Health Policy and Health Services Research
Background: Provider efforts to limit overuse of low value healthcare are often complicated by patient concerns that they are being denied necessary care. Improved communication may help mitigate this concern. Sometimes, providers communicate with patients using probabilistic messages that emphasize risks and benefits. Alternatively, narrative approaches use patient stories to guide choices and have been shown to improve health communication. This strategy, however, has not yet been applied to communication about low value care. The goal of this study is to examine whether narrative approaches increase patient satisfaction with the decision to forgo low value care in the emergency department (ED).
Method: A pilot RCT was conducted at two tertiary care hospitals to compare narrative, probabilistic, and standard discharge instructions (the control group). Participants were enrolled at discharge (n=91) and included ED patients with 1) upper respiratory infections who were appropriately not treated with antibiotics or 2) low risk ankle injuries who were appropriately diagnosed without imaging. Each treatment group watched a two-minute video that used narrative or probabilistic elements to discuss decisions about low value care for the patients’ presenting diagnosis. Both video groups and the control group were compared across three outcomes using one-way ANOVA and χ2 tests: 1) patient satisfaction 2) physician trust, and 3) likelihood of returning after discharge.
Results: Overall, participants were satisfied with their care (median satisfaction score of 21 out of 25, IQR=7), had high physician trust scores (47 out of 50, IQR=7.25), and were unlikely to return for the presenting complaint (11.3%). The narrative arm had lower physician trust than the control group (42.75 vs 47.50, p=.047). There were no other differences in outcomes across the three arms.
Conclusion: The narrative video did not appear to impact satisfaction or rate of return of visits. This may be due to an ineffective intervention or because we were underpowered to detect small effects. The decrease in physician trust in the narrative group compared to the control group suggests a possible unintended consequence of the video intervention. The video may have called attention to the decision to forgo low value care, instead of helping to communicate the decision more effectively to the patient.