Impact of quality-of-care factors on pediatric intensive care unit mortality

Objective. —To determine the importance of the following care factors previously associated with hospital quality on survival from pediatric intensive care: size of the intensive care unit (ICU), medical school teaching status of the hospital housing the ICU, specialist status (pediatric intensivist), and unit coordination. Design. —After a national survey, consecutive case series were collected at 16 sites randomly selected to represent unique combinations of quality-of-care factors. Setting. —Pediatric ICUs. Patients. —Consecutive admissions to each site. Main Outcome Measure. —Patient mortality adjusted for physiologic status, diagnosis, and other mortality risk factors. Results. —There were 5415 pediatric ICU admissions and 248 ICU deaths. The ICUs differed significantly with respect to descriptive variables, including mortality (range, 2.2% to 16.4%). Analysis of risk-adjusted mortality indicated that the hospital teaching status and the presence of a pediatric intensivist were significantly associated with a patient's chance of survival. The probability of patient survival after hospitalization in an ICU located in a teaching hospital was decreased (relative odds of dying, 1.79; 95% confidence interval [Cl], 1.23 to 2.61;P=.002). In contrast, the probability of patient survival after hospitalization in an ICU with a pediatric intensivist was improved (relative odds of dying, 0.65; 95% Cl, 0.44 to 0.95;P=.027). Post hoc analysis indicated that the higher severity-adjusted mortality in teaching hospitals may be explained by the presence of residents caring for ICU patients. Conclusion. —Characteristics indicative of the best overall hospital quality may not be associated, or may be negatively associated, with quality of care in specialized care areas, including the pediatric ICU. (JAMA. 1994;272:941-946)