I want to spend a moment highlighting news that got surprisingly little attention in my favorite health policy media earlier this month.  Dr. Paul Chan and colleagues published an article in the December 3, 2008 edition of JAMA titled "Hospital-wide Code Rates and Mortality Before and After Implementation of a Rapid Response Team."  It reaches a conclusion about Rapid Response Teams (RRTs) that might be surprising to many in the quality community.

The Chan article describes two 20-month periods, one before and one after RRTs were deployed at Saint Luke's Hospital, a 400-bed academic medical center in Kansas City.  Although cardiac arrests, cardiac arrest mortality, and in-hospital mortality all declined, none of the changes were statically significant. 

RRTs have been implemented in scores of hospitals in recent years, due in no small part to the work of the Institute for Healthcare Improvement (IHI), which designated RRTs as one of six strategies to prevent in-hospital deaths as part of their 100,000 Lives Campaign.  (Curiously, even the IHI authors found the evidence for RRTs inconclusive in the previous hyperlink.)   IHI posted a podcast interview response to the Chan article here, but I haven't been able to get it to work. 

The Saint Luke's intervention didn't hurt things - it didn't increase mortality - but it brings up the question of resource allocation.  Dr. Chan argues that widespread RRT implementation should be slowed because of their cost and questionable benefit. 

As Drs. Winters, Pham, and Pronovost wrote two years ago in the same journal, the literature on RRTs have limited, mixed results: "Overall, the observational studies (80% of the RRT literature) tend toward demonstrating a benefit with RRT programs, but there is significant heterogeneity in these studies."  They go on to state that using hospitalists, nurse practitioners, physician assistants, or higher nurse staffing ratios on hospital units - or perhaps the use of automated monitoring systems - might be better ways to treat deteriorating patients early and improve patient outcomes.

The authors of this earlier literature are not trouble-makers trying to cast doubt on new care techniques; they are simply trying to ensure clinicians practice evidence-based medicine.   You might recognize Dr. Pronovost's name from the outstanding Atul Gawande New Yorker article on "the checklist."  Even though RRTs make intuitive sense, if the evidence indicates that they are no better than other less expensive techniques, it would be more appropriate to implement these more cost-effective strategies. 

On the other hand, it is always possible to over-rely on the literature.  Just ask the female patient from the Baylor Health Care System (BHCS) whose life was saved by an RRT called by her husband - who subsequently starred in her own 15-minute training DVD distributed to BHCS staff.   BHCS has found that RRTs were a contributing factor to their declining inpatient mortality rates.  Last year, the WSJ Health Blog wrote about a children's hospital at Stanford University that found similar results.  (Remember to check back here in late January for the news on the publication of our BHCS case study.)  

Or you could even consult a 2003 article from the British Medical Journal that called into question the effectiveness of parachutes reducing mortality for people jumping out of planes; since no double-blind, randomized, placebo-controlled (no parachute!) trial had ever been conducted, their efficacy is judged using mere observational data; not adequate for true believers of evidence-based medicine.

The bottom line is: more research needs to be done.  Whether or not resources are allocated in a particular hospital to developing RRTs or withheld until there is more evidence is up to the individual institution.  That said, until I know my hospital has an automated monitoring system, I think I'd like an RRT in my hospital.

For more, listen to a brief interview with Dr. Chan here.  You can also read a Reuters story here and a HealthDay story here.