An intensivist all day, keeps the bad outcomes away.

T here has been an increasing awareness that organizational factors can influence patient outcomes in the intensive care unit (ICU). For instance, ICU staffing with a trained intensivist is associated with a reduction in ICU length of stay and in-hospital mortality (1, 2). However, access to intensivist care in the ICU varies, with less availability during evening/night, weekend, and holiday hours (" off hours ") (3). Because initial management may be crucial to the outcome of critically ill patients (4), lower level staffing during off-hour admissions could be associated with a worse outcome. The relationship of off-hour admissions to patient mortality has been examined in a number of studies of medical, surgical, and critically ill patient populations, with conflicting results (5–10). In this issue of Critical Care Medicine, Dr. Luyt and colleagues (11) examine the role of off-hour admissions to the ICU on inpatient mortality in a large retrospective multicenter cohort study. Using prospectively collected data from 23 ICUs in the Paris metropolitan region, the investigators examined 51,643 consecutive ICU admissions during a 4-yr period, 33,857 (66%) of which were admitted during off hours. Off hours were defined as periods outside of legally mandated day-shift hours in France, divided into night shifts (6:30 PM to 8:29 AM the next day, Monday to Friday), weekends (1:00 PM Saturday to 8:29 AM Monday), and holidays (8:30 AM to 8:29 AM the next morning). During day-shift hours, ICUs were staffed with a median of three board-certified intensivists, one intensivist-in-training, and two residents. In contrast, during off hours, all ICUs were staffed by only one on-site board-certified intensiv-ist or an experienced intensivist-in-training, with an additional medical resident in ten of the participating ICUs. The primary analysis evaluated the association of off-hour staffing with in-hospital mortality, using a multivariable logistic regression model adjusting for many potential confounders, including age, co-morbidity score, simplified acute physiology (SAPS) II score, and type of admission (direct or transfer). Many sensitivity analyses were conducted, including examining different definitions of off hours, using a propensity score methodology, and analyzing the results by SAPS II quartiles, specific diagnoses, and individual hospital. ICU and in-hospital mortality for the entire cohort were 18% and 22%, respectively. Patients admitted during the day shift were more ill (with a higher mean SAPS II score), had more organ failures, and had a greater need for supportive measures (e.g., mechanical ventilation, hemo-dialysis) than patients admitted during off hours. …