Riders on the ACO bandwagon might want to take a close look at results just published in the Physician Executive Journal (PEJ March-April/2014), which suggest that clinically integrated delivery systems aren’t always better or less expensive. Based upon a sample of 273,000 enrollees of self-funded plans in 52 Midwestern medical group practices, researchers found that integrated delivery systems which had “extensive administrative and organizational capacity” were outperformed by smaller physician owned practices.  The study also found that in some areas, cost was negatively correlated with quality.

The researchers separated the observed practice models into two categories: one category included practices that were owned by not-for-profit integrated delivery systems comprised of at least one hospital, primary care and specialty clinics, and extensive support services; the other category included practices that were physician owned primary care practices, some with one or two additional specialty services.

Cost data, which was adjusted for case mix using the 3M Clinical Risk Group software, was based upon data maintained by an Upper-Midwestern consortium of self-insured employers which provides such data to its clients.  Quality measures were based upon indicators for diabetes care, cervical cancer screening rates, high blood pressure control, avoidable hospitalizations and inappropriate use of the emergency room.

The findings revealed that the integrated delivery systems model had higher costs per-member-per-year (PMPM) than physician owned models, and, multispecialty practices were associated with higher costs compared to single specialty primary care practices.

The researchers also found that the correlation between cost and quality was mixed. In some instances, quality, as measured by diabetes management and high blood pressure control indicators, was better in the less costly practice models.  On the other hand, it appeared that quality, as measured by avoidable hospitalization rates and inappropriate ER utilization rates, was better with respect to the more costly practice models.

ACOs and integrated delivery systems are expensive to create and operate.  The promise of those models is that they will be capable of bending the cost curve.  These data show that in some instance that may indeed be possible.  The question is, will they bend it in the right direction.