Customize your JAMA Network experience by selecting one or more topics from the list below.
Stephen J.LurieMD, PhD, Contributing EditorIndividualAuthorPhil B.FontanarosaMD, Executive Deputy EditorIndividualAuthor
To the Editor: Dr Erickson and colleagues1 concluded that patients with managed care insurance
were less likely than those with fee-for-service (FFS) insurance to undergo
coronary artery bypass grafting (CABG) at hospitals with lower mortality rates.
However, the data in Table 1 of their article provide evidence that CABG inpatient
mortality rates among patients enrolled in managed care plans are equal to
or better than mortality rates among those in non–managed care plans.
The authors minimize this central finding because they focus on the likelihood
of using "lower-mortality hospitals," a very difficult metric to interpret.
Hospital use does not necessarily represent insurer contracting decisions.
Indeed, Erickson et al did not determine which hospitals managed care organizations
did or did not typically contract with. Rather, they contrasted hospital choices
of physicians for very dissimilar patient populations. Given the lower mortality
rates among individuals in managed care organizations compared with those
with FFS insurance, it is apparent that the hospitals used for CABG were very
effective for managed care patients but not as effective for patients with
A survey of more than 2000 consumers conducted by the Kaiser Family
Foundation and the Agency for Health Care Policy and Research found that a
majority of consumers (72%) would choose a hospital that is familiar to them
over a hospital rated higher in quality by experts.2
The same survey concluded that an individual's regular physician has the most
influence over the patients' choice of hospital. Patients choose local practitioners
for convenience, and those practitioners have privileges at local hospitals.
Erickson et al did not consider several confounding variables for the
classification of hospitals as high or low mortality. However, the demographic
compositions of managed care and non–managed care populations were quite
different based on the criteria used in the model. It is likely that additional
confounding variables were overlooked. For example, Erickson et al acknowledge
that patients in the study were more likely to choose a hospital they classified
as a "higher-mortality hospital." In addition, the methods used to control
for travel distance to hospitals were inadequate.
Ensor JE, Grana JR, Diamond CC. Managed Care Insurance and Use of Higher-Mortality Hospitals. JAMA. 2000;284(7):829–831. doi:10.1001/jama.284.7.827
Coronavirus Resource Center
Create a personal account or sign in to: