Comparative Effectiveness of a Complex Care Program for High-Cost/High-Need Patients: A Replication Study

Evidence about the patient benefits of alternative primary care delivery models for high-cost/high needs patients is mixed. We conducted a follow-up study of a program designed to improve outcomes of seriously ill adult patients in an integrated delivery system.

Using a quasi-experimental design, we examined the 180-day mortality of patients empaneled to a complex care program (CCP, n=1445) compared with that of eligible but unempaneled patients who continued to receive usual primary care (UPC, n=6409) for January 2019 through June 2021. Patients in the CCP and UPC were propensity score-matched on demographics, comorbidities, and frailty. In the propensity score matched samples (n=1440 in each group), the hazard of mortality was estimated using Cox proportional hazards regression.

The CCP continued to empanel eligible adults with more comorbidities and greater frailty compared with the eligible patient population in UPC. In the matched samples, CCP patients had a significantly lower hazard of 180-day mortality compared with UPC in this replication cohort (0.71, 95% CI: 0.61-0.82). This was higher than the hazard ratio in the prior inception cohort (0.58, 95% CI: 0.47-0.70).

A reduced hazard of death was reproduced within a second incident cohort of among seriously ill adult patients who were empaneled to a CCP in an integrated health care system compared with matched, but unempaneled patients whose care remained within UPC.