Launching a rapid-response team at Children's Healthcare of Atlanta proved to be easier in concept than in reality, at least initially.

Discussions ebbed and flowed for more than a year, spurred largely by the Institute for Healthcare Improvement's commitment to the concept in which clinicians rapidly converge at the bedside of a patient in crisis. But there was an anemic response from doctors and nurses, in part because it was a concept being pushed nationally that didn't seem clinically relevant to them, says Christiane Levine, R.N., quality manager of surgical services at the nonprofit system, which includes three children's hospitals.

That is, until a series of unexpected patient deteriorations—five during a single month in late 2005—shook up clinicians across the system. An extensive analysis was conducted to identify underlying patterns, says Levine, who had joined the quality department shortly before that pivotal month. "We designed the team based on our failure points," she says. Within short order after releasing the findings, the around-the-clock team was operational.

Led by an intensive care unit nurse and a respiratory therapist, the team has improved patient care. Preventable codes have declined by 76 percent during the most recent four-year period ending December 2010. Other safety efforts were incorporated during that stretch, but Levine credits the team with roughly one-third of the decrease.

Not everyone shares Levine's enthusiasm for rapid-response teams, though, even as they become nearly ubiquitous at U.S. hospitals. By the end of 2008, nearly 3,000 hospitals had committed to developing a team, according to the most recent data compiled by IHI, which first promoted the concept some six years ago as part of its 100,000 Lives Campaign. They wear various monikers: medical-emergency team, rapid-assessment team, critical care outreach team. Their design can vary significantly from hospital to hospital, and they often evolve over time. Still, they all are striving to resolve a central question: How best to save lives?

To date, the findings regarding mortality benefit have been mixed at best. One recent meta-analysis, which looked at 18 studies involving nearly 1.3 million hospital admissions, found that the teams were associated with a nearly 34 percent reduction in cardiopulmonary arrests in adults outside the ICU. But there was no reduction in mortality, according to the results published in the Jan. 11, 2010, issue of the Archives of Internal Medicine journal.

Last year, Johns Hopkins patient safety expert Peter Pronovost, M.D., co-authored a commentary in the Journal of the American Medical Association, in which he questioned whether hospitals are taking enough proactive steps, such as improving patient-flow problems, to reduce their need for deploying the teams. "Rapid-response teams are still at the end of the line," he said in an interview.

In recent years, hospital leaders have evolved their approach, with some developing broader rapid-response systems, says James Bagian, M.D., a University of Michigan professor who until recently served as chief patient-safety officer at the Veterans Health Administration. "This is a process—let's not make it a magical team," he says.

Bagian and others describe attempts that hospital administrators have made to break down some of the hierarchal and communication barriers between doctors and nurses that can impede patient safety.

Some of the most intriguing developments involve using the teams more proactively on the front end to identify at-risk patients, says Kathryn Wood, an assistant professor at Duke University School of Nursing, who conducted a rapid-response survey of academic medical centers. At some hospitals, the teams will round floor to floor. "They will ask the charge nurse, 'Who is your sickest patient today?' or 'Whom are you most worried about on this shift?' "

An Evidentiary Debate

To be effective, a hospital's rapid-response approach must be tailored to that facility's own barriers to patient safety, says Levine, echoing a point frequently made by others. "Your program won't be successful if you try to implement this how everybody else did."

An analysis at Children's Healthcare of Atlanta identified two primary problems: failure to recognize a patient in crisis and failure to speak up. The team's design and broader educational efforts built on those insights, Levine says.

She counsels patience. The system's initial goal had been to reduce preventable codes by 50 percent in the first year. Instead, codes increased by close to 30 percent. New procedures likely identified some codes that hadn't been identified previously. "It was at least 18 months to two years before we saw a sustained decrease," Levine says.

Pronovost worries that the mixed mortality findings by researchers reflects a recurring theme in patient-safety efforts. Hospital leaders, he says, are too prone to embrace quick fixes before scientific studies unravel which approaches work best and why.

Moreover, Pronovost has heard nurses complain that the teams serve as no more than a "cultural Band-Aid" and don't help to address underlying and more systemic quality problems.

While they may not be as pricey as high-tech investments, the teams do carry costs in terms of clinician time, both to create them and then to handle calls, says Paul Chan, M.D., a cardiologist at Saint Luke's Mid America Heart and Vascular Institute in Kansas City, Mo. Chan, who authored the 2010 Archives meta-analysis, found similarly discouraging findings when he analyzed mortality rates before and after the team's launch at Saint Luke's Hospital in late 2005. Neither hospitalwide cardiopulmonary arrests nor patient deaths declined, according to the findings, published in 2008 in the Journal of the American Medical Association.

It could be that mortality and other patient care improvements are more difficult to demonstrate at higher-quality hospitals, Chan says. But at this point, the teams have become so embedded in U.S. hospitals that it no longer may be feasible to design the sort of large-scale randomized trial needed to answer those types of questions, he says.

The team at Saint Luke's Hospital hasn't changed significantly since the publication of Chan's results, says Sally Ling, M.D., a Saint Luke's hospitalist who chaired the committee that oversaw the team's formation. A critical care nurse and a respiratory therapist typically are the first to respond; if needed, a physician is paged.

Despite Chan's findings, the team's influence on patient safety is too important to abandon the concept, Ling says. Some benefits, such as better communication and teamwork between clinicians, are less tangible and thus difficult to quantify, she says.

Kathy Duncan, a registered nurse and IHI faculty member, is far blunter about rapid-response teams. "When your grandmother deteriorates, you want one."

By establishing the systems, hospitals have developed a more responsive culture, Duncan says. As one example, she describes how some hospitals have added "nurse worried" or some equivalent as one trigger to call the team.

"Nurses tend to have this wonderful sense of doom when something is going to happen," she says. "It's hard to put your finger on it. Maybe the patient's blood pressure is fine, their heart rate is fine, but they just don't look like they should."

Wood, whose survey was published in 2009 in The Joint Commission Journal on Quality and Patient Safety, said that today's teams are more likely to include a respiratory therapist, but less likely to provide direct physician involvement. Instead, a doctor might be on call as needed, a change that's at least somewhat driven by financial considerations given the cost of physician time.

Another shift is that hospitals are making better use of the team's down time, Wood says. The team might check on vulnerable categories of patients, such as those recently admitted from the emergency department or transferred out of the ICU.

At Denver Health Medical Center, hospital leaders have developed an approach that Wood describes as less common, but that physicians tend to prefer.

Instead of creating a team, per se, officials at the 477-bed public teaching hospital established a set of clinical scenarios, or triggers, in which the nurse was required to call the physician who had immediate responsibility for the patient and, if no one responded, then moved up the oversight ladder to senior-level residents and attending physicians in 15-minute intervals.

"We didn't believe that a rapid-response team was necessary," says Kendra Moldenhauer, R.N., director of patient safety, quality and regulatory compliance at Denver Health. "But as an organization we felt like there were some outcomes where we could have done a better job. There were signs of clinical deterioration in our patients six, eight to 12 hours prior to the patient's having an event."

Depending on the urgency, the nurse could escalate his calls in 5- or 10-minute intervals, Moldenhauer says. The design of the program, introduced in 2006, was based on lessons learned from the system's own analysis. One of the four major impediments identified, for example, was failure to physically assess the patient. So the program requires a physician to examine the patient, rather than make recommendations by telephone.

The design avoids the additional communication handoffs that can occur when a separate team is introduced to the patient, Moldenhauer says. Another advantage: The physician contacted already is familiar with the patient's care.

She does acknowledge a few drawbacks. Since there isn't an actual team, worried family members don't have an activation number to call. Plus, proactive team rounding of vulnerable patients is not feasible, although Moldenhauer says that nursing supervisors strive to play a similar role at Denver Health.

Since the program was rolled out during the fall of 2006, Denver Health has identified a statistically significant decline in non-ICU cardiopulmonary arrests, according to data provided by the hospital. Prior to the program, the median rate was 5.9 arrests per 1,000 patients on the adult (non-ICU) floors compared with 2.2 arrests from April 2007 through December 2009. The mortality rate was roughly the same before and after. One possible explanation is that the facility's mortality rate already was relatively low, Moldenhauer says.

Ongoing Evolution

Some hospitals may have seen mortality payoffs, but maybe not to the point of statistical significance, Wood says. Also, she says, "It just really takes time to get nurses in the system to be aware of when to call and to feel comfortable to call. A lot of centers told us that the nurses who initially called when they started this team were belittled or derided when the physicians came to the bedside,with comments like, 'What a stupid call this is.' "

Meanwhile, hospitals continue to tweak their rapid-response intervention. Beginning in January, Children's Healthcare of Atlanta allows families to activate their team, Levine says. Until now, the physicians have been reluctant, worried that parents might abuse that option, sounding an alert if they didn't like their child's nurse or that night's meal. But feedback from other pediatric facilities hasn't corroborated that fear, she says.

Also, as the team has become more en-trenched in the pediatric systems, clinicians have realized that its activation doesn't constitute a black mark on their record, but rather an opportunity to help improve patient safety.

"We changed our culture by doing all of this," Levine says, stressing that other hospitals will have to find, and refine, their own path. "Please know that your team needs to be dynamic—otherwise it will be obsolete."

Charlotte Huff is a medical writer in Fort Worth, Texas.