Can Patient-Centered Care Reduce the Overall Cost of Care?

Background/Purpose: Patient-centered primary care emphasizes the principles of comprehensive, personalized, high-quality primary care.  Despite growing interests in implementing more patient-centered care relative to traditional models of care delivery, less is known about the cost implications of transforming individual processes of care to be more patient-centered.  Our study investigates the relationship between a patient’s use of healthcare resources and patient-centered processes of care.  Our evaluation focuses on the effects associated with four specific processes of care: continuity of care, patient-physician communication, medical team communication, and time to appointment.

Methods: We perform a retrospective analysis of managed care patients in a large multi-specialty group practice.  We examine annual resource use, measured in total RVUs, for 115,902 patient-years (86,858 unique patients) in 2009 and 2010.  We compare patients’ utilization (both overall and by department) under different levels of patient-centeredness of care using separate measures for each of the four processes of care.  The comparison is made using multi-variable statistical methods that control for patient risk using demographics, comorbidities, and use of pharmaceuticals.

Results: Three important effects of patient-centered care are identified.  First, patients with the highest continuity of care (100% visits to own PCP) used 2.0% less primary care and 5.1% less specialty care versus patients with moderate continuity of care (50% visits to own PCP).  Reductions in utilization were offset by a 1.6% increase in laboratory services and a 3.0% increase in imaging services, such that there was not a statistically significant net change in utilization across all departments once these are accounted for.  Second, messaging using EHR, both via patient-PCP and PCP-medical team, was associated with an increase in utilization across all departments.  One additional message and reply between a PCP and patient generated approximately 2.8% additional total RVUs per patient-year, while one additional message and reply between PCP and staff generated 7.7% more total RVUs.  Third, patients facing a longer time to appointment with their own PCPs had slightly lower utilization; an additional wait of one day reduced overall utilization by approximately 1.7%, mostly from a 2.2% reduction in ED use.

Conclusion:  Our study is one of the first to simultaneously investigate the individual effects of different dimensions of patient-centered primary care.  The changes in resource use we observe in conjunction with these processes demonstrate a number of policy relevant concerns that future primary care transformations might address.  First, our results suggest that declines in utilization arising from patient-centered care may not materialize immediately.  While some of our estimates show small increases in utilization, these increases occur within a one-year window and they are consistent with high-quality care, which may produce improvements in utilization over a longer term.  Second, our analysis by department shows that changes in utilization across departments may be linked.  Understanding patterns of substitution across departments will be important in designing future primary care transformations.