Patient Use of a Comment Function in Electronic Health Records: Mixed Methods Case Study.

BACKGROUND: Patient access to their electronic health records (EHRs) could improve care and reduce errors by enabling patient contribution to documentation. However, interactive functions that support such engagement remain underresearched. OBJECTIVE: The purpose of this study was to analyze patients' use of the free-text comment function in the Swedish national patient-accessible EHR. METHODS: Comments made by patients between 2015 and 2021 for health care professionals to view were retrieved from the patient-accessible EHR administration service in Sweden. The number and length of comments were analyzed using descriptive statistics and chi-square tests, while the content was examined qualitatively using inductive content analysis. In addition, comments were coded for sentiment (positive, neutral, or negative) and note content, and users were classified according to whether they had submitted at least 1 comment referring to mental health care. RESULTS: A total of 1670 comments were left by 659 patients, of which 15 had irrelevant content (15/1670, 0.9%; 95% CI 0.4%-1.4%), leaving 1655 comments for analysis. Most comments (948/1655, 57.3%; 95% CI 54.9%-59.7%) were related to matters other than note content, and nearly all had a neutral (920/1655, 55.6%; 95% CI 53.2%-58.0%) or negative (702/1655, 42.4%; 95% CI 40.0%-44.8%) sentiment. Patients who submitted a comment referencing mental health care tended to leave more comments overall (median 7 vs 2 comments per patient), and their comments were more frequently negative than comments written by other patients (313/495, 63.2% vs 389/1160, 33.5%; P<.001). The most prevalent categories were care engagement, perceived inaccuracies, complaints, and requests. CONCLUSIONS: Patient engagement with clinical notes has the potential to enhance documentation accuracy and strengthen patient-clinician communication. To realize these benefits safely, such features must be deliberately designed and clearly explained to users to avoid patient safety risks and unnecessary burdens on health care professionals.