Using a Self-Reported Global Health Measure to Identify Patients at High Risk for Future Healthcare Utilization.

Abstract

Research studies have shown that patient-reported outcome measures (PROMs) that assess global health are helpful for predicting health care utilization, but less evidence exists that collection of PROMs in routine care can identify patients with high health care needs. To investigate the association between the PROMIS Global Health (PGH) scores and subsequent health care utilization among patients in a large accountable care organization (ACO). Retrospective cohort study of individuals in the Partners HealthCare ACO who completed at least one PGH during a primary care visit. A total of 2639 individuals who completed at least one PGH and who also had 12months of ACO membership and/or claims data prior to the PROM completion and at least one month of claims data post-PGH completion. The main outcomes were the rates of emergency department (ED) visits and hospitalizations by quartile of PGH physical and mental health scores. We also compared the predictive accuracy of administrative data models with and without the PGH scores to identify the highest utilizers. The group with the worst (lowest) physical and mental health scores had significantly higher rates of hospitalization (RR 5.14, 95% CI 2.37, 11.15; and 2.27, 95% CI 1.06, 4.85, respectively) than those with higher scores. After adjustment for demographic and clinical factors, only the group with lower physical health scores had higher rates of hospitalization (RR 3.15, 95% CI 1.30, 7.90). The addition of the physical health subscore to administrative data increased the sensitivity to detect the top 5% of hospital utilizers compared with administrative data alone (44.0% vs. 36.0% respectively). Worse self-reported physical health, measured during routine primary care, is associated with significantly higher rates of hospitalization. It is not associated with increased rates of ED visits. Self-reported physical health modestly increases the sensitivity to detect the highest hospital utilizers.