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 13.04.16 in Vol 18, No 4 (2016): April

Preprints (earlier versions) of this paper are available at http://preprints.jmir.org/preprint/5207, first published Oct 28, 2015.

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

    Original Paper

    Awareness and Use of the After-Visit Summary Through a Patient Portal: Evaluation of Patient Characteristics and an Application of the Theory of Planned Behavior

    1Division of General Internal Medicine, Department of Medicine, Brigham and Women's Hospital, Harvard Medical School, Boston, MA, United States

    2Brigham and Women's Physicians Organization, Brigham and Women's Hospital, Boston, MA, United States

    3Massachusetts General Physicians Organization, Massachusetts General Hospital, Boston, MA, United States

    4Information Services, Partners HealthCare, Boston, MA, United States

    5Department of Healthcare Policy and Management, Harvard School of Public Health, Boston, MA, United States

    Corresponding Author:

    Srinivas Emani, MA, PhD

    Division of General Internal Medicine, Department of Medicine

    Brigham and Women's Hospital

    Harvard Medical School

    1620 Tremont Street, 3rd Floor

    Boston, MA, 02120

    United States

    Phone: 1 617 732 7063

    Fax:1 617 732 7072

    Email: semani1@partners.org


    ABSTRACT

    Background: Patient portals are being used to provide a clinical summary of the office visit or the after-visit summary (AVS) to patients. There has been relatively little research on the characteristics of patients who access the AVS through a patient portal and their beliefs about the AVS.

    Objective: The aim was to (1) assess the characteristics of patients who are aware of and access the AVS through a patient portal and (2) apply the Theory of Planned Behavior (TPB) to predict behavioral intention of patients toward accessing the AVS provided through a patient portal.

    Methods: We developed a survey capturing the components of TPB (beliefs, attitude, perceived norm, and perceived behavioral control). Over a 6-month period, patients with a patient portal account with an office visit in the previous week were identified using our organization’s scheduling system. These patients were sent an email about the study and a link to the survey via their portal account. We applied univariate statistical analysis (Pearson chi-square and 1-way ANOVA) to assess differences among groups (aware/unaware of AVS and accessed/did not access AVS). We reported means and standard deviations to depict belief strengths and presented correlations between beliefs and attitude, perceived norm, and perceived behavioral control. We used hierarchical regression analysis to predict behavioral intention toward accessing the AVS through the patient portal.

    Results: Of the 23,336 patients who were sent the survey, 5370 responded for a response rate of 23.01%. Overall, 76.52% (4109/5370) were aware that the AVS was available through the patient portal and 54.71% of those (2248/4109) accessed the AVS within 5 days of the office visit. Patients who accessed the AVS had a greater number of sessions with the portal (mean 119, SD 221.5) than those who did not access the AVS (mean 79.1, SD 123.3, P<.001); the difference was not significant for awareness of the AVS. The strongest behavioral beliefs with accessing the AVS were being able to track visits and tests (mean 2.53, SD 1.00) followed by having medical information more readily accessible (mean 2.48, SD 1.07). In all, 56.7% of the variance in intention to access the AVS through the portal was accounted for by attitude, perceived norm, and perceived behavioral control.

    Conclusions: Most users of a patient portal were aware that the AVS was accessible through the portal. Patients had stronger beliefs about accessing the AVS with the goal of timely and efficient access of information than with engaging in their health care. Interventions to improve patient access of the AVS can focus on providers promoting patient beliefs about the value of the AVS for tracking tests and visits, and timely and efficient access of information.

    J Med Internet Res 2016;18(4):e77

    doi:10.2196/jmir.5207

    KEYWORDS



    Introduction

    The adoption and use of patient portals tethered to electronic health records (EHRs) has accelerated in the last decade. A primary driver of this growth has been the Medicare and Medicaid EHR Incentive Program, widely referred to as the EHR Meaningful Use (MU) program, introduced in the Health Information Technology for Economic and Clinical Health (HITECH) provision of the American Recovery and Reinvestment Act of 2009 [1,2]. The objectives of the MU program are to increase the adoption of EHRs and the meaningful use of EHRs to improve delivery of care, decrease medical errors, improve efficiency of care, and enhance patient centeredness of care [2]. The MU program is being implemented in three stages with the criteria for achieving meaningful use of the EHR becoming more rigorous with each stage. Patient portals are expected to play a key role in the MU program by providing patients with timely and efficient access to information, engaging patients in their care, and enhancing patient centeredness of care [3,4]. One of the core objectives of the MU program is to allow patients to view online and download their health information, such as test results, problem and medication lists, and medication allergies. For example, the Blue Button initiative has been implemented by a number of organizations to allow patients to download a copy of their health information by clicking on a blue circle on the patient portal page [5].

    Patient portals are also being used to provide a clinical summary of the office visit or the after-visit summary (AVS) to patients. The Centers for Medicare and Medicaid Services (CMS) has defined the AVS as a clinical summary that “provides a patient with relevant and actionable information and instructions” such as the provider’s office contact information, date and location of visit, an updated medication list, updated vitals, reason(s) for visit, procedures and other instructions based on clinical discussions that took place during the office visit, any updates to a problem list, immunizations or medications administered during visit, summary of topics covered/considered during visit, and time and location of next appointment/testing, if scheduled [6]. Stage 1 of the MU program specified that the AVS should be provided to patients for more than 50% of all office visits within three business days.

    The AVS requirement was controversial in the physician community and, in spite of the Stage 1 requirement of the MU program for the provision of the AVS, there has been relatively little research on how patients view the AVS. In a survey of the printed version of the AVS provided at an office visit, Neuberger and colleagues [7] reported that 88% of respondents said the information on the AVS was easy to understand and 84% said that the AVS was helpful. Chung and colleagues [8] reported similar results in their survey of a printed version of the AVS: 93% of patients agreed that they understood the information on the AVS and 93% agreed that having the AVS was helpful. Ralston and colleagues [9] reported that the AVS was the fastest growing use of their organization’s patient portal and may reflect the patient’s desire for information about their care plan and needs. Pavlik and colleagues [10] found patient satisfaction with the MU version of the AVS did not differ significantly from other content versions of the AVS. In this study, 30% of the patients reported that they plan to keep the AVS for their next appointment.

    Although these studies provide some understanding of patient opinions about the AVS, we know relatively little about predictors of patient access to the AVS through the patient portal, such as do younger and more highly educated patients access (ie, retrieve) the AVS more through the patient portal compared to older and less educated patients, and what role does patient experience with the patient portal play in their accessing the AVS through the portal? There is also a lack of evidence on patient beliefs and attitude toward accessing the AVS. Do patients believe that accessing the AVS through the patient portal will provide information in a timely manner, allow them to track their visits and tests, and reinforce their provider’s instructions? This study contributes to the sparse literature on patient portals and the AVS by addressing the following objectives: (1) assess the characteristics of patients who are aware of and access (retrieve) the AVS through a patient portal and (2) apply the Theory of Planned Behavior (TPB) to evaluate beliefs, attitude, perceived norm, perceived behavioral control, and predict behavioral intention of patients toward accessing the AVS provided through a patient portal.


    Methods

    Theoretical Model

    In a previous study, we pointed to the lack of application of theoretical models in the study of patient adoption and use of patient portals [11]. In that study, Rogers’ Diffusion of Innovation model was successfully applied to assess patient perceptions of a patient portal. The perceived attributes of ease of use and relative advantage of the portal emerged as significant predictors of portal adoption and value. Another technology adoption model, the Unified Theory of Adoption and Use of Technology, was successfully applied by Turvey and colleagues [12] in their study of use of the Blue Button at the Department of Veterans Affairs patient portal, MyHealtheVet. In that study, factors such as knowledge and usability of the Blue Button emerged as significant barriers to the use of the Blue Button associated with the patient portal. However, we continued to find a lack of application of theoretical models in the study of patient portals. Furthermore, there was a need for theoretical models that could predict behavioral intention and behavior toward the use of patient portals and the specific functionality associated with portals such as the AVS. The application of such models could also yield useful prescriptive implications for practitioners who are interested in improving patient use of patient portals.

    In this study, we applied a prominent theoretical model, the Theory of Planned Behavior (TPB), to patient portals with a focus on predicting patients’ behavioral intention toward accessing the AVS provided through the portal. According to TPB, three major factors lead to the formation of a patient’s intention to perform a behavior: attitude toward the behavior, perceived norm, and perceived behavioral control [13]. Underlying each of these factors is a set of beliefs: behavioral beliefs about the positive or negative consequences of accessing the AVS (determines attitude), normative beliefs that important people would approve or disapprove of the patient accessing the AVS or that these referents themselves are accessing the AVS (determines perceived norm), and control beliefs that facilitate or impede the patient accessing the AVS through the patient portal (determines perceived behavioral control). Behavioral intention is a direct predictor of the behavior of accessing the AVS through the patient portal. However, factors such as skills and environmental factors may influence the relationship between intention and behavior. For example, a patient may lack the computer skills to access the AVS through the portal even if they form an intention to access the AVS. Finally, background factors such as education, age, and race may indirectly influence beliefs about accessing the AVS. Over the last three decades, TPB has been successfully applied to predict intention across a wide range of health and risk behaviors, including exercise, breast self-examination, eating a low-fat diet, condom use, alcohol consumption, smoking, and using drugs [13,14].

    Survey Instrument

    To apply TPB to predict patients’ intention to access the AVS through the patient portal, we developed and implemented a cross-sectional survey that captured the different components of TPB. We followed the approach recommended by Fishbein and Ajzen [13] in developing our survey instrument. First, we conducted a pilot study to gather data on the beliefs related to patients’ accessing the AVS through the patient portal. In TPB, the behavior of interest is defined by four elements through the principle of compatibility: the action performed, the target at which the action is directed, the context in which it is performed, and the time at which it is performed [13]. In our pilot study, we defined the behavior of interest as accessing (action) the AVS (target) through the patient portal (context) within 7 days of the visit (time). Our pilot survey consisted of three questions related to this behavior of interest: (1) asking patients to list the advantages and disadvantages of accessing the AVS through the patient portal within 7 days of the visit to identify behavioral beliefs, (2) asking patients to list individuals or groups who would approve or disapprove of their accessing the AVS through the patient portal to identify injunctive normative beliefs, and (3) asking patients to list the factors or circumstances that would make it easy or difficult for them to access the AVS through the patient portal to identify control beliefs. Based on the responses to the pilot survey, we created the items for our survey. Additionally, we changed the time component for accessing the AVS to within 5 days of the visit. Multimedia Appendix 1 lists the survey items on TPB we developed for our study.

    Recruitment

    The study was implemented in the ambulatory care practices of an academic medical center affiliated with Partners HealthCare, an integrated delivery system located in Eastern Massachusetts. Partners developed its own patient portal, Patient Gateway, following its strategy of developing and implementing its own EHR, the Longitudinal Medical Record. The patient portal has functionality similar to other vendor portals, including requests for appointments, prescription refills and referrals, access to certain components of the EHR (eg, laboratory results), and secure messaging with the practice and provider. The AVS is made available to patients through the portal. Patient portal transactions are stored permanently in the Partners clinical information systems and can be accessed for research purposes after institutional review board approval. Over a 6-month period, patients with a patient portal account and an office visit in the previous week were identified using the Partners scheduling system. These patients were sent an email about the study through their portal account. The email included a link to the survey. After 7 days, patients were sent a reminder email with another link to the survey. Patients were not compensated for the survey. All study materials and methods were approved by the Partners Health Care Institutional Review Board.

    Statistical Analysis

    We present frequencies and means of sociodemographic characteristics and factors related to portal experience for the different groups: aware of AVS / unaware of AVS and accessed AVS / did not access AVS. To assess for differences between the groups, we conducted chi-square tests for categorical data (Pearson chi-square for dichotomous and nominal variables) and robust 1-way ANOVA for continuous variables. We employed multiple regression analysis using a forced entry method to assess predictors of patient satisfaction with the AVS. To test the application of TPB, we computed Cronbach alpha for the major factors (attitude, perceived norm, and perceived behavioral control). We then created scales for each factor using the mean of the scores of the items for each scale. We also computed Cronbach alpha and created a scale for behavioral intention to access the AVS through the patient portal. We present means and standard deviations for the belief items captured through our survey and correlations of the belief items with respective factors. Finally, we conducted hierarchical regression analysis to predict behavioral intention from the major factors of TPB as well as external factors such as sociodemographics and portal experience.


    Results

    Response Rate

    Of the 23,336 patients who received the online survey, 5370 responded for a response rate of 23.01%. Table 1 shows the characteristics of the responders and nonresponders. Overall, 61.79% (3318/5370) of responders and 62.51% (11,231/17,966) of nonresponders were female; 90.76% (4874/5370) of responders were white compared to 86.44% (15,530/17,966) of nonresponders (P<.001). Responders were older (mean 56.6, SD 14.0 years) than nonresponders (mean 50.4, SD 15.4 years, P<.001). Responders also had a portal account for a longer time (mean 3.5, SD 3.3 years) compared to nonresponders (mean 2.8, SD 2.9 years) and sent a greater number of messages (mean 7.7, SD 20.1) compared to nonresponders (mean 4.3, SD 13.8). Given the large sample sizes of responders and nonresponders, the statistically significant differences should be viewed with caution.

    Table 1. Characteristics of responders and nonresponders (N=23,336).
    View this table

    Awareness and Access of the After-Visit Summary

    Among the 5370 responders of the survey, 4109 (76.52%) reported that they were aware of the availability of the AVS through the patient portal and 1169 (21.77%) reported that they were not aware of the availability of the AVS through the portal (92/5370, 1.71% did not respond to this question). Table 2 shows characteristics of patients who were aware and unaware of the AVS; 61.35% (2521/4109) of female patients were aware of the AVS and 63.99% (748/1169) of female patients were unaware of the AVS. Patients who were aware of the AVS had a mean age of 56.4 (SD 14.0) years compared to a mean age of 57.1 (SD 13.7) years for patients who were unaware of the AVS. In all, 90.58% (3722/4109) and 91.79% (1073/1169) of patients were white among patients who were aware and unaware of the AVS, respectively. In all, 67.78% (2452/3618) of patients who were aware of the AVA and 67.7% (674/995) who were unaware of the AVS reported their marital status as married or an unmarried couple; 51.11% (1873/3655) of patients who were aware of the AVS and 50.65% (508/1003) who were unaware of the AVS reported their health status as very good or excellent. The proportion of patients who reported that they were a 4-year college graduate or more was lower in the aware group (67.54%, 2461/3644) compared to the unaware group (79.42%, 795/1001, P<.001). The proportion of patients who reported total household income from all sources as US $75,000 or more was also lower in the aware group (62.55%, 2036/3255) compared to the unaware group (66.4%, 570/858, P=.04). Patients who were aware of the AVS had a portal account for a mean of 3.4 (SD 3.3) years compared to a mean of 4.0 (SD 3.4) years for patients who were not aware of the AVS (P<.001). The two groups did not differ on the number of sessions with the portal or the number of messages sent via the portal. Satisfaction with the portal was significantly higher in the AVS aware group (45.28%, 164/3653 reporting excellent) compared to the unaware group (30.1%, 299/993, P<.001).

    Table 2. Characteristics of respondents who were aware/unaware of the after-visit summary (AVS).
    View this table

    Of the 4109 patients who were aware of the AVS, 2248 (54.71%) reported that they accessed the AVS through the patient portal within 5 days of the visit and 1805 (43.93%) did not access the AVS (56/4109, 1.4% did not respond). The top two reasons provided for not accessing the AVS through the portal were did not have a need for the AVS (45.43%, 820/1805) and did not remember that AVS was available through the portal (31.63%, 571/1805). Another 14.68% (265/1805) reported that they had received a copy of the AVS from their doctor’s office. Only 3.49% (63/1805) reported that they did not know how to access the AVS through the patient portal.

    In total, 61.30% (1378/2248) of patients who accessed the AVS through the patient portal were female and 61.61% (1112/1805) of patients who did not access the AVS through the portal were female (Table 3). Mean age of patients who accessed the AVS was 56.5 (SD 14.1) years and mean age was 56.3 (SD 13.8) years for patients who did not access the AVS. In all, 90.52% (2035/2248) of patients who accessed the AVS and 90.47% (1633/1805) of patients who did not access the AVS were white. Of patients who accessed the AVS, 66.08% (1325/2005) had a 4-year college degree or more compared to 69.22% (1102/1592) of patients who did not access the AVS. Also, 62.85% (1132/1801) of patients who accessed the AVS and 62.18% (878/1412) of patients who did not access the AVS had a total household income of US $75,000 or more. Patients who accessed the AVS had a greater proportion reporting a status as a married or unmarried couple (69.69%, 1391/1996) compared to patients who did not access the AVS (65.44%, 1032/1577, P=.007). Patients reporting health status of very good or excellent was similar in both groups (50.59%, 1022/202 in access group and 51.75%, 827/1598 in did not access group). Patients who accessed the AVS had a portal account for a greater number of years (mean 3.6, SD 3.3) compared to patients who did not access the AVS (mean 3.1, SD 3.1, P<.001). Patients who accessed the AVS also used the portal more than patients who did not access the AVS (mean 119, SD 221.5 sessions vs mean 79.1, SD 123.3 sessions, P<.001; mean 8.2, SD 20.8 messages sent via the portal vs mean of 6.6, SD 18.2 messages sent via the portal, P=.01). Patients who accessed the AVS reported greater satisfaction with the portal (excellent: 51.03%, 1032/2022) compared to patients who did not access the AVS (excellent: 38.00%, 602/1584, P<.001).

    Table 3. Characteristics of respondents who accessed or did not access the after-visit summary (AVS).a
    View this table

    Satisfaction With After-Visit Summary

    We assessed patient satisfaction with the AVS by asking patients to rate the AVS on a scale from 1 (poor) to 5 (excellent). The mean satisfaction with the AVS was 3.9 (SD 1.12). Patients 65 years and older reported greater satisfaction with the AVS (mean 3.9, SD 1.1) than patients younger than 65 years of age (mean 3.8, SD 1.1, P=.02). Nonwhite patients reported greater satisfaction (mean 4.0, SD 1.0) than white patients did (mean 3.9, SD 1.1, P=.04). Patients who reported very good or excellent health status reported greater satisfaction with the AVS (mean 4.0, SD 1.1) than patients who reported other (poor/fair/good) health status (mean 3.8, SD 1.2, P=.02). Patients who had less than a 4-year college degree reported greater satisfaction (mean 4.1, SD 1.0) than patients who had a 4-year college degree or more (mean 3.8, SD 1.2, P<.001). We fitted a multiple regression (forced entry) with satisfaction with AVS as the dependent variable and sociodemographics and portal-related variables as predictors (Table 4). Satisfaction with the patient portal was the most significant predictor of satisfaction with the AVS (beta=.679, P<.001). The number of portal sessions was also a significant predictor of the AVS (beta=–.095, P<.001). Among sociodemographic variables, age, gender, race, and education were significant predictors of satisfaction with AVS. Finally, patient-reported health status was a significant predictor of satisfaction with AVS.

    The Theory of Planned Behavior

    Components of Theory of Planned Behavior

    Table 5 shows results of reliability analyses for the three main factors in TPB and for the outcome variable of behavioral intention. Cronbach alpha was very good for all the factors. We created scales for each of the factors using a mean of the scores of the items pertaining to each factor.

    Table 4. Multiple regression with satisfaction with after-visit summary as the dependent variable.
    View this table
    Table 5. Reliability analysis for attitude, perceived norm, perceived behavioral control, and behavioral intentions.
    View this table

    Table 6 shows results for behavioral belief strength, outcome evaluation, and the product of behavioral belief strength and outcome evaluation for the behavioral beliefs included in this study. Mean belief strength was on the positive side for all the behavioral beliefs. The most strongly held beliefs were patients’ ability to track visits and tests, and patients’ having medical information more readily accessible. The two beliefs that were lowest in strength were clarifying issues with their doctor and reinforcing doctor’s instructions. These two beliefs also had the lowest outcome evaluation and the greatest variation in belief strength and outcome evaluation. The mean behavioral belief strength×outcome evaluation products show that the three beliefs with the strongest positive impact on attitude were being able to track visits and tests, having medical information more readily accessible, and a more efficient way to obtain medical information. The correlations between behavioral belief strength×outcome evaluation and attitude were positive and significant and ranged from .45 to .52.

    Table 7 shows results for normative belief strengths, motivation to comply, and the product of normative belief strengths and motivation to comply for the three referents included in the study. Mean normative belief strength was positive for all three referents. The strongest belief strength was associated with the patients’ doctor. Motivation to comply was also positive with the strongest motivation to comply associated with the patient’s doctor. Correlations between normative belief strength×motivation to comply and injunctive norm were all positive and significant.

    Table 8 shows results for control beliefs, power of factor, and the product of control beliefs and power of factor. The strongest control belief was that the patient will have access to the Internet within 5 days of the visit. This was followed by the control belief “it will be easy for me to access the AVS via the patient portal” and the belief “I will remember the user ID and password for the patient portal.” Correlations between the product of control belief and power of factor and perceived behavioral control were all positive and significant. The strongest correlation was for the control belief “it will be easy for me to access the AVS via the patient portal.”

    Table 6. Behavioral beliefs, outcome evaluation, and correlations with attitude.
    View this table
    Table 7. Injunctive normative beliefs, motivation to comply, and correlations with injunctive norm.
    View this table
    Table 8. Control beliefs, power of factors, and correlations with perceived behavioral control.
    View this table

    Predictors of Behavioral Intention

    We employed hierarchical multiple regression to assess predictors of behavioral intention with respect to accessing the AVS through the patient portal and to test the TPB model (Table 9). Given the TPB model, we entered attitude, perceived norm, and perceived behavioral control in the first model. In the second and third models, we entered variables external to the TPB model: variables related to patient portal use were entered in the second model and sociodemographic variables and health status were entered in the third model. The first model consisting of the direct predictors of behavioral intention, attitude, perceived norm, and perceived behavioral control was significant and accounted for 56.7% of the variance in behavioral intention. The second model with portal variables was also significant but added only 5%. Years with portal account, number of sessions, and satisfaction with the AVS were significant predictors in this model. The third model was also significant and added another 3%. Age was the only sociodemographic variable that was significant in this model (beta=.049, P<.001).

    Table 9. Hierarchical multiple regression analysis predicting behavioral intention.
    View this table

    Discussion

    In this study, we assessed patient awareness and access of the AVS provided through a patient portal. A large majority of users of the portal were aware that the AVS was available through the portal. Of those who were aware, just over half reported that they accessed the AVS through the portal within 5 days of an office visit. There were no differences between the groups with respect to sociodemographics (eg, age, gender, race, and self-reported health status). Education and income were related to awareness of the AVS but in a reverse direction than expected: users of the portal with more education and higher income were more likely to be unaware of the availability of the AVS through the portal. However, these differences did not carry over to the access of the AVS through the portal. Given these findings on race, income, and education, our study does not find a digital divide in the case of both awareness and access of the AVS through the patient portal. In a previous study on the digital divide associated with the adoption and use of the patient portal, we found that the digital divide did not carry over to portal use, specifically the relationship between income and the frequency of secure messaging through the portal [15]. In this study, accessing the AVS through the patient portal yields a similar finding: once patients have adopted the portal and are using it, issues of digital divide may not persist at least with respect to some of the functionality of the portal. However, there is a need for additional research on the digital divide with respect to use of portal functionality.

    Although previous studies have reported high levels of satisfaction overall with patient portals, there has been little research on satisfaction with the use of specific portal functionality such as the AVS. Our study found a high level of satisfaction with the AVS similar to the finding by Pavlik and colleagues [10]. At the same time, overall satisfaction with the patient portal was positively and significantly related to both awareness and access of the AVS through the patient portal. Satisfaction with the portal was also the most important predictor of satisfaction with the AVS. We agree with the need for empirically measuring a quality indicator such as patient satisfaction with the portal because it appears to be a driver of patient satisfaction with specific functionality of the portal such as the AVS [4].

    We do not know of other work that has previously applied TPB to assess patients’ beliefs and predict behavioral intention toward accessing their AVS through a patient portal and our findings support the use of the theoretical model in this area. The correlations between beliefs about accessing the AVS through the patient portal and the major determinants of the TPB model (attitude, perceived norm, and perceived behavioral control) were positive and significant, and similar to those reported in studies on other health behaviors. Ajzen [16] reported mean correlations between the expectancy-value index of beliefs and a direct attitude measure ranging from .50 to .53 based on the findings of two meta-analyses of studies applying TPB to health behaviors. The correlations in our study (Table 5) ranged from .45 to .50 with a mean of .48. McEachan and colleagues [14] conducted a meta-analysis of 237 prospective tests of the TPB applied to health behaviors. They reported a corrected (for sampling and measurement error) mean correlation of .57 between attitude and intention, a mean correlation of .54 between perceived behavioral control and intention, and a mean correlation of .40 between subjective norm and intention. In our study, correlation between attitude and intention was .65, perceived behavioral control and intention was .47, and perceived norm and intention was .58. The prediction of behavioral intention to access the AVS through the patient portal from attitude, perceived norm, and perceived behavioral control yielded similar findings to other applications of TPB in the health arena. Ajzen [16] reported a meta-analysis showing the mean multiple correlation between the three major determinants of TPB and behavioral intention to range from .59 and .66. In terms of prediction of intention, McEachan and colleagues [14] reported that 44.3% of the variance in intention was accounted by attitude, subjective norm, and perceived behavioral control. Attitude was the strongest predictor (beta=.35) followed by perceived behavioral control (beta=.34) and subjective norm (beta=.15). In our study, 56.7% of the variance in intention was accounted by attitude, perceived norm, and perceived behavioral control. Attitude was the strongest predictor of intention (beta=.43) followed by perceived norm (beta=.29) and perceived behavioral control (beta=.21).

    In terms of specific beliefs, our study found that behavioral beliefs related to patient access of information through the AVS, specifically the ability to track visits and tests, have medical information more readily accessible, and obtain medical information more efficiently, were more important than beliefs about patient engagement in their health care, such as clarifying issues with their doctor or reinforcing instructions. This finding is similar to the finding obtained in studies on patients accessing their doctor’s notes that patients value access to their information [17-20]. In the VA Open Notes study, patients accessed their doctor’s notes to be better prepared for clinic visits, remember their care plan better, and feel more in control of their health [17,18]. Delbanco and colleagues [19,20] also found patients accessed their doctor’s notes to be better prepared for future visits and have a greater sense of control over their health. In their organization, Ralston and colleagues [9] noted the AVS may serve patient’s information and care needs better because it provides a focused plan of care combined with educational materials hyperlinked to other sources. The question of when patients prefer to access their AVS compared to their doctor’s notes through a patient portal, and the relative value of these uses of the portal, are important topics for future research.

    In the case of normative and control beliefs, the strongest normative belief and motivation to comply were associated with the patient’s doctor. Patients believe that their doctor thinks they should access the AVS and they want to do what their doctor thinks. The importance of clinician encouragement of patients using online tools such as patient portals has been identified in several studies [9,11,21]. Our study finds that clinicians also have an important role in encouraging patients to access specific functionality of portals such as the AVS. In the case of control beliefs, the ease of accessing the AVS through the patient portal was identified as an important belief. The importance of ease of access and use of patient portals has been well documented in empirical and scoping studies [11,12,22,23]. Our study finds that even after patients have adopted patient portals, ease of access of specific functionality of the portals can facilitate or impede intention and behavior to use the functionality.

    This study had two objectives: (1) assess the characteristics of patients who are aware of and access the AVS through a patient portal and (2) apply TPB to evaluate beliefs, attitude, perceived norm, perceived behavioral control, and predict behavioral intention of patients toward accessing the AVS through a patient portal. A majority of users of a patient portal selected for this study were aware that the AVS was accessible through the portal, but almost a third did not remember that the AVS was available and therefore did not access it. Patients may need to be reminded that the AVS is accessible through the portal at the time they leave their office visit, especially if patient preference is to receive information through the patient portal. In terms of patient characteristics, we did not find evidence of a digital divide with respect to income or education in either awareness or access of the AVS in our portal users. On the other hand, portal users in one setting who accessed their doctor’s notes were more educated than those who did not access the notes [18]. Additional research is needed on issues of digital divide with respect to different uses of the patient portal, such as accessing the AVS or doctor’s notes.

    With respect to the meaningful use of EHR incentive program, two goals were envisioned related to EHR functionality such as the AVS [1,2]: (1) provide patients with timely and efficient access to their health information and (2) motivate patients to engage in their health care. Other studies have also identified patient engagement in their health care (eg, shared decision making with their doctor) as an important purpose of the AVS [24]. Our study found that patients had stronger beliefs about the AVS with respect to timely and efficient access of information than with engaging in their health care. This finding may reflect patients’ value of the AVS as a permanent personal record to review whenever the need arises [25]. On the other hand, it is possible that the use of the AVS to engage patients in their health care is not being promoted. Pavlik and colleagues [10] noted the need for concerted efforts to remind patients of important information available through the AVS than simply providing the AVS to patients. Such efforts can lead to patient activation and the use of information by patients to undertake recommended treatment plans and self-management, both of which are important goals for the AVS [24]. Although Stage 1 of the MU program included a core requirement for the provision of the AVS, Stage 2 of the MU program no longer includes this core requirement. This is unfortunate given the value of the AVS for providing timely and efficient access of information as reported by patients in our study.

    We found TPB to be a suitable theoretical model to predict behavioral intention of patients toward accessing the AVS through a patient portal. Our findings match applications of TPB for predicting intention with respect to a variety of other health behaviors. Thus, this study provides an important contribution to the application of theoretical models to the study of patient portals and extends some of the prior theoretical work on this topic [11,12]. Beyond its theoretical contribution, the application of TPB can suggest interventions that are relevant to practitioners. For example, Fishbein and Ajzen [13] recommend interventions that target and change relevant salient beliefs or make new beliefs salient in support of recommended behavior. Our study found that the strongest behavioral beliefs related to accessing the AVS through the patient portal are those related to tracking visits and tests, and having medical information more readily and efficiently accessible. Our study also found that doctors are an important social agent for patients with respect to accessing the AVS. For those patients who are not accessing the AVS through the patient portal, a simple intervention that organizations can implement would be to encourage doctors and support staff to discuss with patients the advantages of accessing the AVS through the portal, such as tracking visits and tests. This intervention would also help those patients who do not remember that the AVS is available through the portal. Similarly, doctors can ask patients to use the AVS for clarifying instructions and engaging in shared decision making. Engaging patients in a dialog about the use of the AVS may also help facilitate two important factors identified by us in a survey of physician beliefs about the AVS: (1) enhancing physician satisfaction with the AVS and (2) promoting positive beliefs about the effect of the AVS on patient outcomes and the care the physician personally delivers [26].

    Although this study yielded valuable insights into awareness and access of the AVS through a patient portal and the application of TPB to this area, it is associated with some limitations. The response rate in our study was low. We relied on a self-report of patients accessing the AVS through the patient portal. With respect to the time component, we chose 5 days from the office visit as the period within which the patient accessed the AVS. There is a need to assess other time periods after an office visit in which the patient could access the AVS. The study was conducted in the setting of a Northeast academic medical center and the results may not be generalizable to other regions and patient populations. However, we have no reason to suspect that the beliefs identified in this study and their respective strengths would differ across different institutional settings in which the AVS is available through a patient portal (academic medical center vs other) or across different platforms (homegrown vs vendor patient portals).

    Acknowledgments

    The authors thank the Partners Siemens Research Council (PSRC) for funding the study, and Mohan Ganasekaran, Partners IS, for help with the programming required for the study.

    Conflicts of Interest

    None declared.

    Multimedia Appendix 1

    Survey Items.

    PDF File (Adobe PDF File), 44KB

    References

    1. Blumenthal D, Tavenner M. The "meaningful use" regulation for electronic health records. N Engl J Med 2010 Aug 5;363(6):501-504. [CrossRef] [Medline]
    2. Buntin MB, Jain SH, Blumenthal D. Health information technology: laying the infrastructure for national health reform. Health Aff (Millwood) 2010 Jun;29(6):1214-1219 [FREE Full text] [CrossRef] [Medline]
    3. Kannry J, Beuria P, Wang E, Nissim J. Personal health records: meaningful use, but for whom? Mt Sinai J Med 2012;79(5):593-602. [CrossRef] [Medline]
    4. Kruse CS, Bolton K, Freriks G. The effect of patient portals on quality outcomes and its implications to meaningful use: a systematic review. J Med Internet Res 2015;17(2):e44 [FREE Full text] [CrossRef] [Medline]
    5. Hogan TP, Nazi KM, Luger TM, Amante DJ, Smith BM, Barker A, et al. Technology-assisted patient access to clinical information: an evaluation framework for blue button. JMIR Res Protoc 2014;3(1):e18 [FREE Full text] [CrossRef] [Medline]
    6. Centers for Medicare and Medicaid Services. Stage 2 Eligible Professional Meaning Use Core Measures Measure 8 of 17. 2014 Nov.   URL: https:/​/www.​cms.gov/​Regulations-and-Guidance/​Legislation/​EHRIncentivePrograms/​downloads/​Stage2_EPCore_8_ClinicalSummaries.​pdf [accessed 2015-10-07] [WebCite Cache]
    7. Neuberger M, Dontje K, Holzman G, Corser B, Keskimaki A, Chant E. Examination of office visit patient preferences for the after-visit summary (AVS). Perspect Health Inf Manag 2014;11:1d [FREE Full text] [Medline]
    8. Chung A, Shea C. After-visit clinical summaries: what is meaningful to patients? 2013 Nov 18 Presented at: American Medical Informatics Association Symposium; Nov 18, 2013; Washington, DC.
    9. Ralston JD, Carrell D, Reid R, Anderson M, Moran M, Hereford J. Patient web services integrated with a shared medical record: patient use and satisfaction. J Am Med Inform Assoc 2007;14(6):798-806 [FREE Full text] [CrossRef] [Medline]
    10. Pavlik V, Brown AE, Nash S, Gossey JT. Association of patient recall, satisfaction, and adherence to content of an electronic health record (EHR)-generated after visit summary: a randomized clinical trial. J Am Board Fam Med 2014;27(2):209-218 [FREE Full text] [CrossRef] [Medline]
    11. Emani S, Yamin CK, Peters E, Karson AS, Lipsitz SR, Wald JS, et al. Patient perceptions of a personal health record: a test of the diffusion of innovation model. J Med Internet Res 2012;14(6):e150 [FREE Full text] [CrossRef] [Medline]
    12. Turvey C, Klein D, Fix G, Hogan TP, Woods S, Simon SR, et al. Blue Button use by patients to access and share health record information using the Department of Veterans Affairs' online patient portal. J Am Med Inform Assoc 2014;21(4):657-663. [CrossRef] [Medline]
    13. Fishbein M, Ajzen I. Predicting and Changing Behavior: The Reasoned Action Approach. New York: Psychology Press; 2010.
    14. McEachan RRC, Conner M, Taylor NJ, Lawton RJ. Prospective prediction of health-related behaviours with the Theory of Planned Behaviour: a meta-analysis. Health Psychology Review 2011 Sep;5(2):97-144. [CrossRef]
    15. Yamin CK, Emani S, Williams DH, Lipsitz SR, Karson AS, Wald JS, et al. The digital divide in adoption and use of a personal health record. Arch Intern Med 2011 Mar 28;171(6):568-574. [CrossRef] [Medline]
    16. Ajzen I. Attitudes and persuasion. In: Deaux K, editor. The Oxford Handbook of Personality and Social Psychology (Oxford Library of Psychology). New York: Oxford University Press; 2012.
    17. Woods SS, Schwartz E, Tuepker A, Press NA, Nazi KM, Turvey CL, et al. Patient experiences with full electronic access to health records and clinical notes through the My HealtheVet Personal Health Record Pilot: qualitative study. J Med Internet Res 2013;15(3):e65 [FREE Full text] [CrossRef] [Medline]
    18. Nazi KM, Turvey CL, Klein DM, Hogan TP, Woods SS. VA OpenNotes: exploring the experiences of early patient adopters with access to clinical notes. J Am Med Inform Assoc 2015 Mar;22(2):380-389. [CrossRef] [Medline]
    19. Walker J, Leveille SG, Ngo L, Vodicka E, Darer JD, Dhanireddy S, et al. Inviting patients to read their doctors' notes: patients and doctors look ahead: patient and physician surveys. Ann Intern Med 2011 Dec 20;155(12):811-819 [FREE Full text] [CrossRef] [Medline]
    20. Delbanco T, Walker J, Bell SK, Darer JD, Elmore JG, Farag N, et al. Inviting patients to read their doctors' notes: a quasi-experimental study and a look ahead. Ann Intern Med 2012 Oct 2;157(7):461-470 [FREE Full text] [CrossRef] [Medline]
    21. Ralston JD, Coleman K, Reid RJ, Handley MR, Larson EB. Patient experience should be part of meaningful-use criteria. Health Aff (Millwood) 2010 Apr;29(4):607-613 [FREE Full text] [CrossRef] [Medline]
    22. Kaelber DC, Jha AK, Johnston D, Middleton B, Bates DW. A research agenda for personal health records (PHRs). J Am Med Inform Assoc 2008;15(6):729-736 [FREE Full text] [CrossRef] [Medline]
    23. Archer N, Fevrier-Thomas U, Lokker C, McKibbon KA, Straus SE. Personal health records: a scoping review. J Am Med Inform Assoc 2011;18(4):515-522 [FREE Full text] [CrossRef] [Medline]
    24. Hummel J, Evans P. Providing Clinical Summaries to Patients After Each Office Visit: A Technical Guide. Seattle: Qualis Health; 2012.   URL: https://www.healthit.gov/sites/default/files/measure-tools/avs-tech-guide.pdf [accessed 2016-03-10] [WebCite Cache]
    25. Tang PC, Newcomb C. Informing patients: a guide for providing patient health information. J Am Med Inform Assoc 1998;5(6):563-570 [FREE Full text] [Medline]
    26. Emani S, Ting DY, Healey M, Lipsitz SR, Ramelson H, Suric V, et al. Physician perceptions and beliefs about generating and providing a clinical summary of the office visit. Appl Clin Inform 2015;6(3):577-590. [CrossRef] [Medline]


    Abbreviations

    AVS: after-visit summary
    EHR: electronic health record
    HITECH: Health Information Technology for Economic and Clinical Health
    MU: Meaningful Use
    TPB: Theory of Planned Behavior


    Edited by A Moorhead; submitted 28.10.15; peer-reviewed by K Nazi, C Kruse, G Vergeire-Dalmacion; comments to author 20.11.15; revised version received 12.01.16; accepted 17.01.16; published 13.04.16

    ©Srinivas Emani, Michael Healey, David Y Ting, Stuart R Lipsitz, Harley Ramelson, Vladimir Suric, David W Bates. Originally published in the Journal of Medical Internet Research (http://www.jmir.org), 13.04.2016.

    This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), 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.