When Good Enough Isn’t…. Good Enough: The Case for Reliability

​Americans don’t always have a favorable view of how the US health care system operates or how much everything costs. But ask someone about the quality of medical care in the US and chances are you’ll hear it’s the best in the world. Yet, clinical studies conducted by California’s RAND Corporation and other research centers consistently show that, for many medical conditions with well-defined criteria for the best treatment, only about 50 percent of the recommended care is ever delivered.
 
Industries in which many lives rest on performing every single task that’s required, again and again, such as airlines and nuclear power plants, are famous for designing high-reliability systems. In health care, however, “Variability is common,” says Roger Resar, MD, a Senior Fellow at the Institute for Healthcare Improvement (IHI). “In fact, it’s part of the culture.” Physicians prize clinical autonomy, meaning the right to do things their own way, says Resar, and because they work so hard during medical training, they come to view hard work as enough to ensure good care, when it’s not. “Human beings are fallible,” says Resar. “They make mistakes even when they’re trying to do the right thing and even when they think they’re doing the right thing.” He estimates that, in health care delivery, a defect — an error, omission, or other failure to accomplish an intended action — occurs, on average, 10 to 20 percent of the time, compared to .0001 percent of the time for airlines and nuclear power plants.
 
An 80 or 90 percent success rate might sound good to an institution, but from the patient’s standpoint it’s unacceptable, says Resar. “A hospital wants to know how well it’s doing for its entire patient population. The patient wants to know how well you’re doing by him or her.” For the individual patient, reliability is an “all-or-none” matter, says Resar. “If the right care has five elements and the medical team accomplishes all five only 90 percent of the time, that’s 100 percent failure for each of the 10 percent of patients who don’t get all the recommended care. There is no partial credit for reliability.” (For a recent discussion of “All-or-None Measurement," see the March 2006 article by Berwick and Nolan.) Of course, even when human beings do their very best, they don’t always succeed, so Resar and other faculty leading an IHI Learning and  Innovation Community have developed a framework for improved reliability based on industrial principles of standardization. It begins with protocols of care that are evidence-based and widely agreed-upon, says Fran Griffin, RRT, MPA, an IHI Director who has worked on IHI reliability programs.
 
Nationwide, 30 hospitals are part of this effort to reduce variability and improve outcomes for conditions ranging from heart failure to surgical site infections. “You start with a well-accepted standard of care so you don’t get off track arguing about what needs to be done,” says Griffin. “Then, for the same reason, you design and standardize the internal processes that work best in your own hospital to make sure that everything that needs to be done gets done.” Standardization is crucial to improvement, explains Griffin, “because that’s what promotes reliability.”
 
To compensate for human limitations, the IHI framework employs a three-tiered strategy: 
  • Prevent Failure: This is the first line of defense and includes stipulated guidelines, tools, and techniques for best-practice treatment of specific conditions; in-house campaigns to promote awareness and vigilance; memory aids such as color-coding or sequential numbering to differentiate similar items; checklists; and making the desired action the default, such as providing smoking cessation counseling to all patients who smoke to ensure that those most at risk, such as cardiac patients, receive counseling.
         
  • Identify and Mitigate Failure: Human factors, such as fatigue or distraction, sometimes thwart initial prevention. Techniques to avoid or at least reduce the impact of failure include standing orders for best-practice treatments that can be altered only with written authorization; warnings when an undesirable event is approaching, such as a computer flag when poorly-interacting drugs are prescribed for the same patient; and independent double-checking of required actions. 
 
  • Redesign for Success: When root cause analysis of persistent failure shows that it is embedded in the system design, redesign is required. It’s necessary to understand where a failure is really occurring in order to figure out the remedy, says Griffin, who offers an example from the automobile industry. “Years ago, there were reports of cars accelerating suddenly, when the drivers swore they were in parking gear. Some of the runaway cars killed or injured people so it was a big liability crisis, but the manufacturers didn’t argue about who or what was at fault for long. They just redesigned their gearboxes so that, today, cars cannot be taken out of ‘park’ unless the driver’s foot is on the brake. Even if the driver is confused, the car won’t let him make a mistake.”
 
One organization that has taken the reliability message to heart is Cincinnati Children’s Hospital Medical Center (CCHMC), a 475-bed, nationally recognized pediatric treatment and research center in Cincinnati, Ohio. In 2005, working with hospital staff and IHI’s Innovation Team, the CCHMC board of trustees targeted 18 areas for improvement, from surgical site infections to asthma care to moving patients efficiently through the system, applying 99 specific measures to track progress. “While the moral case was clear,” says board chairman Lee Carter, “it was also a business decision. You get what you measure and what you pay for, so we’re measuring and paying for improvement.” Indeed, the year-end bonuses of CCHMC senior managers, once based only on meeting financial goals, are now 70 percent dependent on meeting stipulated quality improvement goals. 
 
Board involvement didn’t end there, says Carter. “There’s a saying —  ‘the currency of leadership is attention’ — and this board pays a lot of attention.” The patient care committee, which Carter leads, meets once a month for presentations by improvement teams throughout the hospital. “If a team isn’t moving as well toward its goals as they would like, we work on figuring out what can be done and how the board can help. If they’re moving even better than expected, we discuss whether the goal should be reset. For us, progress reports aren’t just words and numbers on a page; they’re living documents.”
 
Carter acknowledges that achieving some of CCHMC’s goals may take quite awhile. “We want to improve long-term functionality outcomes for chronic conditions such as cystic fibrosis, but it’s not easy work and you don’t know where the tipping point is. You have to just keep going without necessarily seeing an end in sight.” 
 
Transparency and “absolute alignment” between the board and senior staff is the key to moving forward, says Carter. “We won’t adjust our expectations to make people look better or feel better but, at the same time, we’re all working together. Everyone is on the same side.” In fact, says Carter, more than half of the quality improvement component in senior management bonuses is based on general improvement throughout the hospital; less than a third is based on improvement within a manager’s own area of responsibility. “If someone is having a hard time, everyone’s first instinct is to say, ‘What can I do to help you?’” Just as gratifying, though, says Carter, is the pride generated by successful teamwork. “When I lead a tour in the hospital, I see good statistics posted on the walls, even in the bathrooms. Our people are just so motivated by what they’ve been able to accomplish.”
 
One example is CCHMC’s progress on reducing ventilator-associated pneumonia (VAP), which can develop when an intensive care patient on a breathing tube accidentally aspirates fluid or bacteria. 
 
“We went  from delivering perfect care for children with six common conditions 80 percent of the time to delivering it 95 percent of the time,” says Uma Kotagal, MBBS, MSc, Director of CCHMC’s Health Policy and Clinical Effectiveness division. “We then applied what we learned to VAP, where in three to four months we were delivering all components of the bundle close to 100 percent of the time.”
 
The hospital’s patient population presents some special challenges, says Kotagal. “Best-practice guidelines for prevention of VAP are based on evidence developed in adults; we needed to tailor ours to children.” Accordingly, the hospital assembled an expert panel to examine and modify the adult best-practice recommendations. For example, because intubation is very stressful, the adult standard calls for intravenous antacids to counter the formation of peptic ulcers. “Children are not prone to peptic ulcers, so they don’t need antacids,” says Kotagal.
 
In designing their VAP protocol for high reliability, the team not only standardized the elements of the bundle but also located a type of respiratory tubing and heater circuit that so thoroughly resists condensation that it virtually eliminates fluid drainage into the lungs. “That’s the only type we use,” says Kotagal. The hospital also standardized several other improvements, for example:
  • Every-four-hour “huddles” between ICU nurses and respiratory therapists at the bedside of ventilator patients “to make sure key portions of the bundle are being implemented and all is going well”    
  • Immediate bedside huddles when a VAP is discovered to analyze the source and cause (“Our people aren’t looking at month-old data; they’re looking at conditions on the floor.”)
  • Computer screen savers that reiterate the elements of required care, plus laminated pocket-sized cards for all staffers to carry with them (“I guess you could say we’re a little obsessed with eliminating failure,” says Kotagal.)
 
Obsession has paid off. Since last fall, CCHMC’s number of VAPs per 1,000 days of ventilator use across all ICUs has plummeted from seven to less than two.

  

graph_CCHMC.jpg 
 
Lee Carter is pleased but not entirely surprised by results like this. “We planted a flag,” he says. “This is what happens when you’re very, very focused.” He expects the results of the 2005-06 fiscal year effort to be available by September. “Where we haven’t quite met our goals, we’ll keep them in place for the next year. Where we have met our goals, we’ll either up the ante or, if our met goal is unbeatable — say, zero VAPs — our goal will be to sustain it.”
Average Content Rating
(2 users)
Please login to rate or comment on this content.
User Comments