Home Patient Safety To boost patient safety, focus on the evidence, not the buzz

To boost patient safety, focus on the evidence, not the buzz

June 2007

Published in the June 2007 issue of Today’s Hospitalist.

Over the last few years, rapid response teams have received a lot of attention from the quality improvement movement. But take a look at the data on this hot development, and rapid response teams lose some of their sizzle.

According to patient safety expert Kaveh G. Shojania, MD, that’s because the science underlying rapid response teams is weak at best. Despite all the buzz about rapid response teams’ ability to reduce morbidity and mortality in critically ill patients, the only rigorous study on the topic found that the strategy did little to reduce death rates.

For Dr. Shojania, who is the Canada Research Chair in Patient Safety at the Ottawa Health Research Institute, rapid response teams illustrate one of the fundamental challenges of implementing quality improvement strategies. The evidence behind these strategies is not only mixed, but something of a mystery to all but a handful of people.

"The challenge facing patient safety is that most safety practices are still quite speculative," said Dr. Shojania, who is assistant professor of medicine at the University of Ottawa. "Patient safety still doesn’t have its version of aspirin and beta-blocker for acute MI, something that no one is ever going to question."

So how does a strategy like rapid response teams get so much attention in the first place? Dr. Shojania said that much of the enthusiasm for such teams was based on a number of before/after studies that found a striking improvement in morbidity and mortality rates among critically ill patients. When a more rigorous trial was conducted, however, researchers found that rapid response teams definitely increased the number of code blue calls, but with little benefit in terms of reducing cardiac arrests, ICU admissions or deaths.

That’s not to say that Dr. Shojania thinks that hospitals should ignore quality improvement initiatives. Instead, he urged physicians who want to embrace these strategies to behave as informed consumers and choose initiatives that have the best chances of working.

Tried and true approaches
Dr. Shojania, who was speaking at the Spring 2007 Hospitalist CME Series in Cambridge, Mass., said that because the evidence behind certain quality improvement strategies is shaky, hospitals need to select initiatives carefully. Instead of starting with the hippest and hottest intervention, he added, hospitalists making their first foray into quality improvement would do well to stick with a tried and true approach.

Dr. Shojania compared such initiatives to "low-hanging fruit," saying that they were not only relatively easy to implement, but that the evidence for all of them is strong.

While he’s the first to admit that these strategies aren’t exactly sexy, they are effective. Here’s a summary of the quality improvement initiatives he reviewed:

● Ultrasound guidance for central line insertion. A meta-analysis in the Aug. 16, 2003, British Medical Journal found that this simple strategy reduced the insertion failure rate by 86%. But Dr. Shojania noted that just as importantly, the study found that ultrasound guidance led to an 85% reduction in unsuccessful attempts to insert a line.

"We’re always worried about pneumothorax or hitting big red," he explained. "Ultrasound does reduce the rate of that, but the biggest problem is frankly just failing to put in the line."

He noted that the downsides of ultrasound-guided line insertion are minimal. A unit costs between $10,000 to $15,000 to purchase, but Dr. Shojania said that there are cost savings associated with successful line insertions. Besides, he added, ultrasound guidance can also be used for inserting PIC lines, eliminating one barrier to discharge.

In terms of potential harms, Dr. Shojania said that physicians new to the technique say it’s more difficult to remain sterile when using the probe. He noted that those concerns tend to pass quickly.

"It’s just different than what we’ve done before," Dr. Shojania explained, "but everybody I know who has used this technique says that with a few attempts and a little practice, that goes away."

Another common concern is that physicians may lose their line insertion skills and, in an emergency, without any ultrasound guidance at their disposal, they’ll be rusty. Dr. Shojania said there’s a simple solution: Do a procedure manually every so often just to keep your skills sharp.

Besides, he added, evidence suggests that using ultrasound to guide central line insertions actually improves your skills. "There have been some neat studies that have shown that when you use a technique like this," Dr. Shojania pointed out, "you are actually honing your skill with the conventional method. You actually end up with a better understanding of the anatomy."

● Preventing central line infections. The key to reducing central line infections, Dr. Shojania said, is using a number of very mundane but important interventions. They include standard hand disinfection, chlorhexidine for site preparation and full barrier precautions. (Using a pair of gloves and the small drape that comes with the central line kit are not enough.)

He also listed two other strategies that are not as standard, but still pretty basic: avoiding the femoral site and removing unnecessary lines. When it comes to avoiding the femoral site, he said, physicians need to use common sense.

If you are on call with a patient who is coding and needs a line and you feel more comfortable putting in a femoral line, "you should obviously put in the femoral line," Dr. Shojania said. "The point is that over the next 24 to 48 hours, if it becomes clear that the line isn’t necessary, it should be switched to a subclavian or intrajugular line, depending on access."

To provide an example of how these relatively low-tech interventions can reduce infections, he cited a study published in the Dec. 28, 2006, New England Journal of Medicine. That research found that just three months after implementing these strategies, a group of ICUs in Michigan reduced the rate of catheter-related bloodstream infections in half.

"If we invented a new drug that reduced strokes by 15% or 20%," Dr. Shojania said, "we would rejoice. So reducing something by 50% is a very large effect size."

● Stop orders for catheters. Everyone knows that complications from Foley catheters are common, but many are also preventable. While catheters are not indicated about one-quarter of the time, the even worse news is that their continued use is not indicated about 50% of the time. Physicians are often not even aware that their patient has a catheter.

One simple but effective intervention is to make sure that catheters are removed as soon as is appropriate. One solution is an automatic stop order. The order can be part of your hospital’s information system, or it can be as simple as having an order in the chart to remove a Foley catheter in 72 hours. Unless the physician overrides the order, the nurse removes the catheter.

In one study of an automatic stop order, researchers found a three-day difference in how long Foley catheters remained in patients.

Dr. Shojania said he uses an even simpler, hands-on approach: As part of his physical exam, he goes to other side of a patient’s bed and lifts up the blanket to see if there is a bag. "I think we can all admit that the genitourinary exam is not part of our routine physical on hospitalized patients, especially after they have been in the hospital for a couple of days," he said. "If you’re not careful, you can easily forget that the patient has a Foley catheter."

● Condom catheters. Another strategy to minimize the damage from a catheter and to avoid forgetting about it, at least in male patients, is to use a condom catheter.

"I remember being taught quite clearly that condom catheters were no better than Foley catheters," Dr. Shojania said. "The argument was often made that urine is in there and the bacteria will crawl up the urethra."

But a study in the July 2006 Journal of the American Geriatrics Society pointed to the benefit of condom catheters over Foley catheters, with a much lower incidence of adverse outcomes. In patients without dementia, researchers found a roughly five-fold difference in the rate of bacteria, symptomatic UTI or death.

While Dr. Shojania said that hospitalists could take a leadership role in creating a protocol that calls for patients to first receive a condom catheter unless otherwise indicated, he noted there is one catch: You probably need to work with other departments like the ED.

"What often happens is when patients have CHF," he explained, "the physician inserts a Foley before you even see the person."

● Read backs. While the evidence on the effectiveness of read backs is not as solid, Dr. Shojania said that read back protocols for telephone communications have a lot of potential.

"The idea is that when you’re communicating a test result or doing any kind of important phone conversation, you say, ‘I’m going to read back that result for you. This is for Ms. Dorothy Jones and the potassium was 3.4.’ You’re essentially doing what many restaurants do over the phone, which is read back our order to us."

Dr. Shojania said that read backs can help prevent some of the serious mishaps that occur because of phone miscommunications. In one case he cited, a phone call delivering a low potassium as a critical value was received by a clerk, who put the value in the chart of the wrong patient. Because that patient happened to have mild hyperkalemia, the error was serious.

The bad news about read backs is that there aren’t great data to support the strategy, and there likely never will be. While it would be difficult to conduct a randomized, controlled trial, Dr. Shojania pointed to a proof of concept study published in the June 2004 American Journal of Clinical Pathology.

When researchers examined 822 outgoing phone calls for critical results, they found that read backs detected errors in about 3% of calls. (They also found that errors occurred most frequently when physicians received calls.) About one-third of those errors involved the wrong patient, while another third cited the wrong test, and the last third reported the wrong value for a test.

"This is a very simple intervention that you could implement," said Dr. Shojania, "that would be of almost no cost and have some important benefits."

Moving forward
Dr. Shojania also discussed interventions that require a bit more effort and have less evidence, but are of particular interest to hospitalists.

Examples included sign-out systems to help avoid errors due to the "voltage drop" of information during cross coverage, as well as initiatives that target post-discharge adverse events. Those events have been shown to affect approximately 20% of patients discharged from hospital medicine services.

In addition, Dr. Shojania said there is some value to what he called "momentum-building" interventions. While they may have weak evidence, such initiatives carry a low risk of causing harm and offer a chance to foster collaboration and set the stage for other projects.

Dr. Shojania said he counts rapid response teams among those momentum builders, because such teams can boost nursing morale and help physicians spot patterns of problems.

"With a few successful if not sexy projects and a good momentum-generating initiative," he said, "hospitalists may feel up to the challenge of a big-ticket item “like working with the hospital to implement computerized physician-order entry."

Edward Doyle is Editor of Today’s Hospitalist.