Intensivists overnight in the ICU don't help, if you're already good

Nighttime Intensivist Staffing Does Not Help (Again)

Intensivists perform a vital service for humanity -- just ask us. Numerous studies have concluded that a specialized intensivist's presence in an intensive care unit during the day saves lives and results in better use of health care resources. (One contrarian observational study suggests just the opposite; these people will be asked to submit to loyalty tests.) Physicians being a generally masochistic group, certain agitators thence argued that more intensivist coverage should be even better, and that an intensivist's presence in the ICU 24 hours a day should be the standard of care. It's intuitively persuasive, but the evidence supporting this agenda has been a thin and unsatisfying concoction of a few before/after observational trials (in Lancet 2000, Crit Care Med 2008, and AJRCCM 2012). The best-quality study performed to date and published last year in NEJM suggested that 24/7 in-house intensivist coverage did not improve outcomes in ICUs that were adequately staffed by critical care intensivists during the day (mandatory intensivist consultation or intensivist as primary physician). Although a third of U.S. academic medical centers and most ICUs in Europe have adopted in-house overnight intensivist staffing models, with far more ICU beds per capita in the U.S. than most European countries, there simply aren't enough intensivists to go around here, and most U.S. hospitals have balked at the cost. To justify the expense and expansion of intensivist training program quotas necessary to fully implement 24/7 intensivist coverage in the U.S., a clear benefit to be expected from the change would need to be shown.

What They Did

Meeta Prasad Kerlin et al at the University of Pennsylvania randomized their own 24-bed medical ICU to be staffed by either 24/7 by an in house intensivist, or by residents alone overnight with intensivist attendings during the day. They changed allocation in alternating one-week blocks for one year. Almost 1,600 patients were cared for during this time. The primary outcome was ICU length of stay, with numerous secondary outcomes prespecified. They estimated they had about 90% power to detect a 20% difference in ICU length of stay with a P of 0.05.

What They Found

The presence of an overnight intensivist had no measurable effect on any outcome: length of stay in the ICU or hospital, survival in the ICU or hospital, readmission to the ICU, or discharge home. Those admitted in the middle of the night with the most severe illness (organ failure), examined as a subgroup, also did not do better when an overnight intensivist was present. Essentially, 18% of patients died in the ICU regardless of who admitted them or what time of day or night they came, and 24% of patients died in the hospital, again regardless of these factors on admission. The 45% of residents who answered their follow-up survey felt care was better overnight, and felt they learned more effectively with the attending present than when alone, without having their autonomy compromised.

What It Means

At this top U.S. academic medical center, with likely great residents, attendings, and nurses and support staff, care was probably outstanding around the clock, with good communication and handoffs across shifts. When things are done consistently well by a full team of smart, well-trained and caring people, who among us is proud enough to believe that her mere presence itself at night (as opposed to being a phone call away) would save lives or make the complex process of critical care run measurably more efficiently? Apparently not Dr. Kerlin, who told U-Penn's Penn Medicine,

This tells me that residents and nurses are well qualified and completely competent to handle these patients. As long as nurses and residents have access to an on-call attending physician, then the patient will do as well as if the senior doctor was at their bedside.

This study's single center design and substantial differences between academic and community practice make direct extrapolation to the wider U.S. critical care landsape unadvisable. The control group here still had resident physicians at the bedside throughout the night, and U-Penn is presumably a highly resourced MICU. Still, the bar has been set significantly higher for advocates of 24/7 in-house intensivist coverage to provide evidence, beyond face validity, to justify its additional expense to a cash-strapped U.S. medical care system.

[polldaddy poll=7262135]

[polldaddy poll=7262142]

Meeta Prasad Kerlin et al. A Randomized Trial of Nighttime Physician Staffing in an Intensive Care Unit. N Engl J Med 2013; 368:2201-2209.