In Reply We agree with Dr Brown that rigorous research is needed to inform provider payment policy. Because physicians and health care provider organizations are not randomized to different payment systems, however, researchers must turn to quasi-experimental methods to evaluate innovations in health care markets.
Contrary to Brown’s assertion, not only did we recognize selection bias as a potential limitation of the quasi-experimental design we used, we tested for it. Specifically, we compared preintervention spending trends in Medicare between the intervention group of organizations participating in the AQC and the control group of nonparticipating organizations.
McWilliams JM, Landon BE, Chernew ME. Spending and Quality of Care for Medicare Beneficiaries in Massachusetts—Reply. JAMA. 2013;310(24):2675-2676. doi:10.1001/jama.2013.282654