Management of cirrhosis is resource-intensive and disproportionately contributes a growing burden on healthcare. Inpatient care is a sizeable portion of this burden where nearly 30% of admissions result in a readmission within 30 days. Unfortunately, health system-based interventions successful in reducing readmission rates face important barriers to dissemination. In order for successful health delivery redesign to occur, it is important to target the right patient and deliver a tailored intervention. Precisely segmenting patient populations to identify high utilizers is an important first step. Current readmission prediction models based on traditional medical records data have weak performance in cirrhosis. Instead, our novel preliminary data correlate patient reported outcome measures (PROMs) to future healthcare utilization (HCU). Further, the PI?s mentor has shown that when healthcare systems combine real-time PRO tracking with evidence-based management algorithms and patient-facing health tools, HCU burden can be reduced. Based on these data, this proposal will first test the overarching hypothesis that a combination of EHR-based and non-EHR, patient-centered measures will better identify high utilizers in cirrhosis. Taken a step further, this proposal will also test the hypothesis that successful, scalable models of care can be translated to high-risk cirrhotics through adaptation of a health technology tool. We will test these hypotheses via three aims and a robust training plan.
Specific Aim #1 will assess HCU prediction by existing risk models and then utilize a state-wide data source to further refine risk prediction with liver disease-specific and population health data.
Specific Aim #2 will further calibrate prediction of future HCU using PROMs in a prospective cohort of hospitalized cirrhotics. With the ability to identify a vulnerable group of cirrhotics from SA#s 1-2, Specific Aim #3 will build on the co-mentor?s (Dr. Boustani) success in improving HCU in dementia populations by adapting Brain Care Notes, a mobile phone health application designed to support real-time symptom tracking, care-giver support and engagement to reduce HCU in those with cirrhosis. Further, while completing these aims, the PI will accomplish 3 interdisciplinary training goals: 1) develop advanced biostatistical and big data management and analysis skills; 2) acquire experience in methodologies needed for the study of PROMs; 3) gain expertise in healthcare implementation science research all under the guidance of a robust mentorship team led by national experts in the proposed fields. Successful completion of these aims will support the design of a future R01-level intervention that provides innovative, scalable solutions for the chronic disease management in cirrhosis.
End-stage liver disease places a disproportionately large burden on the healthcare system with certain individuals experiencing frequent repeat hospitalization and emergency room visits. Healthcare reform can improve outcomes for these individuals if targeted to their needs. The studies proposed will lead to accurate prediction of this vulnerable group by using multi-dimensional data and will establish the feasibility of using novel mobile phone program to improve healthcare use in liver disease. The proposed studies will bring the field closer to providing care that is (1) more aligned to patient goals, (2) adheres to current guidelines, (3) equitable and (4) at a lower cost.
