The intended and unintended consequences of quality improvement interventions for small practices in a community-based electronic health record implementation project

Abstract

Background: Despite the rapid rise in the implementation of electronic health records (EHR), commensurate improvements in health care quality have not been consistently observed.

Objectives: To evaluate whether the implementation of EHRs and complementary interventions-including clinical decision support, technical assistance, and financial incentives-improved quality of care.

Research design: The study included 143 practices that implemented EHRs as part of the Primary Care Information Project-a long-standing community-based EHR implementation initiative. A total of 71 practices were randomized to receive financial incentives and quality feedback and 72 were randomized to feedback alone. All practices received technical assistance and had clinical decision support in their EHR. Using data from 2009 to 2011, we estimated measure-level fixed effects models to evaluate the association between exposure to clinical decision support, technical assistance, financial incentives, and quality of care. Associations were estimated separately for 4 cardiovascular measures that were rewarded by the financial incentive program and 4 measures that were not rewarded by incentives.

Results: Financial incentives for quality were consistently associated with higher performance for the incentivized measures [+10.1 percentage points at 18 mo of exposure (approximately +22%), P<0.05] and lower performance for the unincentivized measures [-8.3 percentage points at 12 mo of exposure (approximately -20%), P<0.05]. Technical assistance was associated with higher quality for the unincentivized measures, but not for the incentivized measures.

Conclusions: Technical assistance and financial incentives-alongside EHR implementation-can improve quality of care. Financial incentives for quality may not result in similar improvements for incentivized and unincentivized measures.