Maintenance Notice

Due to necessary scheduled maintenance, the JMIR Publications website will be unavailable from Monday, March 11, 2019 at 4:00 PM to 4:30 PM EST. We apologize in advance for any inconvenience this may cause you.

Who will be affected?

Advertisement

Citing this Article

Right click to copy or hit: ctrl+c (cmd+c on mac)

Published on 26.03.20 in Vol 22, No 3 (2020): March

This paper is in the following e-collection/theme issue:

    Review

    Risks and Benefits of Web-Based Patient Narratives: Systematic Review

    1Institute of Biomedical Ethics and History of Medicine, University of Zurich, Zurich, Switzerland

    2Institute of Health Sciences, Zurich University of Applied Sciences, Winterthur, Switzerland

    Corresponding Author:

    Daniel Drewniak, PhD

    Institute of Biomedical Ethics and History of Medicine

    University of Zurich

    Winterthurerstrasse 30

    Zurich, 8006

    Switzerland

    Phone: 41 44 634 40 82

    Email: daniel.drewniak@ethik.uzh.ch


    ABSTRACT

    Background: Patient narratives are illustrative, individual accounts of patients’ experiences with certain health conditions. Web-based patient narratives have become widely available on the internet and in social media, as part of electronically available patient decision aids or Web-based databases. In recent years, the role of patient narratives as a source of information, insight, and support for both health care users and providers has increasingly been emphasized. Although the potential impact of patient stories has high immediate plausibility, it is of interest to know if this impact can be captured in quantitative studies.

    Objective: This review aimed to evaluate whether research-generated Web-based patient narratives have quantifiable risks or benefits for (potential) patients, relatives, or health care professionals.

    Methods: We searched the following databases from August 2017 to March 2019: Medical Literature Analysis and Retrieval System Online, PsycInfo, Sociological Abstracts, Web of Science, and EMBASE. Titles and abstracts of the retrieved studies were reviewed and assessed for the inclusion and exclusion criteria. Papers were included if they studied the following: (1) (potential) patients, relatives, or health care professionals; (2) the effects of Web-based patient narratives that were generated scientifically (eg, through qualitative research methods); and (3) were quantitative studies. Furthermore, 2 authors independently performed an assessment of the quality of the included studies using a validated checklist.

    Results: Of 4226 documents, 17 studies met the inclusion criteria. The studies investigated 10 different sources of Web-based patient narratives. Sample sizes ranged from 23 to 2458. The mean score of the quality assessment was 82.6 (range 61-100). Effects regarding five different purposes were identified as follows: provide information, engage, model behavior, persuade, and comfort. We found positive effects in every category and negative effects in one category (persuade). Several of the reported effects are rather small or were identified under specific experimental conditions.

    Conclusions: Patient narratives seem to be a promising means to support users in improving their understanding of certain health conditions and possibly to provide emotional support and have an impact on behavioral changes. There is limited evidence for beneficial effects on some outcomes. However, narratives are characterized by considerable heterogeneity and the investigated outcomes are hardly comparable with each other, which makes the overall judgment difficult. As there are numerous possible measures and purposes of narratives, quantifying the impact of Web-based patient narratives remains a challenge. Future research is needed to define the optimal standards for quantitative approaches to narrative-based interventions.

    J Med Internet Res 2020;22(3):e15772

    doi:10.2196/15772

    KEYWORDS



    Introduction

    Background

    In their recent report, the Lancet Global Health Commission calls for an improved integration of patient experiences in the evaluation of health care systems, including experiences about competent care, health care utilization, or confidence in the health care system [1]. Such experiences can be collected by using tools such as patient satisfaction surveys. Although quantitative data about patient experiences are essential measures for the accountability and improvement of health care systems, they fall short of capturing a more comprehensive picture of how patients experience health care encounters or illnesses [1,2].

    Patient narratives are illustrative accounts of individual patients’ experiences with a certain illness [3] and are available on social media sites, in patient decision aids, and on databases such as the Database of Individual Patients’ Experiences (DIPEx). There is neither a clear definition of what constitutes a narrative nor any guidance on the length or content [4], which may lead to conflicting research results about the effects of patient narratives because of insufficient operationalization of the term [3].

    Patient narratives are a promising tool that can support people in coping with their illness [5], serve as a resource for preparing health care decisions [6], or help identify questions for physicians [7]. Characteristically, narratives can retrospectively structure actions in ways that convey perceived causality; they are nonlinear and powerful in making sense of complex, emergent phenomena [8]. Furthermore, stories transport images and emotions, which makes them evocative and memorable. Most people recall stories better than statistical information expressed in graphs or numbers [2,8].

    Several qualitative studies report that illness narratives enjoy high acceptance among other patients [9]. Furthermore, positive effects of personal health and illness experiences, including improvements in decision making [10,11] or addressing information needs [12], were identified in qualitative studies. Narratives have a high potential to add unknown insights into patient-focused issues, which can only be provided by a person who has the respective lived experience. For example, as a World Health Organization report states, “qualitative methods help to present narratives that broadly reflect the gendered social norms about parent-child relations. They also provide ‘lived experiences’ from ageing populations about how satisfied they are with the life they have lived” [2].

    On the contrary, there are also serious concerns about the use of patient narratives because they are powerful message formats [13] and are suspected to override statistical information [14,15]. The concern is that patients’ decision-making regarding treatment options could be based on personal experiences of a few, whereas statistical data remain largely ignored [2,14,15]. Furthermore, patients’ experiences presented on the Web may contain unbalanced or misleading messages, which may lead to a manipulation of choices in favor of a particular health care option [16]. A study among mothers of children with genetic disorders, eg, found that several parents put more trust on online communities than on their physicians [17]. Such findings are especially problematic when stories in such communities are biased.

    In recent years, internet platforms, patient blogs, and fora have become important means for individuals to seek information relevant to health, including information describing how other individuals live with illnesses. Such websites often provide biomedical information but lack information on wider experiences [18] or the experiences are not systematically collected, analyzed, and presented [19]. Therefore, in this review, we focused on studies that used established scientific methodologies to elicit the stories [8,20].

    Objectives

    This systematic review aimed to evaluate whether research-generated Web-based patient narratives have quantifiable risks or benefits for patients, relatives, or health care professionals. Patient narratives are understood as immediate personal experience reports.


    Methods

    Search Strategy

    This review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis guidelines [21]. To identify relevant studies, the databases Medical Literature Analysis and Retrieval System Online, PsycInfo, Sociological Abstracts, Web of Science, and EMBASE were searched from August 2017 to March 2019. A search term was developed and was adjusted to the different databases. The search terms were tested and evaluated by the study team. In addition, the search strategy was discussed and evaluated with a member of Cochrane Switzerland and with an employee from the University library who specialized in systematic reviews. The search terms were adjusted based on the discussion and recommendations. The search terms consisted of the following: [Narration: narration, personal narratives, narrative medicine, anecdot*, testimonial*] + [Databases: internet, bibliographic database, online, Web based] + [Participants: patient, health care personnel, relative*, caregivers] + [Study: Surveys and Questionnaires, controlled clinical trials, cohort studies] (Multimedia Appendix 1).

    Selection Criteria

    Titles and abstracts of the retrieved studies were reviewed and assessed for inclusion and exclusion criteria independently by all members of the study team (DD, AG, MH, and NB). Researchers were trained in applying the predefined selection criteria. Nonagreements were discussed until consensus was reached. Papers were included if they (1) studied (potential) patients (with or without an established diagnosis or condition), relatives (or other nonrelated informal caregivers), or health care professionals; (2) studied the effects of Web-based patient narratives (real experiences or fictional stories; collections or single narratives; presentation as text or audio or video clips) that were generated scientifically (eg, through qualitative research methods and not just stories put selectively on the Web with a view to their human interest for marketing purposes); and (3) were quantitative studies such as surveys and questionnaires, observational studies, nonrandomized controlled trials (non-RCTs), RCTs, comparative effectiveness research, cohort studies, or longitudinal studies. We excluded studies that used qualitative study designs such as interview studies, focus groups, or ethnographic studies and studies that were neither published in English or German. Studies that used narratives that were not generated by a scientific method were also excluded (eg, unmoderated blogs or fora). Furthermore, we excluded studies published before 2000 and studies that examined narratives not Web-based. We made no restrictions on the inclusion of studies regarding content, context, length, or depth of the narratives. We decided to focus on Web-based narratives as we felt the range would have been too broad to allow for meaningful comparisons had we included narratives available in different media (books, leaflets, newspapers, etc).

    Quality Assessment

    A protocol was written about all the steps of data collection and analysis, including selection of studies and extraction of content. Researchers were trained in applying the predefined selection criteria. Overall, 3 researchers reviewed and assessed all studies (DD, AG, and MH), whereas nonagreements were discussed with a fourth independent expert (NB). Evaluation tools designed for conventional systematic reviews typically assess the quality of RCTs. However, the diversity of research designs and outcome measures of the included studies required the use of a tool that is able to systematically appraise disparate evidence stemming from different study types. Therefore, 2 authors (DD and AG) independently performed an assessment of the quality of the included studies using the checklist proposed by Hawker et al [22]. This validated checklist consists of nine evaluation sections: abstract and title, introduction and aims, method and data, sampling, data analysis, ethics and bias, results, transferability or generalizability, and implications and usefulness. Each section was assessed by giving a score ranging from 1 to 4 (4=good, 3=fair, 2=poor, and 1=very poor), resulting in a potential score range of 9 to 36. Similar to the Appraisal of Guidelines for Research and Evaluation II instrument [23], we calculated sum scores for each section and an overall score, scaled as a percentage of the maximum possible score over all sections:

    obtained score−minimum possible score/maximum possible score−minimum possible score×100

    Data Extraction

    A data elicitation form was developed and applied systematically to all publications included in the review by 1 author (DD). The form includes information about background characteristics (authors, year of publication, and location), study characteristics (aim, sample size, participants, and study design), narrative (type of narrative and degree of exposure), study measures (attitudes and beliefs, psychometric scales, and preferences), and a summary of findings.

    Data Synthesis

    We extracted study results as they were reported in the results section of the publications. The analysis was based on the comparison of study details using descriptive statistics and text. The analysis was mainly focused on the identification of similarities and differences between the findings of the individual studies. As the study aims, designs, and findings were too heterogeneous, a meta-analysis was not conducted.

    The specific outcomes of the studies were grouped using the taxonomy proposed by Shaffer and Zikmund-Fisher [3]. As several of the included studies provided few details about the content of the narratives, the studies were grouped around the purpose of the narrative. According to Shaffer and Zikmund-Fisher [3], five different purposes of narratives can be described. As most of the studies focus on (potential) patients rather than on relatives or health care professionals, the Shaffer and Zikmund-Fisher [3] taxonomy is suitable for our review. The purposes and their possible outcomes as proposed by Shaffer and Zikmund-Fisher [3] are described in Table 1.

    Table 1. Purposes of narratives.
    View this table

    The definition of effective and preference-sensitive decisions proposed by Wennberg et al [24,25] was applied to assign the outcomes of the included studies to risks and benefits categories: outcomes were assigned to the risk category when they were preference sensitive. In preference-sensitive decisions, the best decision for an individual is unclear because of two reasons: the evidence for specific treatments is inadequate and firm conclusions about risk-to-benefit ratios cannot be drawn and the risk-to-benefit ratio might be clear, but it depends on the patients’ values [24,25]. Outcomes were assigned to the benefit category when they were effective following the definition by Wennberg et al [24,25]. In these cases, the best decision is clear to practitioners and patients. The clinical evidence of harms and benefits is known, and compared with the benefits, the harms are minimal. In effective decisions, there is a widespread consensus among clinicians and patients about known and favorable risk-to-benefit ratios [24,25]. The outcomes of the included studies were assigned to a no-effect category when the corresponding studies reported experimental conditions inferior to the control group or no statistically significant effects (significance level chosen by the individual study). In descriptive studies, thresholds such as significance levels are not available. Therefore, outcomes of descriptive studies that were mentioned by ≤50% of the participants were also assigned to the no-effect category.


    Results

    Literature Search

    Our search strategy identified 4226 documents. Of these, 60 documents potentially fulfilled the inclusion criteria of the study and were assessed in full text. After assessing the full texts, 43 more studies were excluded for specific reasons, including, eg, the study did not focus on systematically generated narratives or the narratives were not Web-based. There were 95.50% (4036/4226) agreements among the raters. Finally, 17 studies were included in the analysis (Figure 1).

    Figure 1. Study flowchart.
    View this figure

    Description of Included Studies

    The studies were taken from Germany (n=5), the United States (n=6), the United Kingdom (n=4), the Netherlands (n=1), and Switzerland (n=1) and covered the period from 2000 to 2018 (Table 2). They investigated 10 different Web sources. The Web source and their specifications are shown in Multimedia Appendix 2.

    Table 2. Characteristics of the included studies.
    View this table

    Sample sizes of the studies ranged from 23 to 2125 (samples of the following substudies were combined: Betsch et al [14], Sullivan et al [36], and Winterbottom et al [37]) with a median of 302 per study. Most of the included studies focus on the effects on (potential) patients. Only one study that met our inclusion criteria focused on (future) health care professionals (medical students) [35]. The measures of the studies were (1) general perceptions of patient narratives, including patients’ expectations and learning experiences [18,30], self‐help and use of patients’ stories [39], and empowering processes and outcomes experienced by website participants [26]; (2) effects of narratives on patients’ and health care professionals’ behavior, including health care participation [38]; information search, treatment intentions, and decision satisfaction [32,33]; self-efficacy coping with cancer and patient competence [29]; physical activity [27,34]; health care utilization and medication overuse [27]; and performance in examinations [35]; and (3) decision making about dialysis modality [37], tobacco cessation treatment [28], vaccination [14,15], reflux and blood pressure drugs [36], and early-stage breast cancer [31,33].

    The degree of exposure to the narrative was reported by 11 out of 17 studies. The reporting included measures such as self-reporting regarding frequency of website visits [26,34], mean visiting times of the websites [27] or mean times spent on the corresponding websites [15,29-31], the length of the narrative videos or expected study length [33,35], the number of participants exposed to the narratives [36], and the timespan for which participants had access to the corresponding websites [38].

    The mean score of the quality assessment was 84.5 (range 61-100). The main issues were concerning appropriate sampling strategies [14,30,37], ethical issues regarding the relationship between researchers and participants [14,15,32,38,39], and the transferability of the study findings to a wider population [14,30,32,33,37,39] (Table 3). Among all the experimental studies, allocation concealment and study blinding were not adequately reported.

    Table 3. Quality assessment of included studies.
    View this table

    Description of Study Methodologies

    The study design varied among the included studies (Table 4): nine used experimental designs, including four RCTs [29-35], and seven used factorial designs [14,15,32,33,36,37], two were descriptive cross-sectional survey studies [26,39], two were mixed method studies [18,30], one was a pre-post pilot study [28], and one was a secondary analysis [38]. Only one study [34] used an intention-to-treat analysis. Furthermore, 6 studies were informed by a theoretical framework, including the social learning theory [34,38], empowerment construct [26], social support features and gamification elements [27], theory of planned behavior [28], and a self-developed taxonomy of patient stories that provides a framework [31].

    Table 4. Description of study methodologies.
    View this table

    Outcomes of Studies

    Table 5 describes the effect directions of the outcomes of the included studies. The outcomes are grouped along the taxonomy from Shaffer and Zikmund-Fisher [3]. Most studies reported more than one outcome. Therefore, the number of outcomes is higher than the number of included studies.

    Table 5. Effects of narratives on outcomes taxonomy.
    View this table
    Provide Information

    Giesler et al [29] investigated patient competence, including self-regulation, effective coping with emotional distress, explicit dealing with cancer threat, and low avoidance, as a secondary outcome in their study. They reported no significant differences between the intervention and control groups. Snow et al [35] examined the effect of patient narratives describing their colposcopy on fifth-year medical students’ proficiency in standard examinations. They reported a significantly better performance in the examination compared with the control group that viewed a clinician describing the procedure (odds ratio [OR] 2.7, 95% CI 1.2-6.1; P=.02).

    Allam et al [27] reported no significant improvements in the knowledge of rheumatoid arthritis. It should be noted that the initial level of the control group was significant. A study among cancer patients testing narrative cancer modules on the website krankheitserfahrungen found that 72% (40/56) agreed or strongly agreed that the internet is supportive to understanding what physicians tell them [18].

    Engage

    A study by Aardoom et al [26] reported that the exchange of information, finding recognition, sharing experiences with others, and feeling better informed were the most often reported empowering processes and outcomes. The authors concluded that online sources where individuals can share their experiences are promising strategies for successful electronic health (eHealth) initiatives such as Proud2Bme. A 5-arm parallel RCT found that levels of empowerment changed over time in study groups having access to online social support (beta=2.59; P=.03) or gamified experiences of a website (beta=2.29; P=.05) [27].

    Participants viewing narratives relating how a patient makes her decision were found to spend more time searching for information regarding breast cancer (narrative condition, mean 5.38 min, SD 2.37, vs no narrative condition, mean 4.92 min SD 2.03 [31]; narrative condition, mean 39.88 min, SD 15.62, vs no narrative condition, mean 35.08 min, SD 16.09 [32]). Furthermore, Shaffer et al [31] reported that participants who viewed narratives containing experiences regarding diagnosis, treatment, or complications with early breast cancer treatments showed greater abilities to imagine how it might be to experience these treatments (imagine a mastectomy in the no narrative condition, mean 4.46, SD 1.21, vs imagine a mastectomy in the narrative condition, mean 4.69, SD 1.02, t=1.72; P=.04; imagine a lumpectomy with radiation in the no narrative condition, mean 4.44, SD 1.19, vs imagine a lumpectomy with radiation in the narrative condition, mean 4.72, SD 0.94, t=2.19; P=.01; measured on a 9-point Likert scale).

    Findings showed that learning about other peoples' health-related experiences is relevant and helpful [18,30]. Furthermore, patients’ stories collected by DIPEx are frequently included in interviews or articles for group newsletters, newspaper articles, or media broadcasts by voluntary organizations [39]. Engler et al [18], eg, reported that 76% (43/56) of their participants agreed that it can be helpful to witness the health-related experiences of others. However, some of the younger participants in the study by Newman et al [30] reported that the site did not cover experiences of younger patients. The participants highlighted the importance of incorporating current and accurate information. Some participants were concerned that the site might be depressing to patients with a new diagnosis [30].

    Model Target Behaviors

    A statistically significant positive effect on physical activity was reported by Allam et al [27]. In contrast, Schweier et al [34] did not find significant effects on physical activity and eating behavior changes. Health care utilization and medication overuse decreased according to the findings of one study [27]. Furthermore, one study investigated the effects of Web-based narratives and didactic information on health care participation [38]. This study reported positive effects of an eHealth program with narratives (audiovisual and text; beta=.123; P<.01) and didactic information (text only; beta=.104; P<.05) on health care participation. Health care participation was measured on a 7-item, 5-point response scale. These effects were reported to be significantly greater for African Americans.

    Persuade

    A total of six studies investigated the effects of narratives on risk judgments [14,36] and treatment decisions, including hypothetical treatment choices between a lumpectomy with radiation or a mastectomy [33], vaccination intentions [14,15], hypothetical dialysis modalities [37], and cessation treatment [28].

    Furthermore, two studies [14,15] focused on the effects of statistical and/or narrative information on vaccination decisions. Betsch et al [14] showed that the perceived risk of vaccination increases the more the narratives report adverse events (F2,58=3.852; P<.05; η2=0.12), and if adverse events are reported in a highly emotional manner (mean 15.33, SD 9.27 vs mean 17.52, SD 11.00; F1,297=4.197; P<.05; η2=0.01). Furthermore, they showed that the intention for vaccination decreases when the number of narratives increases (F2,58=5.729; P<.01; η2=0.17), which is partially mediated by an increased perception of risk [14]. Two years later, the same research group published results from a similar setting, which point in the same direction [15]. Sullivan et al [36] investigated the influence of videos on consumers’ knowledge, perceptions, and behavioral intentions. Participants were randomly assigned to 1 of 10 fictitious prescription drug websites. The video type (patient testimonial, informational video describing the mechanism of action, or none) and whether the video included drug risks was manipulated on each website. They found that participants who were exposed to any of the videos were less likely to recognize drug risks that were presented only on the website text. Videos that included risk information overall led to increased risk recognition. However, in some risk recognition measures such as risk of fracture, risk of special liver tests, or risk of angioedema, risk recognition decreased for risks that were not presented in the videos but risk information was always present in the website text. Furthermore, the study found no significant effects of risk prominence and type of video condition on physician interaction and search intentions on the internet.

    In addition, one pilot study investigated the effect of a website that aimed to engage smokers in a cessation treatment [28]. Among 38 participants who used the website, 18 participants (47%) became abstinent for at least one day, 7 (18.4%) became abstinent for 7 or more days, and 4 (11%) became continuously abstinent. Winterbottom et al [37] demonstrated that hypothetical dialysis treatment choices presented as a patient narrative were more likely to be chosen by the participants than presented by a doctor (both using actors). Another study [33] found no differences in preferences for surgical treatments between women who watched videos that included narratives compared with those who watched a control video.

    Provide Comfort

    Snow et al [35] reported that students in the narrative condition reported significantly more confidence in comfort with patients’ emotions (OR 6.4, 95% CI 2.7-14.9; P<.001). The study by Shaffer et al [31] compared participants who viewed experience narratives with those not viewing experience narratives. They demonstrated increased confidence in the experience narratives condition regarding the ability to make an informed choice (mean 3.77, SD 0.90 vs mean 4.01, SD 0.84; t=2.33; P<.01), to be more thorough in considering relevant factors (mean 4.07, SD 0.73 vs mean 4.21, SD 0.64; t=1.72; P<.04), to be more confident in the awareness of relevant factors (mean 3.29, SD 0.95 vs mean 3.53, SD 0.90; t=2.21; P<.01), and to be more satisfied with their decision-making process (mean 3.76, SD 0.81 vs mean 3.95, SD 0.77; t=2.08; P<.02).

    Giesler et al [29] evaluated the colorectal cancer module of the German DIPEx website with regard to coping self-efficacy as the primary outcome and patient competencies as the secondary outcome. The study results did not support the authors' hypothesis that the website increases self-efficacy for coping with cancer or patient competencies such as self-regulation or managing emotional distress at 2 and 6 weeks after baseline.


    Discussion

    Principal Findings

    There is an increasing number of Web-based sources containing research-based, systematically generated accounts of patient illness and health experiences. Although the evidence on the persuasiveness of narrative information on individuals’ decision making was reviewed over a decade ago [4], we present, to our knowledge, the first systematic review about the effects of Web-based patient narratives on patients, relatives, or health care professionals.

    Our review revealed several beneficial effects for patients and health care professionals. Web-based narratives are an effective way of teaching to improve knowledge and confidence for students as well as for patients [18,35]. Furthermore, research indicates that patients perceive other patients’ health experiences as relevant and helpful [18,30]. This finding points to the importance of the quality of health-related information [41]. Compared with the health-related information and experiences on general social media sites, academic research–based patient narratives might be less susceptible to challenges for the quality of health-related information through, eg, spamming, intentional misspelling, or actuality of information [41]. Several quality measures to evaluate the quality of Web-based health information are available [42,43].

    Another identified benefit is that participants viewing narratives that contain information on how patients make decisions result in longer search times for information [31,32]. This effect can be a resource to increase, eg, patients’ health literacy. However, Shaffer et al [32] also reported that transcripts of the patient videos caused the opposite effect. Participants confronted with text-based narratives spent approximately five fewer minutes for information search. Researchers and health care professionals using patient narrative databases should be aware that the format of patient stories might be similarly important as the content in determining their effect on medical decision making [32].

    On the basis of the findings of this review, it remains unclear whether patient narratives can influence patients’ target behavior. The results regarding physical activity are equivocal [27,34]. Narratives led to an increase of health care participation and decreased unnecessary health care utilization as well as medication overuse [27,38].

    Even though we identified several benefits of patient narratives on the different purposes of narratives, overall, there is little evidence for the effects of Web-based patient narratives in a positive or negative way. The total number of studies we included in the review is small, which is especially challenging in the light of the heterogeneity regarding the sampled population, the study aims, and the heterogeneity of the narratives in itself. Furthermore, the purposes of the narratives presented on the different websites vary considerably from each other. However, patient narratives are not homogenous and have to be evaluated in their context with regard to content, purpose, and patients’ evaluative expression, such as expressions of (dis)satisfaction with processes of health care decision making [3]. We concur with the position of Shaffer and Zikmund-Fisher [3] that the role of narratives can only be fully understood if operational definitions of narratives are sufficient. Furthermore, there is a need for more theoretical conceptions about the impact of narratives on specific outcomes. We found that only 6 out of 17 studies were guided by a theoretical framework [31,32,34,37-39]. The lack of theoretical frameworks might partially be explained by the fact that research on narratives on certain patient-relevant outcomes is a relatively new field with a range of potentially relevant outcomes. Giesler et al [29], eg, found no significant differences at 2 weeks between an intervention group that had immediate access to the colorectal cancer module of the DIPEx website and a control group with regard to self-efficacy for coping with cancer and patient competence. Study participants in the intervention group visited the website on average for 42.21 min with 3.31 mean number of sessions. Such findings do not necessarily indicate that there is no effect of patient narratives. It rather highlights that the specific outcomes that were investigated in this study and in the study-specific setting were not significant. Other psychological measures on self-efficacy for coping or on patient competence may have led to different results. Another indication for the overall little evidence is that the effects of several studies reporting significant outcomes are rather small or are only significant under specific experimental conditions. For example, the difference between the mean search time for information regarding breast cancer reported by Shaffer et al [32] is 0.46 min, with an average search time of 5.38 min in the intervention group and 4.92 min in the control group.

    Almost one-third of the included studies used study samples that cast doubts whether the results can be generalized to broader patient populations [14,30,32,37], Clear definitions of the basic population and appropriate sampling strategies would be desirable in future studies. Schlesinger et al [44] demonstrated that a rigorous collection of patient narratives can also be incorporated into large patient experience surveys.

    At the same time, narratives can bear potential risks in preference-sensitive decisions [24]. There is a growing body of evidence on the effect of narrative bias [4,14,15], where narrative information can override risk judgments. This effect can even occur when base rate information is presented in addition [15]. Narratives are widely used in patient decision aids [45]. Furthermore, it is likely that narratives are used by other patients as decision support tools, although they are not explicitly declared as decision aids. Decision aids are evidence-based tools with an aim to support patients in a value-sensitive way to make specific health care choices [46]. Narratives may reduce the effectiveness of decision aids by presenting unbalanced information or by overriding decision-relevant information through characteristics of the narrator [4,45]. For example, a study conducted by Khangura et al [45] indicates that patient narratives in decision aids were more likely to portray patients that were satisfied with the outcome of their treatment decision. This points to the importance of including disclaimers that highlight the potential for biases in patient narratives [15]. Furthermore, this highlights the need for a careful selection of the presented stories on patient narrative databases by the corresponding research teams in charge for the databases to present a balanced picture of the whole spectrum of health experiences [47]. This might be especially important in narratives about health conditions where public opinions are mixed and biases might be suspected.

    Qualitative studies focusing on how individuals use and value personal health-related experiences [10], decision making regarding prostate cancer [11], or information needs of patients with cancer and their views of internet-based health information [12] indicate improvements in decision making [10,11] and in meeting information needs [12]. These findings are not completely in line with our review of quantitative studies. How can this difference be explained? Both approaches study different phenomena. The foundations of the qualitative paradigm are interpretivism and constructivism, where multiple socially constructed realities are investigated [48,49]. On the contrary, the quantitative approach is based on positivism, which assumes that phenomena can be represented by empirical indicators that represent the one and only truth [49]. It can be speculated that the qualitative findings rather represent the lived experiences of patients’ decision making, whereas the quantitative results represent quantitative measures of the decision-making processes.

    Limitations

    Our study has several limitations. First, we searched only for papers published in journals, and only in English or German. Papers that were published in books or reports are often not indexed in the databases we have chosen for our search strategy and are therefore not included. Therefore, we may have missed some studies published in languages or places outside our scope. Second, we reviewed only published studies regarding patient narratives. Therefore, we may potentially be confronted with a publication bias in such a way that, eg, negative study results were not published. Third, we decided to include only studies that focused on Web-based narratives and that were generated through a research methodology. Although we have done so to ensure comparability among the studies, we also acknowledge that this decision has led to an exclusion of several studies that investigated the effects of non–Web-based narratives or generated in an unstructured, non–research-based way, eg, in chatrooms or fora. Narratives are valuable resources for the narrators themselves, for other patients and their relatives, and for health care professionals and researchers. Despite the limitations, our findings might be helpful for health care professionals and researchers to understand the possible effects of narratives in health care settings.

    Conclusions

    In total, we found 17 studies on the effects of Web-based patient narratives. The effects of narratives were classified by purpose—inform, engage, model behavior, persuade, and comfort—using the taxonomy provided by Shaffer and Zikmund-Fisher [3]. Overall, patient narratives seem to be a promising means to improve knowledge of health care professionals and patients. Learning about other patients’ experiences is perceived as supportive and relevant. Furthermore, they can positively influence patient empowerment. There is some evidence of beneficial effects on some outcomes, such as information search and the modeling of target behavior such as physical activity, health care participation, and medication overuse. The narratives used in the studies are characterized by considerable heterogeneity, and the investigated outcomes are hardly comparable among each other, which makes an overall judgment difficult.

    Acknowledgments

    The authors would like to thank Kevin Selby, Erik von Elm, and Nina Streeck for helpful comments on the setting up and drafting of this review.

    Conflicts of Interest

    All the authors declare that their institution is in charge of DIPEx Switzerland.

    Multimedia Appendix 1

    Search terms.

    DOCX File , 16 KB

    Multimedia Appendix 2

    Investigated databases and descriptions.

    DOCX File , 14 KB

    References

    1. Kruk ME, Gage AD, Arsenault C, Jordan K, Leslie HH, Roder-DeWan S, et al. High-quality health systems in the Sustainable Development Goals era: time for a revolution. Lancet Glob Health 2018 Nov;6(11):e1196-e1252 [FREE Full text] [CrossRef] [Medline]
    2. World Health Organization. European Health Report 2018: More Than Numbers - Evidence for All. Copenhagen, Denmark: World Health Organization. Regional Office for Europe; 2018.
    3. Shaffer VA, Zikmund-Fisher BJ. All stories are not alike: a purpose-, content-, and valence-based taxonomy of patient narratives in decision aids. Med Decis Making 2013 Jan;33(1):4-13. [CrossRef] [Medline]
    4. Winterbottom A, Bekker HL, Conner M, Mooney A. Does narrative information bias individual's decision making? A systematic review. Soc Sci Med 2008 Dec;67(12):2079-2088. [CrossRef] [Medline]
    5. Ziebland S, Lavie-Ajayi M, Lucius-Hoene G. The role of the internet for people with chronic pain: examples from the DIPEx International Project. Br J Pain 2015 Feb;9(1):62-64 [FREE Full text] [CrossRef] [Medline]
    6. Ziebland S, Herxheimer A. How patients' experiences contribute to decision making: illustrations from DIPEx (personal experiences of health and illness). J Nurs Manag 2008 May;16(4):433-439. [CrossRef] [Medline]
    7. Ziebland S, Chapple A, Dumelow C, Evans J, Prinjha S, Rozmovits L. How the internet affects patients' experience of cancer: a qualitative study. Br Med J 2004 Mar 6;328(7439):564 [FREE Full text] [CrossRef] [Medline]
    8. Greenhalgh T. Cultural Contexts of Health: The Use of Narrative Research in the Health Sector. Copenhagen: World Health Organization Regional Office for Europe; 2016.
    9. Burkow TM, Vognild LK, Østengen G, Johnsen E, Risberg MJ, Bratvold A, et al. Internet-enabled pulmonary rehabilitation and diabetes education in group settings at home: a preliminary study of patient acceptability. BMC Med Inform Decis Mak 2013 Mar 5;13:33 [FREE Full text] [CrossRef] [Medline]
    10. Entwistle VA, France EF, Wyke S, Jepson R, Hunt K, Ziebland S, et al. How information about other people's personal experiences can help with healthcare decision-making: a qualitative study. Patient Educ Couns 2011 Dec;85(3):e291-e298. [CrossRef] [Medline]
    11. Chapple A, Ziebland S, Herxheimer A, McPherson A, Shepperd S, Miller R. Is 'watchful waiting' a real choice for men with prostate cancer? A qualitative study. BJU Int 2002 Aug;90(3):257-264 [FREE Full text] [CrossRef] [Medline]
    12. Rozmovits L, Ziebland S. What do patients with prostate or breast cancer want from an Internet site? A qualitative study of information needs. Patient Educ Couns 2004 Apr;53(1):57-64. [CrossRef] [Medline]
    13. Evans DG, Barwell J, Eccles DM, Collins A, Izatt L, Jacobs C, FH02 Study Group, RGC teams, et al. The Angelina Jolie effect: how high celebrity profile can have a major impact on provision of cancer related services. Breast Cancer Res 2014 Sep 19;16(5):442. [CrossRef] [Medline]
    14. Betsch C, Ulshöfer C, Renkewitz F, Betsch T. The influence of narrative v statistical information on perceiving vaccination risks. Med Decis Making 2011;31(5):742-753. [CrossRef] [Medline]
    15. Betsch C, Renkewitz F, Haase N. Effect of narrative reports about vaccine adverse events and bias-awareness disclaimers on vaccine decisions: a simulation of an online patient social network. Med Decis Making 2013 Jan;33(1):14-25. [CrossRef] [Medline]
    16. Ziebland S, Wyke S. Health and illness in a connected world: how might sharing experiences on the internet affect people's health? Milbank Q 2012 Jun;90(2):219-249 [FREE Full text] [CrossRef] [Medline]
    17. Schaffer R, Kuczynski K, Skinner D. Producing genetic knowledge and citizenship through the internet: mothers, pediatric genetics, and cybermedicine. Sociol Health Illn 2008 Jan;30(1):145-159 [FREE Full text] [CrossRef] [Medline]
    18. Engler J, Adami S, Adam Y, Keller B, Repke T, Fügemann H, et al. Using others' experiences. Cancer patients' expectations and navigation of a website providing narratives on prostate, breast and colorectal cancer. Patient Educ Couns 2016 Aug;99(8):1325-1332. [CrossRef] [Medline]
    19. Breuning M, Lucius-Hoene G, Burbaum C, Himmel W, Bengel J. [Patient experiences and patient centeredness : The website project DIPEx Germany]. Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz 2017 Apr;60(4):453-461. [CrossRef] [Medline]
    20. Greenhalgh T, Wengraf T. Collecting stories: is it research? Is it good research? Preliminary guidance based on a Delphi study. Med Educ 2008 Mar;42(3):242-247. [CrossRef] [Medline]
    21. Moher D, Liberati A, Tetzlaff J, Altman DG, PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med 2009 Jul 21;6(7):e1000097 [FREE Full text] [CrossRef] [Medline]
    22. Hawker S, Payne S, Kerr C, Hardey M, Powell J. Appraising the evidence: reviewing disparate data systematically. Qual Health Res 2002 Nov;12(9):1284-1299. [CrossRef] [Medline]
    23. Brouwers MC, Kho ME, Browman GP, Burgers JS, Cluzeau F, Feder G, AGREE Next Steps Consortium. AGREE II: advancing guideline development, reporting and evaluation in health care. Can Med Assoc J 2010 Dec 14;182(18):E839-E842 [FREE Full text] [CrossRef] [Medline]
    24. Wennberg JE, Fisher ES, Skinner JS. Geography and the debate over Medicare reform. Health Aff (Millwood) 2002;21(Suppl1):W96-114. [CrossRef] [Medline]
    25. O'Connor AM, Légaré F, Stacey D. Risk communication in practice: the contribution of decision aids. Br Med J 2003 Sep 27;327(7417):736-740 [FREE Full text] [CrossRef] [Medline]
    26. Aardoom JJ, Dingemans AE, Boogaard LH, van Furth EF. Internet and patient empowerment in individuals with symptoms of an eating disorder: a cross-sectional investigation of a pro-recovery focused e-community. Eat Behav 2014 Aug;15(3):350-356. [CrossRef] [Medline]
    27. Allam A, Kostova Z, Nakamoto K, Schulz PJ. The effect of social support features and gamification on a Web-based intervention for rheumatoid arthritis patients: randomized controlled trial. J Med Internet Res 2015 Jan 9;17(1):e14 [FREE Full text] [CrossRef] [Medline]
    28. Brunette MF, Gunn W, Alvarez H, Finn PC, Geiger P, Ferron JC, et al. A pre-post pilot study of a brief, web-based intervention to engage disadvantaged smokers into cessation treatment. Addict Sci Clin Pract 2015 Feb 1;10:3 [FREE Full text] [CrossRef] [Medline]
    29. Giesler JM, Keller B, Repke T, Leonhart R, Weis J, Muckelbauer R, et al. Effect of a website that presents patients' experiences on self-efficacy and patient competence of colorectal cancer patients: web-based randomized controlled trial. J Med Internet Res 2017 Oct 13;19(10):e334 [FREE Full text] [CrossRef] [Medline]
    30. Newman MA, Ziebland S, Barker KL. Patients' views of a multimedia resource featuring experiences of rheumatoid arthritis: pilot evaluation of www.healthtalkonline.org. Health Informatics J 2009 Jun;15(2):147-159. [CrossRef] [Medline]
    31. Shaffer VA, Hulsey L, Zikmund-Fisher BJ. The effects of process-focused versus experience-focused narratives in a breast cancer treatment decision task. Patient Educ Couns 2013 Nov;93(2):255-264. [CrossRef] [Medline]
    32. Shaffer VA, Owens J, Zikmund-Fisher BJ. The effect of patient narratives on information search in a web-based breast cancer decision aid: an eye-tracking study. J Med Internet Res 2013 Dec 17;15(12):e273 [FREE Full text] [CrossRef] [Medline]
    33. Shaffer VA, Tomek S, Hulsey L. The effect of narrative information in a publicly available patient decision aid for early-stage breast cancer. Health Commun 2014;29(1):64-73. [CrossRef] [Medline]
    34. Schweier R, Romppel M, Richter C, Hoberg E, Hahmann H, Scherwinski I, et al. A web-based peer-modeling intervention aimed at lifestyle changes in patients with coronary heart disease and chronic back pain: sequential controlled trial. J Med Internet Res 2014 Jul 23;16(7):e177 [FREE Full text] [CrossRef] [Medline]
    35. Snow R, Crocker J, Talbot K, Moore J, Salisbury H. Does hearing the patient perspective improve consultation skills in examinations? An exploratory randomized controlled trial in medical undergraduate education. Med Teach 2016 Dec;38(12):1229-1235. [CrossRef] [Medline]
    36. Sullivan HW, O'Donoghue AC, Gard Read J, Amoozegar JB, Aikin KJ, Rupert DK. Testimonials and informational videos on branded prescription drug websites: experimental study to assess influence on consumer knowledge and perceptions. J Med Internet Res 2018 Jan 23;20(1):e13 [FREE Full text] [CrossRef] [Medline]
    37. Winterbottom AE, Bekker HL, Conner M, Mooney AF. Patient stories about their dialysis experience biases others' choices regardless of doctor's advice: an experimental study. Nephrol Dial Transplant 2012 Jan;27(1):325-331. [CrossRef] [Medline]
    38. Wise M, Han JY, Shaw B, McTavish F, Gustafson DH. Effects of using online narrative and didactic information on healthcare participation for breast cancer patients. Patient Educ Couns 2008 Mar;70(3):348-356 [FREE Full text] [CrossRef] [Medline]
    39. Yaphe J, Rigge M, Herxheimer A, McPherson A, Miller R, Shepperd S, et al. The use of patients' stories by self-help groups: a survey of voluntary organizations in the UK on the register of the College of Health. Health Expect 2000 Sep;3(3):176-181 [FREE Full text] [CrossRef] [Medline]
    40. Sullivan J, Gillam L, Monagle P. Parents and end-of-life decision-making for their child: roles and responsibilities. BMJ Support Palliat Care 2015 Sep;5(3):240-248. [CrossRef] [Medline]
    41. De Martino I, D'Apolito R, McLawhorn AS, Fehring KA, Sculco PK, Gasparini G. Social media for patients: benefits and drawbacks. Curr Rev Musculoskelet Med 2017 Mar;10(1):141-145 [FREE Full text] [CrossRef] [Medline]
    42. Charnock D, Shepperd S, Needham G, Gann R. DISCERN: an instrument for judging the quality of written consumer health information on treatment choices. J Epidemiol Community Health 1999 Feb;53(2):105-111 [FREE Full text] [CrossRef] [Medline]
    43. Boyer C, Selby M, Scherrer JR, Appel RD. The Health On the Net Code of Conduct for medical and health Websites. Comput Biol Med 1998 Sep;28(5):603-610. [CrossRef] [Medline]
    44. Schlesinger M, Grob R, Shaller D, Martino SC, Parker AM, Rybowski L, et al. A rigorous approach to large-scale elicitation and analysis of patient narratives. Med Care Res Rev 2018 Oct 6:1077558718803859. [CrossRef] [Medline]
    45. Khangura S, Bennett C, Stacey D, O'Connor AM. Personal stories in publicly available patient decision aids. Patient Educ Couns 2008 Dec;73(3):456-464. [CrossRef] [Medline]
    46. Stacey D, Légaré F, Lewis K, Barry MJ, Bennett CL, Eden KB, et al. Decision aids for people facing health treatment or screening decisions. Cochrane Database Syst Rev 2017 Apr 12;4:CD001431 [FREE Full text] [CrossRef] [Medline]
    47. Grob R, Schlesinger M, Parker AM, Shaller D, Barre LR, Martino SC, et al. Breaking narrative ground: innovative methods for rigorously eliciting and assessing patient narratives. Health Serv Res 2016 Jun;51(Suppl 2):1248-1272 [FREE Full text] [CrossRef] [Medline]
    48. Berger PL, Luckmann T. The Social Construction of Reality. Frankfurt am Main: Fischer; 2004.
    49. Sale JE, Lohfeld LH, Brazil K. Revisiting the quantitative-qualitative debate: implications for mixed-methods research. Qual Quant 2002 Feb;36(1):43-53 [FREE Full text] [CrossRef] [Medline]


    Abbreviations

    DIPEx: Database of Individual Patients’ Experiences
    eHealth: electronic health
    OR: odds ratio
    RCT: randomized controlled trial


    Edited by G Eysenbach; submitted 05.08.19; peer-reviewed by C Holmberg, H Sullivan; comments to author 14.09.19; revised version received 01.11.19; accepted 16.12.19; published 26.03.20

    ©Daniel Drewniak, Andrea Glässel, Martina Hodel, Nikola Biller-Andorno. Originally published in the Journal of Medical Internet Research (http://www.jmir.org), 26.03.2020.

    This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on http://www.jmir.org/, as well as this copyright and license information must be included.