Patients play a crucial role in managing their health, and that role is even more central in the context of chronic conditions that require ongoing coping skills and self-management efforts [1,2]. Broadly, the term “patient engagement” refers to patients taking part in or actively participating in their care [3], including how patients work with health care providers and systems to manage and improve their health. Hibbard [4] conceptualized patient engagement as degrees of active involvement in their care, termed patient activation. Patient engagement can operate at multiple levels and exists on a continuum [5]. We focus primarily on the level of direct patient care where engagement can range from patients receiving information about a diagnosis to patients participating in treatment plan decisions [5]. Although it has been characterized in a variety of ways, a common thread is that patient engagement is a multi-faceted concept incorporating actions on multiple levels to promote patient-centered care and self-management.

Patient-facing eHealth technologies are often described as a means to improve patient engagement [6,7]. Online patient portals have received particular attention since their features often support healthcare system transactions (eg, prescription refills, scheduling), healthcare team communication, access to patient medical records, and the delivery of health education resources [7,8]. Previous studies of patients living with diabetes, for example, suggest that individuals who use online patient portals are more likely to achieve desirable outcomes, including improved glycemic control [9-12].

Few studies have attempted to specify how engagement may be realized through patient portals or the pathways through which portals may help patients engage in their care and self-management. One study of direct-release of test results via patient portals found that this access to information improved engagement by helping patients to monitor results over time and prepare before communicating with their provider [13]. However, a greater understanding of pathways through which a range of portal features support engagement is needed.

A framework for patient engagement by Barello and colleagues may help us understand the different ways in which a patient portal may support patient engagement [14]. This framework incorporates patients’ actions, thoughts, and feelings and considers patient engagement within three dimensions: emotional, cognitive, and behavioral [14]. In this model, the emotional dimension of engagement encompasses the patients’ emotional state and feelings about their care. The cognitive dimension captures patients’ understanding of their disease and treatment. The behavioral dimension of engagement captures patients’ actions to address their health.

In this paper, we examine whether and how an online patient portal supports patient engagement for individuals living with diabetes. We take a qualitative approach to this by reviewing patient portal experiences among veterans living with diabetes and describe several ways in which an electronic patient portal facilitates patient engagement.

Data were collected in a more extensive study examining the role of patient portals in diabetes management, described previously [15]. Briefly, from November 2017 to January 2018, 1200 patients living with diabetes were invited to participate in a mailed survey about their diabetes self-management and use of the United States Department of Veterans Affairs (VA) online patient portal, My HealtheVet [16]. The My HealtheVet portal offers several features to help patients manage their care. Patients can request medication refills, schedule appointments, receive appointment reminders, communicate electronically with providers through secure messaging, and use the Blue Button feature to access portions of their medical records including labs and clinical notes [17].

All 1200 patients had a diabetes diagnosis in their record since at least 2011, had experienced uncontrolled diabetes in 2012 (mean HbA_1c >8.0% and less than 25% of the year with an HbA_1c <8.0%) and were actively engaged patient portal users who used secure messaging at least four times between January 2016 and June 2017. Since 2012, half had achieved glycemic control, defined as having mean HbA_1c <8% in 2016 and HbA_1c <8% for at least 75% of 2016, and half had not. We stratified each group based on urban/rural residence, according to the Rural-Urban Commuting Areas [18], and the presence of comorbid mental health diagnoses. Individuals were considered to have a comorbid mental health diagnosis if they had two outpatient diagnoses or one inpatient diagnosis within each diagnostic group. The diagnostic groups selected using International Classification of Diseases (ICD), 9th and 10th Revision (ICD-9 and ICD-10) codes were anxiety, bipolar, depression, personality disorder, post-traumatic stress disorder, schizophrenia, and substance use disorders. We received 449 completed surveys, of which 350 patients responded that they would be interested in participating in an interview about how they managed their diabetes. This study was approved by the Edith Nourse Rogers Memorial Veterans Hospital Institutional Review Board.

Purposeful sampling was used to identify and select patient portal users for telephone interviews. In total, 160 respondents provided positive responses to an open-ended survey question, “Can you tell us about an ‘A-Ha!’ moment when you realized you could use the My HealtheVet portal to better manage your diabetes?” We selected interviewees to represent a variety of responses to this and other survey items about My HealtheVet use. Interviewees were selected to represent those who used a variety of My HealtheVet portal features, those with controlled and uncontrolled diabetes, urban and rural patients, and those with and without comorbid mental health diagnoses. Women and minority veterans were oversampled to broaden the representation of patient demographics.

Forty telephone interviews were completed between February and May of 2018. Each interview was conducted by two researchers and averaged 70 minutes long. The interview covered how the patient managed their diabetes, including their feelings about the efficacy of their self-management, how they gathered and used information, how portal features and other technologies supported their diabetes management, and their suggestions for improving the portal. Interviewers were intentionally blinded to the patient’s glucose control at the time of the interview.

Across all 40 patient interviews, 30 touched on concepts related to patient engagement through portal use. Overall, patients reported feeling engaged by their use of the portal. We describe this below, followed by illustrations of the four pathways of interaction with the portal that support patients’ feelings of engagement. Finally, we share patient insight on how portal functionality may hinder engagement. To further contextualize our data, we describe select patient characteristics after each illustrative quote.

Through 40 interviews with patients with diabetes who used a patient portal, we engaged in a wide-ranging discussion regarding patient engagement and pathways by which portals may facilitate engagement. Access to detailed health information in the portal facilitates engagement by allowing patients to learn about their condition, remember information from provider visits that they otherwise may forget, and prepare for medical appointments. Tools to facilitate communication through the portal allow patients to partner with their providers to manage their health. Many patients found secure messaging and Blue Button features supportive of engagement; difficulties that patients reported with these features may, by extension, be barriers to patient engagement.

Patient narratives in this study align with Barello and colleagues’ framework [14] of three engagement dimensions (emotional, cognitive, and behavioral) and highlight an interconnectedness between cognitive and behavioral engagement. In terms of emotional engagement, patients felt that interacting with providers via the portal improved the patient-provider relationship. Portal use cognitively engaged patients to understand their health by using portal tools such as secure messaging to ask questions and Blue Button to review clinical notes. Patient narratives revealed how cognitive engagement is translated into behavior. Access to information through the patient portal led patients to engage in behaviors to support their health and self-management. Actions facilitated by the portal’s information and communication platform included preparing for visits, coordinating care among providers, and making lifestyle or medication changes between visits.

There are concerns in the literature that eHealth initiatives may worsen disparities [24] or weaken patient-provider relationships [12]. Disparities could worsen if vulnerable groups have less access to technology that promotes engagement or if vulnerable populations are less likely to use patient portals [25]. Our data suggest varied backgrounds and degrees of glucose control among patients using the portal to enhance their engagement. Portal benefits for engagement seem to be experienced by all types of patients, including individuals in urban and rural settings, with and without mental health conditions, those with diabetes in control, and those with diabetes not in control. Furthermore, several patients who were not in control described activities related to engagement. A portal may offer further opportunities to work with these patients to improve diabetes control. However, engaging patients via portal use does not necessarily indicate they will have better outcomes. In this analysis, we do not measure whether engagement facilitated by the portal influenced diabetes outcomes or patient satisfaction scores or whether this varied based on additional patient characteristics. The interviews were overwhelmingly positive about the benefits of the portal and suggested patient satisfaction is improved among those who use it. Future research should explore whether portal engagement predicts health outcomes and patient satisfaction and whether the findings are consistent among patients with other chronic conditions.

Patient-provider relationships are essential for individuals with chronic conditions like diabetes. There is some concern that portals may hurt the patient-provider relationship if portal interactions replace face-to-face interactions [12,26]. However, in our study, patients described the portal interactions between patient and provider as a key factor that facilitated their engagement. Some patients reported feeling that their interactions with providers through the portal help strengthen their relationship with providers. A point of caution: we must recognize the unintended consequence that a portal’s design may have on patient perceptions of their provider’s willingness to communicate with them. In our analysis, most patients found the portal helpful, but the design of the secure messaging feature led some patients to feel their providers may not want to communicate with them. Employing human-centered design may be one strategy to help mitigate portal design issues by engaging patients and providers early in the design process. One patient also complained about the impersonal, “cookie-cutter” nature of some clinical notes, which could make patients feel like their providers didn’t know them as individuals, or that the notes did not capture the essence of what was discussed during the clinical encounter. Future qualitative work should continue to explore patient and provider perspectives on how portals affect their relationships and could examine provider attitudes and approaches to the portal as a possible moderator of patient engagement.

This study has several limitations. The study population was limited to United Stated military veterans with diabetes who used the My HealtheVet patient portal. We found overwhelmingly positive comments regarding the portal, perhaps because all interviewees were portal users. Our study was not designed to determine the prevalence of patient engagement, but our findings do reveal potential ways in which a portal may facilitate engagement.

In this study, patient narratives helped identify pathways by which a portal may facilitate patient engagement. Patients found the portal helped strengthen their relationship with providers and helped the patient feel engaged. Patients reported finding the portal useful for receiving help in managing symptoms, coordinating their care, and learning about their health. Patient portal users included a variety of individuals who described the engagement benefits of the portal. The group included individuals living in urban and rural settings, those with and without mental health conditions, and those with controlled and uncontrolled diabetes. Thus, the portal may help a wide range of portal users engage with their care.