Replicating Behavioural Insights in Health: A Quasi-Experimental Phase 2 Trial of integrating descriptive social norms and institutional cost in SMS Reminders to Reduce Missed Hospital Appointments

Missed hospital appointments (DNAs) undermine healthcare efficiency and access. A high-profile study by Hallsworth et al. (2015) found that adding descriptive social norms (DSN) or specific institutional cost (SIC) messages to SMS reminders could substantially reduce DNAs. This prompts optimism that integrating behavioural insights, besides reminders themselves, offers a cost-effective approach to mitigate DNAs. However, subsequent similar interventions have reported heterogeneous findings, echoing broader debates on recent meta-analyses about how to evaluate such findings. We address this issue by framing Behavioural Insights as Applied Science (BIAS), which structures validation in three phases inspired by clinical research. We then treat Hallsworth et al. as a Phase 1 proof of concept and conduct a Phase 2 replication under comparable operational conditions in a quasi-experimental, time-blocked field trial at South-western Jutland Hospital (20,867 appointments) across Cardiology, Endocrinology, and Pulmonology. Patients received SMS reminders rotating every two months between a standard message, DSN framing, or SIC framing. Neither DSN nor SIC reduced DNAs overall. SIC increased cancellations (OR = 1.41, p < .001) but not DNAs; DSN reduced DNAs in Cardiology (OR = 0.76, p = .027), while SIC increased DNAs in Endocrinology (OR = 1.31, p = .021). Our findings underscore the importance of applying a systematic approach in the evaluation of Behavioural Insights.