Many hospitals rushed to establish rapid response systems after the Institute for Healthcare Improvement made them a key component of its “100,000 Lives Campaign” in 2005. Now, with teams established, hospitals must follow through with the next step in IHI's model for improvement: measuring effectiveness.
“How is it working and how do you know?” said Michael Westley, ACP Member, medical director of critical care and respiratory therapy at Virginia Mason Medical Center in Seattle. “We need to close the loop by showing how well we are identifying, and getting resources to, patients in trouble.”
While the benefits of deploying critical care specialists to the bedside before a life-threatening event occurs may seem obvious, there is limited evidence to prove that the rapid response systems, known as medical emergency teams (METs) or rapid response teams (RRTs), save lives.
A number of studies have suggested benefits in terms of patient safety and cost, but others have raised questions. For example, one meta-analysis concluded that there is “weak evidence” that RRTs are associated with reduced mortality and that large randomized trials are needed before they become a standard of care (Critical Care Medicine, May 2007).
“That's the tension,” said Dr. Westley of the conflicting evidence. “Inside the hospital we know that it takes a while to get the resources we need when someone is in trouble, but how do we show that METs are the solution?”
While it is difficult to prove that rapid response systems save lives, it is possible to demonstrate benefits, said Dr. Westley. To do that, hospitals must be diligent about tracking statistics, such as number of calls received, response times and number of unexpected transfers to intensive care. Armed with such data, RRTs can demonstrate that they are reliably identifying patients at risk before a life-threatening event occurs.
There are also the benefits that don't show up in statistics, said Bruno DiGiovine, MD, associate chief medical officer for quality at the Wayne State University Physician Group, who has worked with IHI helping hospitals set up and implement rapid response systems.
“We measured satisfaction among residents, and they were very happy with the overall program,” said Dr. DiGiovine, who established an RRT at Detroit-based Henry Ford Health System before assuming his current position. “Most felt like they would not want to practice at a hospital in the future without an RRT.”
Assessing your response
One of the hardest things to assess is whether a hospital's criteria for triggering an RRT call are working, said Dr. DiGiovine. It is easy to record the number of calls received, but harder to know whether the team is missing patients who should have triggered a call.
One way to determine this is to compare the volume of calls coming in with generally accepted norms. According to the IHI, hospitals need to reach a threshold of 20-25 calls per 1,000 discharges to demonstrate a decrease in overall hospital mortality. If your team isn't reaching that threshold, you are probably missing potential calls, said Dr. DiGiovine.
Quick, focused chart reviews of unscheduled transfers to the ICU can also help determine whether you are missing calls, he said. At Henry Ford, Dr. DiGiovine's team randomly screened unscheduled transfers to the ICU to see whether an RRT was called and if so, whether the call was made at the first sign of an established trigger, such as abnormal heart rate or respiratory rate. If a call was not made when a patient's respiratory rate was greater than 30, for example, it was considered a missed call.
At Yale-New Haven Hospital, hospitalist Victor Morris, ACP Member, was able to demonstrate his team's effectiveness by looking at the clinical data leading up to codes or cardiac arrests outside of the ICU.
“Before we [created the RRT], we had data showing that if you go back and look at patients who coded, up to two days before the code they had evidence of clinical instability,” said Dr. Morris, whose team is made up of a hospitalist, respiratory therapist and ICU certified nurse. “So, we predicted that if we intervened it would decrease codes, and that's what happened. It's pretty clear that [the decrease] is because of the RRT.”
In its “How-to Guide” on RRTs, the IHI cites several studies suggesting that certain warning signs signal pending cardiac arrest. For example, one study, published in Critical Care Medicine in February 1994, identified these warning signs often present within six hours of arrest:
- Mean arterial pressure <70 or >130 mm Hg
- Heart rate <45 or >125 beats per minute
- Respiratory rate <10 or >30 breaths per minute
- Chest pain
- Altered mental status
Beyond the numbers
Even though your team may have set the proper criteria for calls, it can take a while to see statistical changes as staff members adjust to the new system, the experts said. For example, the call volume may be low at first because people don't feel comfortable calling.
“When you start, it will be obvious you aren't getting all the calls you should be getting,” Dr. DiGiovine said. “You have to continue to advertise the service and make sure the team is customer-focused, that they thank people for calling and don't belittle anyone for calling.”
Kathy J. Simpson, RN, director of the MET for Intermountain Medical Center in Salt Lake City, attributes a recent drop in codes to her intensive “public relations” campaign to educate staff about using the MET since it was launched in 2005. Codes at the 742-bed hospital decreased from six to two per month between August and December 2010.
“It takes time to educate people and let them know that this is what we do,” said Ms. Simpson, who noted that she's made a concerted effort to educate physicians about the MET. Some physicians initially perceived the MET as an intrusion into their territory and chastised nurses for making a call, she said. When that happened, Ms. Simpson and a team physician intervened to address the issue.
Those efforts are starting to pay off and Intermountain has seen its call volume increase by about 50% a year, she said. “Now, nurses who do call, call over and over because they know they're safe,” she said. “Eventually, word gets out.”
It took years to reach the current call volume of 68 calls per 1,000 discharges at Virginia Mason, said Dr. Westley. He advised teams to anticipate 18 to 24 months of focused work before calls exceed the threshold level.
“We spent a lot of time getting nurses comfortable with summoning the team and getting the team's response to be helpful, respectful and affirming rather than, ‘Why'd you call me?’” he said. “It takes a while for nurses to become confident [about calling] and for teams to mature as responders.”
At Henry Ford, Dr. DiGiovine found that confidence in the RRT grew virally because residents who rotated through the pilot unit saw the value of an RRT and spread the word as they moved to other parts of the hospital. “They saw that calling the RRT could stabilize a patient before they got to the ICU, or prevent an ICU transfer,” he said.
One of the biggest benefits of Yale-New Haven's RRT has been a rise in nurse satisfaction, said Dr. Morris. “Nurses were ecstatic when we rolled this out,” he said. “They were frustrated with trying to tell the doctor something was wrong and having no one listen. Now, they can use the RRT.”
Dr. Morris has presented the results from nursing and other staff satisfaction surveys to hospital administrators to show that mortality statistics are only one impact of a rapid response system, he said. Looking at mortality rates alone can be misleading, he noted, because the RRT is one of many factors that influence inpatient mortality. The actual reduction in mortality might be very small if a hospital is already performing at a high level.
Sharing stories of successes or positive changes is another way to highlight the team's impact, said Dr. DiGiovine. For example, after Henry Ford's RRT nurse noticed that pulse oximetry checks were being omitted from routine vital checks, the procedure was changed to include them.
“The impact is hard to measure, but there's no question in my mind that there are patients who were hypoxemic who we would not otherwise have recognized,” he said.
One of Dr. Morris' favorite stories is of sitting beside a hospitalist when a rapid response pager goes off.
“They walk off immediately, no matter what they're doing,” said Dr. Morris, who notes that his team has a two-minute average response time. “We're doing a good job of identifying patients in distress and getting them in before they get to the point of needing ICU.”