To the Editor.
—Given the complexity of the health care system and its interaction with patients, studies such as the MOS are prone to methodological weaknesses. Dr Greenfield and colleagues1 candidly identify a number of significant flaws. However, the authors fail to identify one additional source of variance that may further confound their findings.Both the physician and patient groups were assigned to systems of care (eg, HMO, independent practice association [IPA], or fee for service [FFS]) as of the mid to late 1980s.2 However, as managed care has matured in each of the three study cities, a significant number of physicians and their patients are likely to have moved from one system to another during the 7 years for which outcome data were collected. In addition, the MOS predates the arrival of point-of-service plans, which would cloud the assignment of patients to any system of care. Finally,
Kastner T. The Medical Outcomes Study. JAMA. 1996;275(14):1084. doi:10.1001/jama.1996.03530380025017