January 19th, 2010

Do Rapid Response Teams and Remote ICU Monitoring Actually Prevent Deaths?

Hospitals devote a lot of resources to preventing in-hospital cardiac arrests. That makes sense because when one actually occurs, the patient has only a 1 in 6 chance of surviving to discharge. Two popular strategies for helping clinicians recognize and treat clinical deterioration before an in-hospital cardiac arrest are remote ICU monitoring and rapid response teams. After all, it’s intuitively appealing to pay close attention to patients who are doing poorly and, when they deteriorate, to intervene quickly and change their course. Yet, evidence regarding the effectiveness of these two strategies has been mixed.
Recently, my colleagues and I did a meta-analysis of studies about rapid response teams. After analyzing data from 1.3 million admissions, we found that the use of such teams was not associated with lower in-hospital mortality in adults. Similarly, two recent multicenter studies, one published in JAMA and the other in Critical Care Medicine, showed that remote ICU monitoring did not reduce ICU or in-hospital mortality rates, or length of stay.
These findings are especially disappointing because of the vast financial and human resources required to develop and maintain rapid-response-team and remote-ICU-monitoring programs. But I wonder, given the evidence, how much hospitals are accomplishing with these strategies. More important, what if institutions get complacent after they’ve implemented them?
I think we need more data to resolve such questions. But I’d like to ask you: What’s your hospital’s experience with rapid response teams and remote ICU monitoring? Do you think hospitals should continue them in the face of disappointing survival data? Can other “outcomes” touted by these programs, such as nursing satisfaction, be achieved at lower costs? And if your hospital has managed to reduce cardiac arrest rates or improve survival after cardiac arrest using other types of quality-improvement initiatives, please share them with us here on CardioExchange.

7 Responses to “Do Rapid Response Teams and Remote ICU Monitoring Actually Prevent Deaths?”

  1. The results on RRT in terms of reduction of in hospital mortality do not surprise me. My observation is that in our institution RRT is mainly used to help transport the patient to a higher level of care bed (step down, ICU) and provide continuous monitoring of vital signs while this is being arranged, in a seting of minor clinical instability such as worsenig respiratory status without need for intubation or atrial fibrillation with RVR. I believe it has its role but more as system improvement rather than direct impact on mortality. I think patient/nursing satisfaction would be an interesting outcome measure to analyze.

  2. There is literature to support that RRTs improve nursing satisfaction. Nurses are the ones on the front line and feel these teams are a potential resource when their patients become ill. However, the Institute for Healthcare Improvement has promoted RRTs because it is believed that they reduce hospital deaths. Given the limited personnel and financial resources that hospitals have for quality improvement, I wonder if efforts to develop RRTs are sufficiently justified or if these efforts could be channeled toward other hospital quality improvement programs with proven benefits. My sense is that improving nursing satisfaction without demonstrated improvements in patient outcomes is not enough.

  3. Paul. Your thorough analysis showing that RRTs do not improve outcome is a reminder that programs or initiatives that “make sense” need to be studied and proven before they are broadly implemented (and become a peformance measure). As you pointed out, the Institute for Healthcare Improvement’s 100 000 Lives Campaign5 recommended that hospitals implement RRTs to reduce preventable in-hospital deaths. Likewise, routine preoperative beta blocker administration has been used as a peformance measure, prior to recent studies showing it is deleterious in low risk individuals.

  4. I agree with you, Richard, that the push for quality improvement has sometimes gotten ahead of itself. There has been a tremendous amount of enthusiasm among hospitals and health quality organizations (and rightly so) to critically examine the quality of care delivered to their patients. Over the past decade, the Institute for Healthcare Improvement has helped to sharpen our national focus on gaps in patient care. However, the solutions for these gaps in quality are not always based on robust evidence, such as in the case of rapid response teams. Then there is the Leapfrog Group’s recognition of hospitals which provide round-the-clock intensivist coverage, regardless of whether this care is provided in-house or via remote ICU monitoring. While there is evidence to support the role of in-house intensivists, the evidence for remote ICU monitoring is less clear. The initial studies were problematic, as they were led by investigators with patents for the technology. Over the past month, we have 2 multi-centered studies which found no benefit with this expensive technology, either in the form of reduced lengths of stay or improved survival. In the meanwhile, hospitals pay several million dollars up front simply to acquire the technology to set up remote ICU monitoring. I think there is a need to take a step back and critically examine whether proposed solutions for improving care are indeed rooted in solid evidence and to not let the our enthusiasm for promoting quality get in the way. Otherwise, at best, we are diverting useful resources to things that may not work.

  5. Without a doubt, this is a very good study that you have completed as well as being important and timely. In terms of Dr Lange’s comments about initiatives that “make sense” in this area, I often refer to a CHEST paper from 2002 (Snider et al CHEST August 2002 vol. 122 no. 2 517-523) which failed to demonstrate a benefit from the use of telemetry in patients admitted with atypical chest pain and normal ECG. However, despite that literature this technology which, if we recall, was initially developed to monitor the heart beats of astronauts in space, is still being used nationwide at incredible expense with limited benefit.

  6. DAM studies

    Paul. I’m interested in know if St Luke’s Hospital (University of Missouri–Kansas City) has a rapid response team. John….thanks for the Chest reference regarding futility of telemetry monitoring in patients admitted with atypical chest pain. As you point out, once a hospital embarks on an “obviously beneficial program”, it is loathe to discontinue it, even when emerging data shows it is not efficacious. Results that are counter to our preconceived notions are usually filed in the “DAM” (“Doesn’t Apply to Me’) folder

  7. John–you bring up a good point about other uses of medical technology. I wonder if care providers in hospitals continue to admit patients with atypical CP with normal ECGs to telemetry wards because: 1) the reimbursement for telemetry ward admissions may be higher, especially if the admitting diagnosis is coded differently, 2) history-taking is often inadequate in this era when we rely more on biomarkers and stress imaging than talking with our patients (so we automatically admit patients to telemetry), and/or 3) the medical-legal climate.

    Richard–it is funny you ask, but yes, Saint Luke’s does have a rapid response team. In fact, we recently published (JAMA;2008;300:2506-2513) our hospital’s results, in which we found no effect on hospital mortality. Nevertheless, even after discussions with our hospital’s QI committee members, they are convinced that they work and our RRT remains in place. This is a classic example of doing what one thinks should work, rather than going with the evidence base–in this case, with our own hospital’s data (so the DAM [Doesn’t Apply to Me] argument doesn’t even work here).

    Having watched RRTs in action at various hospitals, I am not surprised at the lack of survival benefit. While the criteria to identify patients with evidence of clinical deterioration are rather similar across studies, the implementation of these criteria is not consistent and many patients may not be identified early enough to potentially alter their clinical course. Moreover, there is no standardized approach to treating and triaging patients, so even if one hospital’s RRT was found to be truly successful in improving survival, there is no way of easily replicating that hospital’s approach.