Ben Rosner Receives UCSF RAP Grant
Congrats to Ben Rosner for receiving the UCSF RAP grant for a project entitled: "Leveraging Patient Generated Health Data to Close the Hospital Readmission Information Gap."
Patient generated health data (PGHD) are central to our national health information technology roadmap,1 and their use in clinical care is expected to improve health outcomes and reduce costs. Knowledge gaps about the accuracy and completeness of PGHD, however, remain critical barriers to their adoption in clinical decision making. Addressing these gaps for patient-reported readmissions is a compelling use case, both methodologically (PGHD for readmissions have objective external sources of validity, whereas subjective PGHD do not), and from an outcomes perspective (readmissions are costly and actionable). It is estimated, for example, that 34% of Medicare patients are readmitted within 90 days of discharge, and that 31%-65% of readmissions occur at secondary facilities unbeknownst to the index institutions,6,7 resulting in $17B-$25B in annual costs.8,9 Under value-based reimbursement, index hospitals are bearing new financial risk, incurring penalties in some cases for all excess cost-drivers in the 90-days post-discharge, wherever they occur. This has driven intense interest in identifying any readmissions, and designing interventions prior to discharge to mitigate them. One academic medical center, for example, was able to identify previously unrecognized readmission “hot spots,” re-design its processes pre-discharge, and reduce its readmission rate by 25% within one year of implementing a claims-based near-real-time readmission feedback system. Such closed loop feedback is central to organizational learning and quality improvement, but healthcare has been slow to adopt such approaches post-discharge. PGHD-based sources of near-real-time readmission feedback are promising means to close this knowledge gap, but their accuracy and completeness are not known.
In this study, we will assess the extent to which: 1. Patients are active, accurate, and complete self-reporters of 90-day readmissions through post-discharge electronic surveys, and 2. Smartphone-based geofencing, a location-based technology that detects when the phone crosses a hospital boundary, offers an accurate and acceptable passive means to supplement any gaps in active patient self-reporting. The central hypothesis is that active patient self-report offers accurate feedback on readmissions for a portion of patients, but that a complete picture on readmissions may only be possible when passive sensing technology fills the remaining gap. The objective of this proposal is to assess the accuracy and completeness of two novel sources of post-discharge readmission PGHD: 1. The patient him/herself, and 2. Smartphone-based geofencing technology. The REDCap survey platform and UCSF’s Eureka Research Platform will be used to gather active (survey based) and passive (geofencing-based) data for adult patients following UCSF hospital discharge respectively. Subgroup analyses based on demographics and disease burden will be conducted to understand potential limitations of these approaches and opportunities for future implementation.