Chronic illnesses, including cardiovascular disease, arthritis, diabetes, asthma, cancer or chronic obstructive pulmonary disease (COPD), affect more than 145 million Americans, representing a large increase over projections made only 10 years ago [1-3]. The individual burden of chronic illness is magnified by the high rate of two or more co-occurring chronic illnesses, termed multimorbidity [1]. This population is at particularly high risk for a variety of adverse health outcomes including poor functional status, unnecessary hospitalizations, and adverse drug events [4-6]. Chronic illnesses account for 7 out of 10 deaths each year. In addition, chronic illness has resulted in huge costs to medical systems and society, accounting for 84% of the total US health care spending [1], or almost 15% of the gross domestic product (GDP). For these reasons the Agency for Health Care Research and Quality (AHRQ) has designated people with multimorbidity as a priority population [7].

Individuals with chronic illness generally receive care in ambulatory settings and often look to primary care providers to direct their care. However, the rise in chronic illness is shifting the role of the physician to that of facilitator in the patient’s self-management process. The increase in availability of personal disease monitoring tools such as hand-held, self-monitoring blood glucose systems [8], and a focus on patients as consumers of health care is moving expectations around management of chronic conditions away from laboratories and physicians to the patients themselves. Evidence shows that enhanced patient self-management can lead to better control of chronic illness [9-11], and health information technology (HIT) is a potentially important mechanism to facilitate patient self-management [12-15]. Large studies in outpatient settings have found that providing patients with access to their medical record, physician progress notes, personalized health information, and reminders (ie, patient-centered functionalities) leads to increases in adherence to guidelines, health status, and patient satisfaction [16,17].

Patient portals are a class of electronic personal health records (PHRs)—tools that patients can use to track and manage their health. The PHR is “an electronic record of an individual’s health information by which the individual controls access to the information and may have the ability to manage, track, and participate in his or her own health care” [18]. PHRs can be stand-alone or tethered. Stand-alone products allow the patient full control over what data are entered and accessed and are independent of the provider, allowing the patient to input their own data from any provider regardless of the electronic health record (EHR) system (if any) the provider uses, and to carry their data with them across providers. Tethered PHRs, also called patient portals, are offered through a health care provider and are connected to the patient’s EHR with that provider. These patient portals provide access to information in the patient’s EHR, controlled by the provider, as well as other functions such as viewing and scheduling appointments and secure communication with the provider [19]. Both types of PHRs show promise for assisting patients in self-management of chronic conditions by allowing patients to input and track health information, facilitating communication between patients and providers, and providing access to consumer-friendly information about diseases [20-24].

Research has shown that patients with special health care needs, such as those with multiple chronic conditions, have the greatest interest in patient portals [25-27]. The first large-scale study of the use of patient portals within a large health system found that having more chronic conditions predicted both adoption and intensity of patient portal use [28]. However, there is still much to understand about why, how, and which patients use portals, and how the health care system can best support them. Currently, little is known about what motivates patients to adopt and continue to use portals, and what functionalities patients consider important for self-management of their conditions [13-15].

Patient engagement, defined by the Institute for Healthcare Improvement as “actions that people take for their health and to benefit from care” [29], is critical to the management of chronic diseases. A 2013 editorial in Health Affairs referred to patient engagement as the “next blockbuster drug of the century” [29]. Patient portals are positioned as a central component of patient engagement through the potential to change the physician-patient relationship and enable chronic disease self-management [30-33]. Studies of outpatient portals suggest that patients want accurate and timely information provided across the continuum of care that they can apply to their care and communicate with providers in a secure and trusted manner [14,34].

Despite the focus on patient engagement [29,35], research on patient engagement in the inpatient setting is in its infancy. A recent systematic review found only four studies testing the provision of patient-specific information in the inpatient setting [36]. The focus of these studies included providing access to the patient record and information on the care team through a mobile phone app [37], a tablet computer app to view care team profiles and hospital medication records, a tablet app with the plan of care, diet and safety information [38], and large in-room information displays in an emergency department [39]. While these small-scale qualitative case studies reported positive findings, including patient reports of enhanced engagement in the care process and satisfaction with care, none included patient-centered functionality such as the ability to send messages to the care team, allowing patients to input information or record notes—elements that have been demonstrated to further enhance patients’ engagement [17,40,41].

Information and technology are insufficient to fully engage patients in their care; patients also need motivation to engage [42-46]. A common element of health behavior change theories is the need for a trigger to action [47,48]. This is supported in studies of individual behavior change across a variety of health behaviors [49-54]. For patients with multiple chronic conditions, hospitalization is often due to exacerbations of one or more condition. We assert that hospitalization can serve as the necessary trigger that engages these patients in managing their care [47,48]. In other areas, times of acute crisis have been linked to a greater perception of risk and increased focus on health behaviors [49-52]. Therefore, hospitalization may create a window of higher engagement in which to initiate behavior change and foster interest in tools for managing health.

Most patient portal implementations have assumed that internal documentation is sufficient, and that the application is sufficiently intuitive for use with only a supplemental list of “frequently asked questions (FAQs)”; however, documentation and FAQs alone are not always sufficient (see Pilot Study section). In particular, individuals with low health literacy and low technological literacy may experience greater difficulty navigating these tools and may simply give up. The Institute of Medicine defines health literacy as “the degree to which individuals have the capacity to obtain, process, and understand basic health information and services needed to make appropriate health decisions” [55]. Electronic health (eHealth) literacy, then, is “the ability to seek, find, understand, and appraise health information from electronic sources and apply the knowledge gained to addressing or solving a health problem” [56]. Technological literacy, including knowing how to operate computers and mobile devices, and how to make use of the information provided, are essential components of eHealth literacy.

The multimorbid tend to be older and more socioeconomically disadvantaged than the population at large [57], so people with multiple chronic illnesses may be especially challenged in developing eHealth literacy skills. However, people of all ages and status struggle with health literacy. Poor health literacy is a better predictor of poor health status than income, employment status, or education [55]. Interventions thus need to provide training specifically designed to overcome these challenges. Further, the introduction of new technology into the care process, such as tablet computers, may create barriers to participation due to the competing demands on a patient’s time inherent in hospitalization. If these tools are to accomplish their intended aims of empowering patients through greater engagement, it is critical to understand the training and support materials necessary to meet diverse patient needs.

We initiated a pilot study in 4 units within Ohio State University Wexner Medical Center (OSUWMC) that tested the specifications for implementation of MCB. This preliminary study served to formatively guide our High Tech/High Touch (HT2) proposal. For instance, the pilot identified the need for a High Touch intervention, the need to use TNs (rather than existing staff) to provision tablets, and the need for easily accessible support materials. Videos created for this pilot serve as the initial basis for our Low Touch intervention, and we will make modifications as needed for our study purposes.

In preparation for our proposed HT2 study, we secured ongoing report data on the use of MCA. From these data we know that the mean age of MCA users is 47.6 years (SD=16); and that most are white, have managed care insurance, and have used MCA for nearly 2 years. It was found that 44.82% (31,439/70,152) of all patients created a MCA account; however, only 16.00% (5029/31,439) had logged on and used their account during the pilot study time frame. This pilot served to reinforce our need to focus on the use of functions within MCA and MCB, rather than the standard model promoting account sign up and leading to only intermittent use.

Dr Huerta has been assessing usability of functions within Epic. Working with students to engage health practitioners, preliminary work was conducted to explore the usability of the system. Pilot research results have recently been made available to the health system.

By focusing specifically on engaging patients with multimorbidity in the use of patient portals across the continuum of care, HT2 addresses one of the main goals set forth in the 2010 Department of Health and Human Services report, Multiple Chronic Conditions: A Strategic Framework [4], to implement and effectively use health care technology. As multimorbidity research is not disease-specific, our study focuses on self-efficacy and satisfaction as the outcomes of interest because they have been associated in previous research with positive health outcomes [7] and are considered generic outcome measures that are responsive to change over time [6]. Current studies of patient portal use focus mainly on metrics such as number of users and email response times, but these do not necessarily reflect use in a way that impacts an individual’s health. We plan to study the potential for how introducing patient portals during an inpatient stay can influence patient-centered care and outcomes including patients’ perceived self-efficacy for managing chronic conditions across the continuum of care. Further, we expect to be able to examine both readmission rates and rates of ambulatory patient portal utilization (MCA) after discharge in this priority patient population.

Our study also highlights the importance of patients’ engagement in their care. Patient portals offer several tools to help patients become and remain engaged. In the inpatient setting, when patients may be particularly ready to learn about managing their health, patients can learn how and why it is important to track biometric measures such as weight or blood glucose levels, receive feedback from and ask questions of their providers specific to the management of their conditions, and access disease-specific educational materials. They can also send secure messages to their hospital and family physicians. This communication opportunity has the potential to change patient engagement with their nursing team as well. In our MCB pilot study (discussed below), nursing staff considered it important to disseminate educational material throughout the hospital stay rather than simply at discharge, because this allows education to happen when the patient is ready to engage. These same materials can then be made available in MCA when the patient leaves the hospital. With this study, we seek to clarify and measure how engagement in the inpatient setting (use of MCB) facilitates continued engagement once the patient leaves the hospital (ongoing use of MCA). At issue is whether use during an inpatient stay reinforces ambulatory use.

Our study aims to improve health care and outcomes through implementation of enhanced patient-centered HIT. The integration of patient portals within EHR systems has the potential to improve the patient experience and the quality of patient care [28,58-60]. However, use of patient portals beyond creating an initial account remains limited, and is based on the technology's relevance to the patient and the ability of the technology to enhance the physician-patient relationship [34,61,62]. Our study focuses on increasing relevance and improving communication to promote patient portal use in order to improve care, outcomes, and experience for patients with multiple chronic conditions. We specifically explore the impact of education strategies as a means to compensate for differences in technology and health literacy. In doing so, we intend to explore the use of both technology and literacy training to increase self-efficacy in the use of these tools. Evidence shows that enhanced patient self-efficacy for management activities can lead to better control of chronic illness [9-11].

The patient portals to be studied at OSUWMC are available through the Epic EHR system in use at Ohio State. Epic currently touches over 50% of all Americans as they receive health care and is available to 50 million patients in the United States [20,63]. Both MCA and MCB patient portals are integrated into the Epic EHR at OSUWMC.

The study randomizes individual patients within 1 of 6 general medicine units from OSUWMC into 1 of 4 study arms (Figure 1). While the unit of analysis and intervention is the patient, absent unit level controls, the study would suffer from endogeneity issues related to unit workflow dynamics that are changed by the presence of the tools, and as such the use of a unit-level control is justified. We will identify 6 matched pairs of general medicine units (12 total) on the basis of Case Mix Index, average length of stay, number of beds, nurse staffing levels and “nurse to bed staffing ratio,” and select units for the intervention in each pair. Matched controls offer the ability to calculate a de-identified reference cohort of patients against which readmission and MCA uptake can be measured. In addition, we will conduct interviews with a purposive sample of health care professionals across the system to examine how the roll out of MCB impacts patient discussions in both the inpatient and ambulatory care settings; these interviews will include topics such as expectations of MCA usage, and perceptions of the technology and touch interventions.

Sample size and power calculations were made using current patient satisfaction scores as a baseline for each measure using G*Power 3.1 software. We expect this design to be sensitive enough to identify changes of between 2.2% and 3.8% across the panel of patient satisfaction scores in the Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) with a minimum of 968 observations per cell to accomplish this level of sensitivity. Our approach entails conservative oversampling (1500) to mitigate potential participant drop out. Inclusion criteria for the study is presented in Textbox 2.

This phase of the study will begin immediately upon funding and is expected to take 9 months. In total, 20 patients and/or caregivers will be involved in the pre-intervention phase.

The intervention phase of the study is expected to be completed within 18 months. The targets for recruitment are 6000 patients and/or caregivers, and 100 providers.

OSUWMC is only the second health system in the nation to offer this highly innovative patient portal designed specifically for the inpatient environment. MCB provides patients with situation-specific information, such as daily schedule information, routing of questions to their care team, information about the care team, and immediate release of lab and test results. This will be the first large-scale study of an interactive tablet-based patient portal available at the inpatient bedside. This is a significant step toward using HIT to engage patients across the continuum of their care. OSUWMC has a strong relationship with Epic, and the research we propose will likely influence Epic implementations throughout the United States, thus potentially improving care delivery and outcomes nationwide.

Further, while health systems have not yet deployed such systems, the issue is more associated with an absence of tools provided by vendors as opposed to an unwillingness of health systems to implement. As the technology matures and more information comes online detailing the value proposition, these tools will see increasing use as well as greater diversity of tool availability (eg, tablet-based, mobile PHRs, etc). Further, as patients are better able to manage their disease states, they will “bend the cost curve” [80] by moving from high-cost interventions (eg, emergency room visits) to lower cost management of their conditions. The supposition is that the better we can engage patients in self-management of their own care and be effective at that practice, the lower the cost of care. Our study thus seeks to explore and gain a greater knowledge of how one such tool ̶ an inpatient patient portal ̶ can serve to not only influence outcomes related to the inpatient setting but throughout the entire continuum of care, including transitions and the outpatient care environment.

While it would have been possible to conduct a smaller study that sought to explore the implications of patient portal adoption in greater detail, the need for larger-scale quantitative studies is without question. Existing studies have been resource limited and have not explored their implications in a RCT. As HIT tools such as patient portals become more prevalent, the likelihood of being able to execute a controlled trial is diminished. Rarely is a health system willing to subject the organization to such a broad scale intervention as Ohio State University is proposing.

In addition, our Touch intervention will train patients not only in the technical aspects of utilizing the inpatient patient portal (MCB, the how) but also in the general use of an ambulatory patient portal (MCA) to manage their specific health condition over time (the why). Because the tablet MCB app allows an assessment of the patient’s comfort with technology and health literacy at the time of training, the Touch intervention can be tailored to the individual patient’s needs.

Finally, our study will collect quantitative and qualitative data from a sufficiently large and diverse sample of patients to enable us to make comparisons across patient subgroups. We will assess the impact of our interventions in a variety of subgroups including those identified by demographics, by health condition, and by experience and comfort with technology. The results of this study will thus help health systems and providers understand what levels of technology and training work best to engage which patients.