The feasibility and acceptability of a data science powered personal health library HealthPAL.: Presenter(s): Reed Bratches, Dartmouth College, United States. (April 2023)

• Record Type: Journal Article
• Authors: Barr, Paul; Oh, Lisa; Ganoe, Craig; Tarczewski, Susan; Carpenter-Song, Elizabeth; James O'Malley, A.; Onsando, Wambui; Arend, Roger; Piper, Sheryl; Masel, Meredith; Cavanaugh, Kerri; Das, Amar; Bruce, Martha
• Is Part Of: Patient education and counseling. Volume 109(2023)Supplement
• Page Start: 130
• Publication Date: 2023-04
• Subjects: Patient education -- Periodicals; Health counseling -- Periodicals; Health education -- Periodicals; Counseling -- Periodicals; Patient Education -- Periodicals; Éducation des patients -- Périodiques; Counseling -- Périodiques; Éducation sanitaire -- Périodiques; 615.5071
• DOI: 10.1016/j.pec.2022.10.298
• Languages: English
• ISSNs: 0738-3991

Abstract

Abstract : Background: Digital recordings of clinic visits offer an opportunity to promote patient and family engagement in care. Using advances in natural language processing (NLP), we have developed HealthPAL (Personal Audio Library). HealthPAL is open-access software that highlights key information (e.g., medications) from visit recordings, and links to trustworthy web-based resources. The objective of this trial is to demonstrate the successful use of HealthPAL at home by older adults. Methods: We conducted a three-arm, parallel-group, patient-randomized, blocked, controlled, pilot trial at Dartmouth-Hitchcock. Patients were ?65 years with two or more chronic diseases. Patients visiting primary care were randomized to 1) usual care, 2) simple recording or 3) HealthPAL. We assessed recruitment feasibility and retention. Assessments were completed at baseline and 2 weeks from enrollment. System usability, feasibility, acceptability, and appropriateness of the recording interventions were assessed using validated metrics. While not powered, we also explored potential impact on patient activation, satisfaction, adherence, QOL, interpersonal communication, electronic health literacy, and caregiver preparedness. Results: We enrolled 4 of 4 clinicians approached, exceeded patient enrollment milestones (n=45 of 45 patients), and achieved high retention (91%) and high fidelity to our trial protocol. Patients were 78 years old (SD 5.9) and 65% female. Feasibility, acceptability, and appropriateness of intervention were a median of 4 out of 5, and system usability score was 70.6 for HealthPAL, indicating above-average usability. As expected, no significant difference in outcomes was found, however, there was a non-significant increase in electronic health literacy in the HealthPAL group from baseline (+3.2). Implications: Visit recordings are a highly usable, feasible, and acceptable way of communicating visit information to older adults. The use of NLP to curate visit recordings is an innovative strategy to optimize visit communication and promote greater self-management. A powered trial is needed to determine the impact on patient outcomes.