essential for translating biomedical research into routine patient care. Online platforms offer unique opportunities for medical knowledge dissemination. Objective: In this study, we report: (1) the profiles of healthcare providers participating in the New York State HIV-HCV-STD Clinical Education Initiative (CEI) online program; and (2) their evaluations of the CEI online continuing medical education (CME) and continuing nursing education (CNE) courses. Methods: We compiled the professional and personal background information of the clinicians who completed at least one CEI online CME/CNE course. We collected their self-reported program evaluation data with regard to the course content, format, change in knowledge, and impact to clinical practice. Results: We recorded a total of 4,363 completions of 88 online CME/CNE courses by 1,976 unique clinicians during a 12-month study period. The clinicians’ background was much diversified with regard to the demographics, education levels, professional disciplines, practice years, employment settings, caseloads, and clinical services. The evaluation of the CEI online CME/CNE courses was very positive (usefulness/relevant 91.08%, easy comprehension 89.09%, knowledgeable trainer 92.00%, appropriate format 84.35%, knowledge increase 48.52%, intention to use knowledge 85.26%, and plan to change practice 21.98%). Comparison with the reference data indicated that the CEI online program successfully reached out to the primary care communities. Both the younger generation and the senior healthcare providers were attracted to the CEI online program. High quality multimedia resources, flexibility of access, ease of use, and provision of CME/CNE credits contributed to the initial success of the CEI online program. Conclusions: We have successfully characterized a diversified group of clinicians participating in the CEI online program. The evaluation data have demonstrated effective use of online resources to disseminate HIV, HCV, and STD clinical evidence to primary care clinicians.

  • Background: Cognitive impairments after stroke are not always given sufficient attention despite the important limitations they impose in activities of daily living (ADL’s). Although there is important evidence on cognitive rehabilitation benefits, its implementation is limited due to time and human resources demands. Moreover, many cognitive rehabilitation interventions lack a solid theoretical framework in the selection of paper-and-pencil tasks by the clinicians. In this endeavor, it would be useful to have a tool that could generate standardized paper-and-pencil tasks, parameterized according to patients’ needs. Objective: In this paper, we present a framework for the creation of personalized cognitive rehabilitation tasks based on a participatory design strategy. Methods: We have selected eleven paper-and-pencil tasks from standard clinical practice and parameterized them with multiple configurations. A total number of sixty-seven tasks were assessed according to their cognitive demands (attention, memory, language and executive functions) and overall difficulty by twenty rehabilitation professionals. Results: After assessing the internal consistency of the data (α values from .918 to .997), we identified the parameters that significantly affected cognitive functions and proposed specific models for each task. Through computational modeling we operationalized the tasks into their intrinsic parameters and developed a web-tool that generates personalized paper-and-pencil tasks – the Task Generator (TG). Conclusions: Derived from expert knowledge, our framework proposes an objective and quantitative personalization strategy tailored to each patient in multiple cognitive domains (attention, memory, language and executive functions), materialized in the TG application, an online available cognitive rehabilitation Web Tool.

  • Background: Technology is nowadays omnipresent. Using New Information and Communications Technologies in the health sector is promising and intended to be beneficial, both for patient disease prevention and care. The Lab e-Santé, a French think tank, aims to further advance knowledge of the use and practices of digital health tools, identify obstacles in their use, and issue recommendations. Objective: The Lab e-Santé conducted an exploratory study in order to get preliminary data regarding the appropriation, satisfaction and global use in real life of health applications and connected objects. This initial survey would allow to explore, within MHCs (Multidisciplinary health centers), the usability and the perceived contribution of the tools from both HCPs and patients’ viewpoints and provide preliminary inputs regarding the barriers and the drivers of connected health. Methods: The Lab e-santé recruited a total of 10 facilities from the exhaustive database of French MHCs. HCPs from those MHC were provided with iPads on which the m-health application was installed. They were also provided with three types of connected devices: weighting scales, blood pressure monitors and activity trackers. VirtualSanté and VirtualSanté Pro applications were especially developed for the survey. VirtualSanté allowed to compile all the patient’s data collected from a connected health device, and share these data with the health professional in their dedicated professional application, VirtualSanté Pro. During a 5-month period, participating health professionals were instructed to include patients with chronic diseases, aged 18 years and above, who agreed to participate. Results: This exploratory study addressed the perceived contribution and usability of health technologies from both patients and their HCPs. Our findings as described below suggest that there is a dichotomy between HCPs and patients’ perception about connected health usefulness. HCPs are more confident in a potential contribution of connected health for their patients than the patients themselves. Patients didn’t express any benefit 6 months later and the device became a source of anxiety for some of them. One possible explanation may be the multiple technical issues users faced during the survey including the installation of the applications on non-compatible devices, the initial connection between the object and the application or technical bugs during use. For the same reasons, HCPs found it time-consuming and difficult to integrate in their everyday practice. Conclusions: Today many people are optimistic about the concept of mHealth. However, to use these devices and have high acceptance, both HCPs and patients need to be aware of the existence and usefulness of applications and connected objects then trust them. Tools should be personalized and adaptive, as far as possible, or at least be adapted to the good user profile. The latter should be motivated, properly trained, educated and supported for assuring an appropriate, useful and sustainable use.

  • Background: Clinical trials targeting ophthalmic conditions are currently underway and some gene therapies targeting ocular diseases have been proven to be safe and efficacious/present a positive risk:benefit; its potential as a therapeutic tool has yet to be fully exploited in clinical ophthalmology. Pharmacogenomics suggests that diseases with similar symptomatic presentations often have varying genetic causes, affecting an individual patient’s response to a specific therapeutic strategy. Gene therapies and somatic cell therapies offer unique therapeutic pathways for ocular diseases and often depend on increased understanding of the genotype-phenotype relationship in disease presentation and progression. While demand for personalised medicine is increasing and the required molecular tools are available, its adoption within paediatric ophthalmology remains to be maximised in the post-genomic era. To address the individual hurdles encountered in the field of genomic-related clinical trials and facilitate the uptake of personalised medicine, we propose to conduct a review that will examine and identify the digital technologies used to facilitate data analysis in somatic and gene therapy trials in paediatric patients with ocular diseases. Objective: To present an outline of HIT/ICT resources used in somatic and gene therapy clinical trials in children with ocular diseases. This review will enable authors to identify challenges and provide recommendations facilitating the uptake of genetic and somatic therapies as therapeutic tools in paediatric ophthalmology. The review will also determine whether conducting a systematic review will be beneficial. Methods: The review will be guided by Arksey/O’Malley’s and Levac’s methodological frameworks. Following the development of Medical Subject Headings (MeSH)/subject headings and keywords, a systematic search of MEDLINE/PubMed, Embase and Scopus will be conducted. Two suitably qualified independent reviewers will determine study eligibility. Following identification of studies, data will be extracted and analysed. Results: Database searches will be initiated in May 2018. We expect to complete the review in August 2018. Conclusions: Based on review findings, the authors will summarise methods used for facilitating IT integration in personalised medicine. Additionally, it will identify further research gaps and determine whether conduction of further reviews will be beneficial. Clinical Trial: Because this is a scoping review protocol, it is ineligible for registration with PROSPERO.

  • Designing a digital person-centered self-management support intervention for people with type 2 diabetes

    Date Submitted: Apr 5, 2018
    Open Peer Review Period: Apr 7, 2018 - Jun 2, 2018

    Background: Self-management is a substantial part of treatment for patients with type-2 diabetes (T2D). Modern digital technology, being small, available and ubiquitous has promising properties that might work well in supporting self-management. This study follows the process of designing a digital person-centered self-management support intervention for people with T2D. Objective: The aim of this paper was to explore and describe the preparatory activities and the subsequent choice of digital support for a digital person-centered self-management support intervention for people with T2D treated in primary health care. Methods: A user-centered, mixed-methods approach was used to collect and analyze data from different sources. Results: Results show that the requirements for a digital platform for T2D self-management care is complex and has to be negotiated, often with trade-off relations. It is also shown that the focus on user needs is complicated as nurses’ project organizational and regulatory needs that causes limitations in the design space. Conclusions: A user-centered approach can be useful in identifying goals for design that both show possibilities as highlight challenges and limitations. There are limits in the user-centered approach and thereby we suggest more research to find ways to deal with the nurse-organization perspective. Clinical Trial: This study is a preparatory part of a larger randomized intervention project aimed at designing and implementing person-centred interactive self-management support (iSMS) in primary healthcare in northern Sweden. The overall project has a user-centred design, described in this paper and the project is registered at ClinicalTrials.gov, Identifier NCT03165084.

  • TECHsex: Youth Sexual Health and Relationships Online

    Date Submitted: Apr 4, 2018
    Open Peer Review Period: Apr 6, 2018 - Jun 1, 2018

    Background: According to a 2015 report from PEW Research Center, nearly 24% of teens go online almost constantly and 92% of teens are accessing the internet daily. To explore the sexual health information-seeking behaviors and sexual relationship building practices of youth online. In addition, young people also turn to the web for sexual health resources and romantic relationship building. This drastic increase in accessibility and feasibility has changed the way young people are able to access information, build relationships, and socialize. An examination of the unique relationships young people have with their sexual health and relationships online and the implications for adolescent health programs has yet to be fully explored. Objective: Although there is a growing concern for sexually transmitted infections (STIs), human immunodeficiency virus (HIV), and healthy romantic relationships among young people, many programs continue to neglect online spaces as locations for health interventions. Little is known about the online sexual health practices of young people, including digital flirting and online dating. This data provides information relevant to youth sexual health at the emergence of social media tools and online dating, which provides insights into youth behavior and intervention opportunities. Methods: Through January and December 2016, the TECHsex study used a mixed-methods approach to document information-seeking behaviors and sexual health building behaviors of youth online in the US. A national quantitative survey of 1,500 youth and twelve qualitative focus groups (66 youth) were triangulated to understand the experiences and desires of young people as they search for sexual health information online and begin building their sexual relationships through social media, online chat, and online dating. Results: Sixty-two percent of young people throughout the US reported using the internet, most notably Google, as a source for sexual health information. At the same time, respondents most often reported a doctor, nurse, or clinic to be the single most effective way to learn about sexual health (30%). Young people are not only turning to online resources for sexual health information; they are also using the internet to begin sexual relationships with others, including dating, online flirting, and “hooking up”, with 34% of young people reporting the use of online dating apps and 70% having met a partner they met online in real life. Conclusions: While youth are turning to the internet for sexual health information, they are also turning to online spaces to build sexual relationships. Sexual relationship building included online flirting and online dating websites and/or apps. These findings have implications for sexual health, risk, and implementing future programs interested in enhancing youth sexual health. We implore researchers and organizations to consider the relationships young people have to technology for these sexual health purposes and implement these platforms to create successful and youth-centered programs.

  • Implementing a Web-Based Intervention in HIV Primary Care Clinics: A Pilot Study on the Feasibility of Positive Health Check

    Date Submitted: Apr 4, 2018
    Open Peer Review Period: Apr 5, 2018 - May 31, 2018

    Background: Web-based interventions can help people living with human immunodeficiency virus (HIV) achieve better clinical outcomes and behaviors, but integrating them into clinical practice remains challenging. Several studies report on the appropriate content for web-based interventions and their outcomes, but there is a gap in understanding the feasibility of implementing these interventions in HIV clinic settings from the clinicians’ perspective. Objective: We examined whether Positive Health Check (PHC)—a web-based, tailored video counseling tool—was acceptable, appropriate, and feasible for HIV primary care clinics to implement. Methods: Four primary care clinics implemented PHC during a 1-month pilot. Nine key clinic staff across clinics who implemented the intervention participated in 54 interviews. We used framework analysis to code the data and identify themes related to implementation outcomes; acceptability, appropriateness, and feasibility. We also analyzed patients’ (n=104) intervention usage metrics to quantify patients’ intervention completion rates (n=68). Results: At the start of implementation, and with increasing frequency as enrollment was opened to more diverse patient groups, clinic staff reported that PHC was an acceptable and appropriate intervention. Stakeholders successfully dedicated space for patients to use the tool, delivered patients’ handouts, and adjusted workflow to facilitate implementation. Clinic staff reported that PHC is feasible to implement in clinic settings, but challenges remained for workflow integration and the best time to administer PHC. Conclusions: To increase implementation effectiveness, we examined implementation outcomes related to the integration of PHC into clinic workflows. Findings from this pilot implementation suggest that PHC can be integrated into clinic workflow systems, but clinicians and researchers need to better understand the role of implementation outcomes to leverage the benefits of tailored web-based interventions for people living with HIV.

  • Looking for and having sex with partners met online: the Second Australian study of health and relationships

    Date Submitted: Apr 4, 2018
    Open Peer Review Period: Apr 5, 2018 - May 31, 2018

    Background: Studies have reported on the proportion of the population looking for potential sexual partners online, but few have investigated those who have sex with these partners, arguably a more important target group for health promotion. Objective: To determine the proportion of people who have had sex someone met online in the previous year. Methods: Data were analysed from the 2012–13 Second Australian Study of Health and Relationships, a nationally representative telephone survey of 16–69-year-old Australian residents (n=20,091). The participation rate for the telephone survey was 66%. The prevalence of looking for a potential partner, physically meeting, and having sex with someone first met online (through internet site or smartphone application) was estimated. Multivariate logistic regression was used to determine demographic and behavioral factors associated with having had sex with someone met online in the last year, for men and women separately. Results: Overall, 12% of respondents had looked for potential partners online, 5% had done so in the last year. In the last year, 3% had met someone in person, and 2% reported having had sex with someone first met online. Prevalence of all behaviors were greater in men than women, and in younger than older respondents. Among sexually active men, factors associated with having had sex with someone met online included: identifying as gay or bisexual (Adjusted odds ratio [AOR]: 15.37, 95% confidence interval [CI]: 8.34 –28.35), having either 2–3 or >3 sexual partners in the last year (AOR: 9.20, 95% CI: 9.20 – 34.68 and 35.77, 95% CI: 18.04 – 70.94, respectively), having had an STI test in the past year (AOR: 2.02, 95% CI: 1.21 – 3.38) or an STI in the last year (AOR: 3.15, 95% CI: 1.25 – 7.97). Among sexually active women, factors associated with having had sex with someone met online were: having either 2–3 or >3 sexual partners in the last year (AOR: 32.01, 95% CI: 13.17 – 77.78 and AOR: 71:03, 95 % CI: 27.48– 183.57, respectively), very low/low income (vs. very high AOR: 3.40, 95% CI: 1.12–10.35) and identifying as lesbian or bisexual (AOR: 2.27, 95% CI: 1.04–4.49). Conclusions: More than a third of adults who had looked for potential partners online each year had sex with such partners, and those who had done so were more sexually active, suggesting that dating and hook-up websites and applications are suitable settings for targeted sexual health interventions.

  • Background: Long term recovery takes longer than expected despite improved surgical techniques and Enhanced Recovery After Surgery programs. An eHealth care program (‘ikherstel’) was developed to partially substitute perioperative care for patients undergoing colorectal surgical procedures. Successfully tested eHealth programs are not always implemented in usual care and it is therefore important to evaluate the process to optimize future implementation. Objective: To evaluate whether the eHealth intervention was executed as planned. Methods: A mixed method process evaluation was carried out alongside a multicenter randomized controlled trial (RCT). This evaluation was performed using the Linnan and Steckler framework for the quantitative part of this study measuring the components reach, dose delivered, dose received, fidelity and participants’ attitudes. Total implementation scores were calculated using the averaging approach, in which the sum of all data points is divided by the number of data points and measured the total adherence to the protocol. For the qualitative part the Unified Theory of Acceptance and Use of Technology framework (UTAUT) was used. The quantitative data was based on participants’ questionnaires, a logistic database, a weblog and participants’ medical files and was obtained by performing semi-structured interviews with participants of the RCT. Results: A total of 151 participants of 340 eligible patients were included in the RCT, of which 73 participants were allocated to the intervention group. Based on the quantitative process data, total implementation scores for the website, mobile application, eConsult and activity tracker of 64%, 63%, 44% and 67% were calculated. Participants in the qualitative part experienced the program as supportive and provided guidance on their recovery process after colorectal surgery. Most frequently mentioned barriers were the limited interaction with and feedback from health care professionals and the lack of tailoring of the convalescence plan in case of a different course of recovery. Conclusions: The intervention needs more interaction with and feedback from health care professionals, and needs more tailored guidance in case of different recovery or treatment courses. To ensure a successful implementation of the program in daily practice, some adjustments are required to optimize the program in a blended care form. Clinical Trial: NTR5686.

  • Top-Cited Articles In Medical Professionalism: A Bibliometric Analysis versus Altmetric Scores

    Date Submitted: Apr 2, 2018
    Open Peer Review Period: Apr 4, 2018 - May 30, 2018

    Background: Citation counts of articles have been used by universities and funding bodies to measure scientific outcomes and assess suitability for grant applications. However, citation counts are not without limitations. With the rise of social media, altmetric scores may provide an alternative assessment tool. Objective: The aims of this study are to assess the characteristics of highly cited articles in medical professionalism and altmetric scores. Methods: The Web of Science was searched for top-cited articles in medical professionalism, and the characteristics of each article were identified. The altmetric database was searched to identify report for each article. A model to assess the relationship between the number of citations and each of key characteristics as well as altmetric scores were developed. Results: No correlations were found between the number of citations and number of years since publication (p=0.192), number of institutes (p=0.081), number of authors (p=0.270), females in authorship (p=0.15), or number of grants (p=0.384). The altmetric scores varied from zero to 155, total= 806, median=5.0, (IQR=20). Twitter (54%) and Mendeley (62%) were the most popular altmetric resources. No correlation was found between the number of citations and the altmetric scores (p=0.661). The top topics covered were learning and teaching professionalism, curriculum issues, professional and unprofessional behavior, defining and measuring professionalism. Only the number of countries correlated with the number of citations (p= 0.001). The articles were mainly published in Academic Medicine, Journal of the American Medical Association, Journal of General Internal Medicine, and Annals of Internal Medicine. Conclusions: Except for the number of countries involved, no correlation was found between citations and any of the article parameters. Altmetric scores of articles were not significantly correlated with citations. Highly cited articles were produced mainly by the United States, Canada, and the United Kingdom. The study reflects the emerging role of altmetric and social media in the dissemination of research. Future studies should investigate the specific features of highly cited articles and factors that reinforce distribution of research data among scholars and non-scholars. Clinical Trial: Not applicable.

  • Background: Telediagnosis is a promising tool in situations of limited access to specialty care. Considering that, an oral health telediagnosis platform (EstomatoNet) was created in the state of Rio Grande do Sul, Brazil. EstomatoNet provides specialist support over a web-based platform to primary care dentists for diagnosis of oral lesions. Objective: To evaluate the usability of EstomatoNet and to identify user perceptions regarding their expectations and difficulties with the system. Methods: Sixteen dentists were selected for the study: 8 were frequent users of EstomatoNet and 8 were residents who had never used the Platform. To assess usability, participants were required to request telediagnosis support for a fictional case provided by the research team. During the process of uploading the information and sending the request, users were asked to “think out loud,” expressing their perceptions. The session was observed by an examiner with remote access to the user's screen (via Skype). After the simulation, users completed the System Usability Scale (SyUS). Results: The mean SyUS score assigned by frequent users was 84.7±6.6, vs. 82.2±9.3 for residents (satisfactory usability: score above 68). The difference between the groups was not statistically significant (Student t test, P = .55). The inexperienced group took longer (347.1s±101.1) to complete the task than frequent users (252.8s±80.3), however the difference between the groups was not statistically significant (Student t test, P = .06). In their subjective evaluation, users suggested the inclusion of a field to add further information on outcomes and resolution of the case and changes in the position of the “Send” button to improve work flow. Conclusions: The present results indicate good usability of EstomatoNet. The Platform seems to meet the needs of users regardless of how experienced they are; nevertheless, a few minor changes in some steps would improve the tool.

  • Background: There is increasing use of on-line surveys to improve data quality and timeliness, and to reduce costs. While there have been numerous cross-sectional studies comparing responses to on-line or paper surveys, there is little research from a longitudinal perspective. Objective: In the context of the well-established Australian Longitudinal Study on Women’s Health, we examined the patterns of responses to on-line or paper surveys across the first two waves of the study in which both modes were offered. We compared: 1. Differences between women born in 1946-51 and in 1973-78; 2. Types of device used for on-line completion; 3. The socio-demographic, behavioral and health characteristics of women who responded on-line or using mailed paper surveys; 4. Associations between mode of completion of the first survey, and participation and mode of completion in the second survey. Methods: Participants in this study, who had responded to regular mailed surveys since 1996, were offered a choice of completing surveys using paper questionnaires or web-based electronic questionnaires, starting in 2012. Two groups of women were involved: an older cohort born in 1946-51 and aged in their 60s, and a younger cohort born in 1973-78 aged in their 30s when the on-line surveys were first introduced. We compared women who responded on-line on both occasions, women who responded on-line at the first survey and used the paper version of the second survey, women who changed from paper to on-line, and those who used paper for both surveys. Results: Of the 9663 women in their 60s who responded to one or both of the surveys, more than 50% preferred to continue with paper surveys (5290 at the first survey and 5373 at the second survey). If they chose the on-line version most used computers. In contrast, of the 8628 women in their 30s, more than 50% (4835) chose the on-line version at the first survey. While most favored computers to phones or tablets, many did try these alternatives at the subsequent survey. Many women who completed the survey on-line the first time preferred the paper version at the subsequent survey. In fact, for women in their 60s, the number who went from on-line to paper (n=1151) exceeded the number who went from paper to on-line (n=734). The on-line option was more likely to be taken up by better educated and healthier women. In both cohorts, women who completed paper surveys were more likely than on-line completers to become non-respondents at the next survey. Due to the large sample size almost all the differences were statistically significant with P < .001. Conclusions: Despite the cost-saving advantages of on-line compared to paper surveys, paper surveys are likely to appeal to a different population of potential respondents, with different socio-demographic, behavioral and health characteristics and potentially greater attrition from the study. Not offering a paper version is therefore likely to induce bias in the distribution of responses unless weighting for respondent characteristics (relative to the target population) is employed. Therefore, if mixed mode (paper or on-line) options are feasible, they are highly likely to produce more representative results than if only the less costly on-line option is offered.

  • Background: Experiential learning plays a critical role in learner development. Kolb’s four-part experiential learning model consists of concrete experience, reflective observation, abstract conceptualization, and active experimentation in a recurring cycle. Most clinical environments provide opportunities for experiences and active experimentation, but rarely offer structured means for reflection and abstract conceptualization that are crucial for learners to learn through experience. Traditional approaches such as narrative medicine and recent use of online blogs to encourage reflection are too narrowly focused on professionalism and ethical topics and fail to address crucial gaps in the experiential learning cycle in everyday clinical learning. We created Learning Moment, a novel web-based educational tool that integrates principles of asynchronous learning and learning portfolios to fulfill the reflection and abstract conceptualization aspects of Kolb’s learning cycle in the modern clinical learning environment. Medical students log concise clinical “pearls” in the form of “learning moments” for reflection, review, and sharing with peers in a community of practice. Objective: We sought to evaluate learners’ experiences with LM via a qualitative study. Methods: We implemented Learning Moment at a busy, urban, tertiary care emergency department that hosts an emergency medicine residency and robust third and fourth year medical student clerkships. We encouraged rotating medical students to use Learning Moment on a voluntary basis to log self-selected learning experiences in the form of a “learning moment” and to view fellow students’ “learning moments” through a “Community Feed”. Experienced clinical faculty facilitated monthly in-person “Learning Moment Reflection” small group discussion with medical students as they reflected on, and expounded upon their own and peer “learning moments” with faculty guidance. We employed purposive sampling to recruit medical students who used Learning Moment during their rotation. We conducted 13 interviews (10 individual interviews and one 3-person group interview) between January and March 2017 using an ethnographic approach and utilized a general inductive method to analyze and code for potential themes. Results: Thirteen students voluntarily participated in our qualitative interviews. Emerging from our analysis were three key themes relating to the perceived impact of Learning Moment on student learning: 1) The process of logging of “learning moments” enhanced memorization; 2) Improved learning through reflection; and 3) Sharing of knowledge and experiences in a community of practice. Conclusions: Learning Moment was successfully implemented into the educational infrastructure in our department. Students identified three mechanisms by which the application optimizes experiential learning, including: enabling the logging of “learning moments” to promote memorization, encouraging reflection to facilitate learning, and fostering the sharing of knowledge and experiences within a community of practice. The Learning Moment concept is potentially scalable to other departments, disciplines, and institutions as we seek to optimize experiential learning ecosystems for all trainees.

  • Background: Online medical service provision is increasingly common. However, it is unclear how physicians are responding to this trend and how online and offline medical services are linked. Objective: The objectives of this study were to examine physician use of social media for STD consultation and identify the physicians who frequenly used social media to evaluate STD patients. Methods: In June 2017, we conducted a nationwide cross-sectional survey among physicians registered on a physician mobile phone application (app) in China. We collected data on physicians’ demographic information, institutional information, and online medical practices. We compared physicians who used social media to evaluate STD patients frequently (at least once a week) to infrequent users. Bivariate and multivariate logistic regressions were used to identify physicians who frequently evaluated STD patients on social media. Results: A total of 501 physicians participated in the survey. Among them, three-quarters were men and the average age was 37.6 ±8.2 years old. Nearly all physicians (92.8%) recommended their last online STD patient to subsequently see a physician in the clinic. More than half of physicians recommended HIV/STD testing to online patients, and 43.9% provided treatment advice to STD patients. Over twenty percent of physicians used social media to evaluate STD patients online more than once a week. 85.2% conducted follow-up consultation for STD patients using social media. Physicians working at institutions with STD prevention materials (aOR =2.10, 95%CI: 1.18 -3.74) were associated with frequent evaluation of STD patients on social media. Conclusions: Physicians increasingly use social media to provide clinical services. The high rates of referral to clinics suggests that social media are used to promote clinic-seeking, and not replace it. Social media medical services may be useful for STD patients and sexual minorities who hesitate to seek formal clinic-based services.

  • A large-scale study of food purchases and health outcomes: study and implications

    Date Submitted: Mar 31, 2018
    Open Peer Review Period: Apr 1, 2018 - May 27, 2018

    Background: Conditions associated to the metabolic syndrome have an enormous impact on people’s health. It is estimated that more than 300k premature deaths in Europe are caused by obesity and more than half of European citizens will be obese by 2050. That represents a heavy burden on healthcare, with more than €70B spent every year in Europe. Since healthy eating is one of the most effective intervention to counter the risks of metabolic syndrome, monitoring food consumption at scale is key for effective prevention. Traditional nutrition studies are costly and, most often, of limited scale. To partly fix that, researchers have resorted to digital data to infer what people eat and to estimate how that relates to their health. Objective: For the entire city of London, we study the association between food purchases in grocery stores, as measured by the digital traces of customer loyalty cards, and consumption of medicines prescribed to treat different aspects of the metabolic syndrome. Methods: We combine two sets of geo-referenced data. The first contains the record of all the 1.6B food items that 1.6M loyalty card owners bought in 2015 in all the London stores of Tesco, the largest grocery retailer in UK. From it, we extract food nutrients and other indicators of food healthiness. The latter contains information about all the 1.1B medical prescriptions written by every General Practitioner in 2016 in London, from which we infer the incidence of three aspects related to the metabolic syndrome: hypertension, high cholesterol, and diabetes. We study the data at the level of 942 Middle Super Output Areas, geographical areas containing an average of about 8,000 residents. Using correlations, regression, and classification models, we explore the association between and medicine prescriptions and food purchases. Results: Areas with high incidence of the ailments considered are characterized by high consumption of fat (Spearman rank correlation ρ~[.28, .36]), sugar (ρ~[.31, .49]), and especially carbohydrates (ρ~[.45, .57]) and low consumption of protein (ρ~[-.22, -.50]) and fibre (ρ~[-.19, -.50]). A regression model that uses the diversity of nutrients and calorie intake can explain more than half of the variability in diabetes incidence (coefficient of determination R2=0.598) and about a third for high cholesterol (R2=0.345) and hypertension (R2=0.388). A classifier trained over all the available features can identify unhealthy areas from their food consumption with up to 91% accuracy. Conclusions: Our study shows that analytics of digital records of grocery purchase can be used as a cheap and scalable tool for health surveillance: the distribution of the food nutrients is far more predictive of food-related illnesses than socio-economic conditions. Being able to identify areas at higher health risk is just a first step. We detail how different stakeholders could help with prevention strategies.

  • Effectiveness of a Mobile Prenatal Care App to Reduce In-Person Visits: A Prospective Trial

    Date Submitted: Mar 28, 2018
    Open Peer Review Period: Mar 31, 2018 - May 26, 2018

    Background: Risk appropriate prenatal care has been asserted as a way for cost effective delivery of prenatal care. A virtual care model for prenatal care has the potential to provide patient-tailored, risk-appropriate prenatal educational content and may facilitate vital sign and weight monitoring between visits. Previous studies have demonstrated safe reduction of frequency of in person prenatal care visits in low risk patients, but have noted a reduction of patient satisfaction. Objective: The primary objective of this study is to test the effectiveness of a mobile prenatal care app to facilitate a reduced in-person visit schedule, for low risk pregnancies, while maintaining patient and provider satisfaction. Methods: This controlled trial compared a control group receiving usual care to an experimental group receiving usual prenatal care plus a mobile prenatal care app. The experimental group had a planned reduction in the frequency of in-person office visits while the control group had the usual number of visits. The trial was conducted at two diverse outpatient obstetric practices that are part of a single academic center in Washington, DC, USA. Women were eligible for enrollment if they presented to care in the first trimester, were between the ages of 18 and 40, had a confirmed desired pregnancy, were not considered “high-risk,” and had an iOS or Android smartphone that they used regularly. We measured the effectiveness of a virtual care platform for prenatal care via the following measured outcomes: (1) the number of in-person Obstetric (OB) visits during pregnancy; (2) patient satisfaction with prenatal care; (3) and provider satisfaction. Results: 88 patients were enrolled in the study, 47 in the experimental group and 41 in the control group. For patients in the experimental group, the average number of in-person OB visits during pregnancy was 7.8 and the average number in the control group was 10.2 (P=0.01204). There was no statistical difference in patient satisfaction or provider satisfaction in either group. Conclusions: The use of a mobile prenatal care app was associated with reduced in-person visits and without a reduction in patient or provider satisfaction. Clinical Trial: ClinicalTrials.gov: NCT02914301

  • Background: Background: Instagram (a popular image-based social media app) with millions of posts each day can be used to inform public health interventions and policies but current research relying on image-based data often relies on hand coding of images which is time consuming, costly, and may be subject to researcher bias. What is more, current best practices in automated image classification (e.g., support vector machine (SVM), Backpropagation (BP), and artificial neural network) are limited in their capacity to accurately distinguish between objects within images. Objective: The goal of this study is to demonstrate how convolution neural network (CNN) can be used to extract unique features within an image and how SVM can then be used to classify the image. Methods: Images of waterpipes or hookah (an emerging tobacco product possessing similar harms to that of cigarettes) were collected from Instagram and used in analyses (n=840). CNN was used to extract unique features from images identified to contain waterpipes. SVM classifier was built to distinguish between images with and without waterpipes. Methods for image classification were then compared to show how CNN + SVM classifier could improve accuracy. Results: As the number of the validated training images increased, the total number of extracted features increased. Additionally, as the number of features learned by the SVM classifier increased, the average level of accuracy increased. Overall, 99.5% of the 420 images classified were correctly identified as either hookah or non- hookah images. This level of accuracy was shown to be some improvement over earlier methods that used SVM, CNN or Bag of Features (BOF) alone. Conclusions: CNN extracts more features of the images allowing the SVM classifier to be better informed which results in higher accuracy compared with methods that extract fewer features. Future research can use this method to reduce computational time in identifying millions of images. By taking images of waterpipes from Instagram, we place our methods in a context that can be utilized to later inform health researchers. Automated and scalable approaches to identifying waterpipes in images from Instagram may facilitate reaching out to Instagram users who may need tobacco-related education to reduce misconceptions about hookah use.

  • How did Ebola Information Spread on Twitter: Broadcasting or Viral Spreading?

    Date Submitted: Mar 27, 2018
    Open Peer Review Period: Mar 28, 2018 - May 23, 2018

    Background: It has been argued that information and emotions towards public health issues could spread widely through online social networks. Although aggregate metrics on the volume of information diffusion are available, we know little about how information spreads on online social networks. For example, health information could be transmitted from one to many (i.e. broadcasting), which is similar to how traditional mass media passes information to the general public. Health information could also be transmitted from many to many (i.e. viral spreading), which is analogous to the spread of infectious diseases. Objective: The aim of this study is to determine the spreading pattern of Ebola information on Twitter and identify influential users regarding Ebola messages. On Twitter, influential users are those whose tweets receive a large number of retweets. Methods: Our data was purchased from GNIP, the official Twitter data provider. We obtained all Ebola-related tweets (including retweets and replies) posted from March 23, 2014 to May 31, 2015. We reconstructed Ebola-related retweeting paths based on Twitter content and the follower-followee relationships (who follows whom on Twitter). Social network analysis was performed to investigate retweeting patterns. In addition to describing the diffusion structures, we classify users in the network into four categories (i.e., influential user, hidden influential user, disseminator, common user) based on following and retweeting patterns. Disseminators received fewer retweets than expected based on their number of followers, common users and influential users received as many or fewer retweets than expected, and hidden influential users received more retweets than expected. Results: On average, 91% of the retweets were directly retweeted from the initial message. Moreover, 47.5% of the retweeting paths of the original tweets had a depth of 1 (i.e., from the seed user to its immediate followers). These observations suggested that the broadcast model was more pervasive than viral spreading. Furthermore, we found that influential users and hidden influential users can trigger more retweets than disseminators and common users. Disseminators and common users relied more on the viral model for spreading information beyond their immediate followers via influential and hidden influential users. Conclusions: The broadcast model was the dominant mechanism of information diffusion of a major health event on Twitter. It suggests that public health communicators can work with influential and hidden influential users to get the message across, because influential and hidden influential users can reach more people that are not following the public health Twitter accounts. Although both influential users and hidden influential users can trigger a lot of retweets, recognizing and using the hidden influential users as the source of information could potentially be a cost-effective communication strategy for public health promotion, because the hidden influential users can receive more retweets than expected based on their limited number of followers. However, challenges remain due to uncertain credibility of these hidden influential users.