Skip to main content
Management in Practice

Eliminating medical errors

Michael Apkon is using techniques from manufacturing to improve the efficiency and safety of medication delivery at Yale-New Haven Hospital, and he’s finding some solutions can be very simple.

By Ted O'Callahan

More than 1.5 million people are harmed by medication errors, creating $3.5 billion in treatment costs in the Unites States every year, according to a National Academies report. More dramatic than the numbers is talking to a mother whose child has experienced one of those errors, says Michael Apkon, vice president and executive director of Yale-New Haven Children's Hospital.

Apkon describes one incident that made a deep impression on him. A child coming out of surgery was given pain medication. The normal thing to do, but the patient's reaction wasn't normal. After some terrifying moments, the medical team realized that a hurried caregiver had picked up the wrong concentration of the right drug. The error was caught and corrected. Apkon met with the child's mother afterward. "Her child was fine. She wanted to know how the next child would be protected," he says. Apkon was then in charge of performance management for Yale-New Haven Health System, and had been working to create safer systems in the Children's Hospital for several years. "Having her sit there and talk it through created a sense of urgency."

Apkon is proud of the care delivered at YNHH, where he has served, in various capacities, since beginning his residency in 1989, but he is also proud of the efforts the hospital has made to improve. "We can focus on the metric of drug errors," he says. "But what's an appropriate number? What's an acceptable failure rate?" Yale-New Haven Hospital has 944 beds and dispenses 1.4 million doses of medicine a year. "No system can ever do it perfectly 100% of the time. But if you look at what could have happened to that child, that creates a tremendous amount of motivation to close that gap, regardless of whether it's one failure in ten or if it was the one failure out of a billion."

Apkon is working to "create a culture of a high reliability organization" at the children's hospital using the tools of the management profession. "The language that I like to use actually comes from other industries," he says.

These process improvement efforts draw on Six Sigma and Toyota Lean management processes. Six Sigma comes from the practices developed by Motorola to eliminate defects by reducing variation in process outputs. And Lean comes from Toyota's continuous effort to reduce waste. Apkon's medical research in critical care was always systems-focused, and he views these systems-focused management strategies as variations on the scientific method. "It's a hypothesis-driven approach to quality management," he says. "You frame a hypothesis about what you think is impacting performance. You use that hypothesis to come up with an intervention. And you test whether your hypothesis was right by putting that intervention in place and seeing whether you see the change you expected."

In one example, last summer the pediatrics unit held an improvement event focused on a core process in all acute-care hospitals: delivering medicines. The system needed to be both accurate and efficient. "When there are delays getting medicines to patients, they can be uncomfortable, they might not get the care that they need, staff get frustrated, and there are a lot of workarounds," Apkon says.

Building on data gathered by SOM students who had participated in a course Apkon teaches called Field Studies in Healthcare Management, staff from the pediatrics unit and the pharmacy spent five days planning and implementing ways to improve the medicine-delivery system. When staff from both sides of the medicine-delivery process worked together, Apkon says, "they got a different feeling for the demands placed on each other. And that created a much greater willingness to solve problems, rather than basically saying, 'This is your problem.'"

Managers were held accountable for producing measurable, budget-neutral results, not for implementing a specific solution. Improvement was measured by tracking a defect rate defined as a medication not being available to a nurse seeking to administer it.

Simply analyzing the data hadn't yielded many clues to the problem. "There were days where we had no defects and there were days we had up to a 16% defect rate. It was all over the place," says Lori Lee, the pharmacy manager involved. To figure out why, the team went and watched work being done. "We actually observed nurses looking for meds. We called it hunting and gathering." Without a clear procedure, medications could end up in several possible locations, leading to informal workarounds that created confusion and potential for mistakes.

"All we did was standardize the process," says nurse manager Kim Carter. The various delivery systems from the pharmacy to the unit now all end in one place. "It was almost silly. We labeled a plastic bin 'pharmacy in.' You would think it would have always been like that, but the nice thing about Lean is that very simple implementations make a huge difference."

Lee adds, "In one instance by putting the med bins in a specific spot and the medication ordering station right by it, we saved 50% of walking time" for the nurses.

Steps in the pharmacy were similarly concrete. Medicines ordered as critical will arrive in 15 minutes. Standard orders will arrive in 60 minutes. "That simple move of defining expectations meant there were fewer frustrated people in the process. And we were able to meet our turnaround time much better," says Lee. She has found that there is far less rework created by nurses checking on the status of medications and pharmacy staff now know what is a priority at any given moment.

"When I took over, frustrations with medicine delivery were about all I heard. Now it's not even an issue anymore," says Carter. The starting defect rate was 0–16%. Eight months later the range is 0–3%. And the changes made in the pediatrics unit are rolling out to other departments.

Apkon puts this kind of effort into the context of the broader healthcare picture: "We spend a tremendous amount of money on healthcare in this country. It's the biggest segment of the economy, around 16% of the GDP, and about the most rapidly growing. We ought to be paying for value."

He adds, "When patients get harmed by the system, when they wait too long, when they can't get the services that they need, when staff are running in circles trying to get tasks done, all those destroy value. They take away from the value that we could be creating in delivering safe and effective treatment. I think the response we've got to have as a healthcare delivery system is to work to continuously improve the value of the work that we do by providing safer care, more effective care, timelier care, care that's more customized to the patient, care that's efficient and equitable."