Published in the August 2006 issue of Today’s Hospitalist
When Missouri Baptist Medical Center decided to make reducing adverse drug events a top priority, it faced a dilemma. While most hospitals enlist the help of sophisticated computer systems to combat drug errors, the St. Louis hospital had little in the way of high-tech tools. There was no computerized ordering system, for example, and no decision-support system.
As the hospital quickly learned, however, it didn’t need sophisticated computer systems to fight adverse drug events. By implementing a series of low-tech strategies that even small hospitals with no technology budget can use, Missouri Baptist slashed the rate of drug errors within its walls.
That’s not to say the task was easy. The hospital spent years working on the effort, and it used just about every quality improvement strategy under the sun. But Missouri Baptist was able to get the job done without having to spend a small fortune on information technology.
Here’s a look at a few of the key strategies that the hospital used to reduce adverse drug events, including techniques that hospitalists can implement in their own facilities.
Chart reviews
It all started in early 2001, when Missouri Baptist undertook an effort to audit its rate of adverse drug events. Because the hospital lacked many of the computerized tracking tools used at other hospitals, it adopted an approach that relied on auditing charts by hand.
Auditing charts by hand may sound like a time-consuming process, but Max Cohen, MD, says that nothing could be farther from the truth. By using the Institute for Healthcare Improvement’s “trigger tool” for detecting adverse drug events, he adds, the hospital was able to detect drug problems relatively quickly and easily.
Instead of pouring through hundreds of charts, the IHI tool calls for auditing as few as 10 patient charts to determine a hospital’s adverse drug event rate. As Dr. Cohen explains, previous research has found that increasing the sample size beyond 20 didn’t change the results, regardless of the size of the hospital being studied.
The tool uses 24 “triggers” to very rapidly detect patients who may have experienced harm. In one instance, for example, that trigger was a patient on coumadin who had an INR greater than 5. That individual went on to develop GI bleeding.
Getting started
Dr. Cohen, who was lead author of a study on the initiative that was published in the June 2005 issue of Quality and Safety in Healthcare, explains that two reviewers at the hospital examined patient charts. When they found one or more of the 24 triggers, they would look at the chart more closely to determine if an adverse drug event had occurred.
Dr. Cohen, who is now chief medical officer at New York University Medical Center in New York City, says the problems the hospital uncovered were fairly typical of what’s been reported in the patient safety literature. The biggest problems were related to agents like anticoagulants, narcotics and insulin.
While Missouri Baptist used a wide range of quality improvement techniques to reduce adverse drug events, Dr. Cohen says he was particularly impressed by the IHI’s trigger tool. That’s why he encourages hospitalists to embrace the tools to measure adverse drug event rates in their facilities, or even within the wards where they work.
“It’s a great place to start,” he says, “and it’s not so difficult to do. Reviewing 20 charts a month takes one person a couple of hours. It gives you such a good handle on how much harm you are causing patients. As the quality improvement people say, if you don’t measure something, you can’t improve it.”
Importing protocols
Once the hospital had a handle on the scope of the problem, it needed to take action. Instead of starting from scratch to address the drug events that were found during chart audits, the hospital turned to the expertise of organizations like the Institute for Healthcare Improvement and the Institute for Safe Medication Practices.
“We stole everyone’s good ideas,” quips Dr. Cohen. “We didn’t want to reinvent the wheel.”
Importing solutions crafted by other organizations, however, required some diplomacy. As Dr. Cohen explains, medical staff would not want to think that outside protocols were being foisted upon them by medical administrators.
The hospital identified members of the attending staff who were interested in the areas being targeted by the protocols. Those physicians then assembled teams of clinical staff that were charged with scrutinizing “and changing “the protocol.
Dr. Cohen notes that while the protocols received some tweak¬ing, most didn’t undergo major changes. As a result, he says, most were rolled out in three to six months.
Starting point
One question was where to pilot protocols designed to detect and prevent adverse drug events. Dr. Cohen says that the ICU was the logical choice for two reasons.
First, the ICU was the locus of many of the adverse drug events recorded during chart reviews. In addition, Missouri Baptist’s team of full-time intensivists would provide leadership and some much-needed consistency in rolling out new protocols.
“We found that the ICU was a great place to pilot protocols,” he explains. “We had a full-time staff of critical care physicians who provided 24/7 coverage of ICUs, so one of them was present to make sure that the protocols were being followed all the time.”
Once protocols had been pilot-tested in the ICU, they were disseminated through the rest of the hospital. Dr. Cohen says that at that point, the hospital experienced some resistance from physicians, often from attending physicians who didn’t work exclusively in the hospital.
That’s where the hospitalists came in. “We worked very hard with our group of hospitalists to influence other physicians to buy into the protocols,” he explains. “We found that they were very useful in getting the word out to the entire medical staff.”
The bottom line
To determine the success of the effort to reduce adverse drug events, Missouri Baptist gathered information at three intervals. The differences in problems in each of those three periods speak for themselves.
The hospital, for example, found that the median rate of adverse drug events per 1,000 doses of medication administered was 2.04 in the baseline period. During the transition period, when the protocols were being phased in, the rate had already dropped to 1.26.
After the interventions were complete, the rate had dropped to a stunning 0.65. Dr. Cohen says that rate later dropped to 0.25, where it has remained for several years.
Viewed from a different perspective, the hospital saw a 3.6-fold lower risk of adverse drug events, and the number of drug problems associated with patient harm showed an eight-fold drop. At the time that the hospital was putting the finishing touches on the first phase of the project, 4,400 adverse drug events were being prevented for an annual savings of at least $10 million.
Dr. Cohen is quick to point out that the effort involved large numbers of people and long-term strategies. Hospital administrators spent a lot of time and effort assuring physicians and nurses that they wouldn’t be punished for reporting errors.
Physicians like Dr. Cohen also had to find ways to get the attention of hospital leaders so the initiative didn’t wither in committee meetings. The solution was to create a new kind of committee that included senior members of the hospital’s executive team and physician leaders.
“This immediately gave safety a very high profile,” he explains, “and it created a mechanism that allowed us to address patient safety issues at the highest level without going through a series of committees to reach management for resolution. It basically short-circuited the traditional process, so questions that were raised about safety went straight to the top.”
While a long list of tools and strategies made the project a success, Dr. Cohen says the fact that it didn’t break the bank was a huge key to its success. He thinks it’s an approach that other community hospitals without huge budgets can emulate.
“One of the things that we were most proud of,” he says, “is that we made these improvements without spending a lot of money on information technology.”
Edward Doyle is Editor of Today’s Hospitalist magazine.
