Background
Hospital settings present important challenges to communication of medical information between providers and patients. Patients often have multiple active conditions, undergo numerous tests, and receive complex treatments that evolve throughout their hospital stay. Care teams are large and team membership is dynamic because of the need to provide care 24 h a day, 7 days a week [
1‐
3]. Further, the verbal information provided to patients during daily rounds is seldom complemented by any other mode of communication [
4]. As a result of these challenges, hospitalized patients often have incomplete comprehension and less than optimal engagement in their care [
5‐
8].
Patient portals, now commonly used in ambulatory settings, leverage electronic health records (EHR) in an effort to inform and engage patients [
9,
10]. Until recently, little attention had been given to using portals as a tool for hospitalized patients. Several small studies have evaluated the use of tablet computers with mobile patient portal applications designed for hospitalized patients [
11‐
14]. Although results indicate patients are generally enthusiastic about such applications, the impact on patient comprehension, engagement, and clinical outcomes is largely unknown. Our research team recently developed and tested a hospital-based mobile patient portal which included information about the medical team, scheduled tests and procedures, and a list of active medications [
15]. While the portal improved patients’ knowledge of physician team members, we found no effect on patient knowledge of the care plan or activation. Importantly, patient use of the portal was highly variable. We hypothesized that limited use of the portal by some patients may have contributed to these findings. Thus, the overarching goal of this study was to identify barriers to portal use and solicit user input to inform the design of future versions of a hospital-based patient portal.
The specific objectives of this study were to explore: (1) patient perspectives regarding portal content and features perceived to be most beneficial, (2) patient barriers to portal use and recommendation for improvement, and (3) healthcare provider perspectives regarding both benefits and challenges that a hospital-based patient portal may pose.
Discussion
We found that participants generally felt the portal was useful as a tool to inform and engage patients. However, patients and healthcare providers readily identified opportunities for improvement. Several themes emerged which reveal opportunities to enhance the design of portals intended for hospitalized patients.
Regarding portal content, both patients and providers described the utility of including information about current medications and several suggested enhancements to the portal such as including timing of the last dose for “as needed” medications, use of both generic and brand names, and indicating the class or purpose of medications. Regarding potential additional features, patients expressed interest in test results. This finding is consistent with research by Dykes, Dalal, and colleagues demonstrating that hospitalized patients have a strong interest in all types of test results [
14,
17]. Importantly, healthcare providers in our study worried that, should results be made available to patients, the large volume, high percentage of abnormal results, and complex medical terminology could cause anxiety and confusion for patients. Importantly, studies of outpatient portals have found that few patients report increased anxiety or difficulty understanding the clinical information [
18‐
20]. Less is known about providing access to clinical information for hospitalized patients. Prey and colleagues published a study in which they provided hospitalized patients with paper copies of their medical records including physicians’ progress notes, laboratory test results, radiology reports, and operative reports [
21]. Patients perceived the information as highly useful even if they did not fully understand complex medical terms. Pell and colleagues also recently published a study in which they provided tablet computers to hospitalized patients, allowing access to medication schedules, laboratory results, and plain radiography results to patients on a medical unit [
12]. Although comprehension was not objectively assessed, use of the portal did not increase patient reported anxiety or confusion. These studies suggest that clinicians’ concerns may be unfounded, but further research is needed to objectively assess the impact of providing clinical information via a portal on the levels of anxiety and comprehension among hospitalized patients.
Healthcare providers expressed a desire to control which results were released to patients, the timing of release, and to annotate the results. However, providers also identified the potential for such manual release to increase their workload. Overall, these findings highlight the need for careful decisions related to type of results to release, as well as the timing, rules, and format of display. While principles of transparency, patient empowerment, and shared decision making would support immediate release of results, the most pragmatic approach likely entails pre-selecting certain results for timed automated release. For example, an identified rubric could require that all basic chemistry and blood count results finalized before 7 AM as well as plain radiograph, CT, and MRI results finalized before midnight be automatically released at noon, whereas pathology reports and sensitive laboratory results (e.g., HIV, CD4 counts) might be communicated only through verbal discussion.
We found that both patients and providers were interested in giving patients the ability to record notes and questions. More advanced communication options, like two-way messaging via the portal, were strongly opposed by healthcare providers. Physicians feared that patients would send more messages than they could manage, have unrealistic expectations regarding timeliness of response, inappropriately use messaging for urgent clinical matters, and that messaging might damage the patient-physician relationship. Although studies of ambulatory based portals have found a high level of patient satisfaction with secure messaging, data on physicians’ perceptions is lacking [
10,
22,
23]. Dalal and colleagues recently published a study of a hospital-based patient portal in which two-way messaging was available [
14]. The volume of messages was fairly low (~1.8 messages per patient), which they attributed to efforts to set expectations at the time of enrollment and lack of timely responses from providers.
Recognizing the importance of surrogate decision makers and that surrogates are often unable to be present during physician rounds, healthcare providers suggested that family members and loved ones be given access to the patient portal. This recommendation is supported by a recent study by Torke and colleagues who found that surrogates were involved in decision making for nearly half of hospitalized older adults [
24]. Providing remote access to surrogates would require specific steps at both a local and national level [
25]. At the local level, hospitals need to develop policies allowing patients to appoint designees and strong authentication procedures using a unique login and password for each designee. At a higher level, federal support is needed to develop standards in this area and for research to understand how to best provide remote access.
The game and entertainment applications on the tablets were extremely popular among patients. Several patients described how use of games helped them cope with their illness. This finding is unique among studies of hospital-based patient portals, but consistent with prior research showing that video games can serve as a positive distraction and a complement to other symptom management strategies [
26,
27]. In light of the variable use found in our initial hospital-based portal study and generally low adoption of ambulatory-based portals [
11,
15,
28], patients’ interest in games and entertainment could also be leveraged to promote use of the portal and enhance engagement in care. For example, patients could be asked to review and confirm understanding of their current medication list prior to being given access to additional game applications. Gamification of the portal itself, may also promote use [
29]. A recent systematic review published by Otte-Trojel and colleagues on the development of patient portals found that few studies had evaluated the use of promotional initiatives incorporated into portal design to attract patient attention and encourage use [
30].
Patients generally found the portal application easy to use. Several patients identified opportunities to improve the portal interface, all of which appear easily achievable. Some patients identified barriers related to the tablet itself, highlighting the need to provide devices that accommodate patients with physical and visual impairment.
Our study has several limitations. We conducted a qualitative evaluation of a custom designed, mobile patient portal application at a single site. Though our sample size was relatively small, we achieved thematic saturation. Prior research has shown that saturation often occurs in interview studies within the first 12 interviews and basic elements for themes may be present as early as six interviews [
31]. Our portal was designed for hospitalized patients and lacked content and features frequently available in ambulatory-based portals. Our overarching goal was to identify barriers to portal use and solicit user input to inform the design of future versions of a hospital-based patient portal. Though patient characteristics and provider workflow may vary, we believe our findings are generalizable to other hospital settings. Importantly, our findings can help inform a range of decisions during implementation and adaptation of EHR vendor developed patient portals for use in hospital settings (e.g., Epic MyChart Bedside). Finally, a single individual (AK) conducted the interviews, moderated the focus groups, and served as one of the two primary analysts. Though relatively common in qualitative research, this arrangement may have increased the potential for the researcher to become anchored prematurely to initial impressions.