The trusted source for
healthcare information and
Do Rapid Response Teams Reduce Hospital Mortality or Simply Increase Costs?
Abstract & Commentary
By Michael Young, MD, Associate Professor of Medicine, Fletcher Allen Health Care/University of Vermont, Division of Pulmonary and Critical Care, The University of Vermont College of Medicine. Dr. Young reports no financial relationship to this field of study.This article originally appeared in the August 2010 issue of Critical Care Medicine. It was edited by David J. Pierson, MD, and peer reviewed by Saadia Akhtar, MD.
Synopsis: This meta-analysis casts serious doubt on the ability of rapid response teams to significantly reduce hospital mortality.
Source: Chan P, et al. Rapid response teams: A systematic review and meta-analysis. Arch Intern Med 2010;170:18-26.
In the past decade, rapid response teams (RRTs) were broadly implemented to identify and treat patients on medical and surgical wards at risk for catastrophic deterioration and thus prevent death. The impetus to form RRTs came after multiple observational studies in the 1990s suggested that ward patients who experience a marked clinical deterioration often suffer excess morbidity and mortality from delayed treatment and under-resuscitation. A number of single-center studies indicated that RRT implementation was associated with reductions in cardiopulmonary arrests outside the ICU and decreases in hospital mortality. However, two meta-analyses on RRTs published in 2007 were far more cautionary on their purported benefits. Despite equivocal evidence favoring RRTs, RRT implementation was an integral part of the "Save 100,000 Lives Campaign" lead by the Institute of Healthcare Improvement (IHI), and in 2008 the Joint Commission listed implementation of a RRT (or equivalent) as a National Patient Safety Goal.
The present meta-analysis by Chan and colleagues examines the effect of RRT implementation on cardiopulmonary arrest rates and hospital mortality. The authors' search strategy yielded 143 potentially relevant articles on the basis of title and abstract. Additional articles were excluded because they were simply review articles, lacked a control group, failed to evaluate mortality or CPR rates, were duplicates of other articles, or provided insufficient data. Eighteen articles, including 13 adult and five pediatric studies, published between 2000 and 2009, were included in the meta-analysis. Of the 6 studies classified as "high-quality," 2 were randomized clinical trials and 4 were observational in design.
Some of the adult RRT studies included in the meta-analysis showed a reduction in cardiopulmonary arrest rates while others did not. Cumulatively, the adult high-quality studies identified a 21% decrease in non-ICU cardiopulmonary arrest rate compared to a 48% decrease in the other studies. In contrast, the pooled mortality rate for adult patients was unaffected by implementation of RRTs in both high- and low-quality studies, and in studies with high- vs low-RRT activation rates. In the pediatric population, there was a lower mortality rate (relative risk [RR], 0.79; 95% confidence interval [CI], 0.63-0.98) with RRT implementation, but the improvement was not robust. Furthermore, when the adult and pediatric results were pooled, mortality rate with RRT implementation did not change (RR, 0.92; 95% CI, 0.82-1.04).
Among the five studies that reported a lower mortality rate with RRT implementation, Chan et al measured whether the reduction in cardiopulmonary arrest rate could explain the improvement. Their analysis found that the observed reduction in cardiopulmonary arrest rates could explain < 1% to as much as 61% of the mortality reduction.
This study is now part of growing series of meta-analyses indicating that the benefit of RRTs may be more modest than widely believed. If RRTs' benefits are measured largely by their impact on hospital mortality, the enthusiasm for RRT formation by the Joint Commission and IHI is not easily supportable. On the other hand, the data do indicate that RRTs reduce the number of cardiopulmonary arrests outside the ICU among both adult and pediatric patients. Given the discouraging mortality and morbidity associated with in-hospital cardiopulmonary arrest, this is an important benefit. There may be other benefits that RRT implementation provides that have not been well quantified. Nurse and patient/family satisfaction are reported to be improved with RRT implementation. Clinician education may be improved and delays in transfer to the ICU from the ward may be decreased. More research in these areas is needed.
Perhaps equally important, this study reminds us how difficult it can be to interpret the results of single-center studies, especially those that use a "before-and-after" design. RRTs may be another example of where we, as clinicians, adopt therapies or interventions that simply make "sense" to us without waiting for well-done outcomes studies. Once rigorously studied, we may find the story more complex than we suspected and the therapy less helpful than we anticipated and sometimes even more harmful than helpful for example, lidocaine post-myocardial infarction, aggressive use of total parenteral nutrition in the ICU, and routine use of pulmonary artery catheters among patients with acute lung injury and the ARDS.
Why the remarkable variation in outcomes seen in the before-and-after RRT implementation studies? There are probably a number of factors at play. RRT composition is widely variable. The clinical thresholds used to prompt RRT activation vary. What are the sensitivities and specificities of those clinical thresholds' ability to predict catastrophic deterioration? In about 50% of cases, RRT are activated because of a "worried clinician." When should clinicians become worried? What interventions should the RRT provide when the RRT arrives at the bedside of a patient suffering measurable physiologic or subjective deterioration? What is the right rate of RRT activation? Once activated, how long should the RRT stay at the patient's bedside? When and which patients should be transferred from the ward to the ICU? How often should the vital signs of patients on medical and surgical wards be measured and reported? How accurate are those measurements?
Should RRTs be disbanded? In my view, no. However, until we conduct serious research to find answers to these questions, the real potential of RRTs to prevent morbidity and mortality among patients on hospital wards will remain unknown.