A Safety Catch System for Patient Self-Service Appointment Booking

In introducing a small-scale pilot focused on self-service appointment booking, the authors found that automating queries of health records and applying natural language processing (NLP) technology can effectively enable triage processes, particularly for red-flag or high- risk cases. As Kaiser Permanente's care delivery arm, the Southern California Permanente Medical Group serves 4.8 million patients; historically, patients have relied on call center agents for appointment booking, which offers the advantage of real-time transfers to telephonic triage. Unfortunately, self-service booking does not inherently have this capability. Therefore, Southern California Permanente Medical Group implemented a pilot program utilizing a multicomponent virtual safety net (VSN) ecosystem, starting in 2019, to ensure clinical safety in this self-service space. The VSN runs data queries to identify red-flag cases among self-booked appointments. Afterward, nurses review reports from the queries to support telephonic triage. In 2021, Microsoft's Language Understanding Intelligence System was integrated into the VSN to enable more automation, which led to greater efficiency. Since its launch, as of September 2023, the VSN has been associated with 19 lives saved, which are counted among 665 safety catches - instances where VSN-driven action helped patients by facilitating timely diagnosis, treatment, referrals, or hospital admissions. Insights gleaned from this project, including data from the pilot, will help to drive improvements to the VSN and support plans for broadening its scope.