In a retrospective study, mortality rose with specialized care, but other factors could explain these results.
Past studies (with substantial methodological limitations) have suggested that critical care physician (intensivist) management can lead to better clinical outcomes. To examine this hypothesis, researchers evaluated more than 100,000 patient records from 123 intensive care units at 100 hospitals. Critical care physicians were defined as those who were recognized as such by their institutions, whether or not they were board-certified or fellowship-trained in critical care; they had responsibility for overall patient management. In most units, the choice of whether to involve an intensivist was left to each patient’s admitting physician.
Patients who were managed by intensivists were more severely ill and underwent more procedures. In an effort to account for this source of confounding, mortality analyses were done separately for different levels of illness severity and included the likelihood that a patient would be cared for by an intensivist (based on a technique called a propensity scoring). Intensivist management was associated with higher mortality (odds ratios, 1.18 to 2.83). The largest effects on mortality occurred among the least severely ill patients.
Comment: This study methodology is about as good as it could be with retrospective data. The findings are plausible and might be explained by differences in clinical practices (e.g., more procedures and their attendant complications). But editorialists (both practicing intensivists) note that the findings are in conflict with prior literature, and they express concern that open ICUs with elective intensivist consultation — rather than closed ICUs run by intensivists — were the source of much of these data. Hence, they call for replication of these findings and identification of a mechanism for the effect. I think there is a fair chance that the patients cared for by intensivists were sicker in ways that could not be accounted for in the analyses. In any case, this study should lead to a careful examination of assumptions about the best management approaches for critically ill patients.
— Richard Saitz, MD, MPH, FACP, FASAM
Published in Journal Watch General Medicine June 12, 2008
Levy MM et al. Association between critical care physician management and patient mortality in the intensive care unit. Ann Intern Med 2008 Jun 3; 148:801.