Rapid response teams reduce preventable in-hospital deaths-some of the time. Research hasn't conclusively shown that they're effective in lowering cardiac arrest and mortality rates outside the ICU. Now a large prospective cohort study conducted by Chan and colleagues at a tertiary hospital in Kansas City, Missouri, has found that they don't improve rates of cardiac arrest or death hospital-wide-that is, both inside and outside the ICU.

A rapid response team consists of ICU personnel who can be summoned to assess and treat any patient outside the ICU who shows signs of deterioration and who may be at risk for cardiac arrest or death. Team makeup varies but often includes one or more ICU nurses, a respiratory therapist, and a physician who can be called upon when needed.

Since the Institute for Healthcare Improvement (IHI) made the implementation of rapid response teams one of the six strategies in its 100,000 Lives Campaign in 2004 (and carried over as one of 12 interventions in its more ambitious 5 Million Lives Campaign in 2006) to reduce preventable deaths in hospitals, initiatives promoting the use of such teams have been taken up by major foundations and accrediting agencies, as well as hundreds of hospitals.

But little is known about the cost-effectiveness of rapid response teams, and until now their effectiveness had only been studied outside the ICU-not in the hospital as a whole. Concerned that "a primary action of rapid response teams is to transfer patients to the ICU," and that cardiac arrests in the ICU were not included in recent studies, Chan and colleagues examined hospital-wide cardiac arrest and mortality rates before and after the implementation of a long-term (20-month) rapid response team.

The team in the study consisted of two ICU nurses, a respiratory therapist, and an ICU physician. During the 20 months of the study, the team was called on 376 times. Cardiac arrest rates were 11.2 per 1,000 admissions before implementation of the rapid response team and 7.5 afterward; after adjusting for variables, the reduction was not significant. Nor did unadjusted hospital-wide mortality rates "meaningfully change" after the intervention, with rates of 3.22 per 100 admissions before the intervention and 3.09 after the intervention.

Should support of rapid response teams be reevaluated? IHI vice president Joe McCannon says that although the organization is "always interested in new evidence," the IHI still believes that "rapid response teams and other early detection strategies are important" for addressing this problem.

The study authors suggest that secondary effects of having rapid response teams may be as important as the primary outcomes they measured. Noting that "a sizable number of patients who survived their initial rapid response team intervention subsequently obtained [do not resuscitate] status during their hospital stay," the authors speculate that "rapid response teams may not be decreasing [cardiac arrest] rates as much as catalyzing a compassionate dialogue of end-of-life care among terminally ill patients." The authors also call for more research into the best composition of rapid response teams, their most appropriate use in hospitals, and the "optimal triggers for rapid response team activation."

Jacob Molyneux, senior editor

Chan PS, et al. JAMA 2008;300(21); 2506-13.