Background: Sub-optimal discharge preparation during hospitalization may adversely impact safety and lead to a poor patient experience. As part of an AHRQ-funded study, we designed and developed interactive digital health tools (Figure 1) to engage patients and caregivers in self-assessing discharge preparedness: an educational video and 16-item discharge checklist addressing 4 domains (understanding the plan, medications, self-care, follow-up) administered to patients 24-48 hrs prior to their expected discharge date (EDD); a safety dashboard integrated into the electronic health record (Epic, Inc.) displaying patient-reported discharge concerns from the checklist for clinicians to view in real-time; and a mechanism for patients to request secure text messaging with a physician to address issues that arise after discharge for up to 7-days. The purpose of this study was to evaluate our implementation experience.

Methods: This IRB approved study was conducted on general medicine units at a large academic medical center in Boston, MA from 1/2018 through 8/2018. Research assistants approached patients or a designated healthcare proxy for enrollment. Research assistants coached patients/caregivers to watch the video and complete the checklist on a mobile device (patient portal or web-based REDCap survey). We categorized reasons for non-participation. We used descriptive statistics to quantify frequency of patient-reported discharge concerns by domain as well as usage of each component. We used a 2-person consensus approach to analyze qualitative data collected from interviews and focus groups of patients and clinicians to identify implementation barriers.

Results: Of 756 patient-admissions, the discharge checklist was successfully submitted for 510 (67.5%) patients (480 unique patients, mean age of 58.6 ±17.9 yrs; 230 (45.1%) male; 340 (66.7%) Caucasian; 305 (59.8%) government insurance, mean LOS of 8.78±7.93 days). For the remaining 246 patient-admissions, the patient was unavailable (126); not appropriate per nurse (97); declined to participate (41); did not speak English / no caregiver was available (33); did not respond by email when reminded (8); or encountered technical issues (8). The video was watched prior to completing the checklist in 416 (81.6%) patient-admissions. On average, patients reported 4.24 concerns based on checklist responses, most commonly about medications (30.7%) and follow-up (30.3%). In 210 (41.2%) patient-admissions, a member of the patient’s care team–most often a nurse–accessed the dashboard discharge column to view concerns. For 422 patient-admissions in which secure messaging was offered, 141 (33.4%) patients provided their mobile number during checklist submission; of these, a communication thread was initiated by a physician for 3 (2.1%) discharges. From our qualitative analysis, we identified 10 key implementation barriers.

Conclusions: Most patients approached were willing to watch a video and complete a checklist to self-assess discharge preparation, and many reported concerns about medications and follow-up. Use of the dashboard by clinicians was modest and could be improved by improving specificity of patient-reported concerns, linking to clinical actions, and stratifying patients by readmission risk. Novel features, such as post-discharge texting, likely require institutional and policy-level changes to facilitate adoption. The strategies (Table 2) we propose could help address implementation barriers and promote adoption.

IMAGE 1: Figure 1 & Table 1

IMAGE 2: Table 2