Nearly 13 years after the release of the Institute of Medicine’s landmark report To Err Is Human, which called national attention to the rate of preventable errors in U.S. hospitals and galvanized the patient safety movement, six out of every seven hospital-based errors, accidents and other adverse events still go unreported.
That’s the troubling conclusion of a report released in January by the U.S. Department of Health and Human Services’ Office of Inspector General (OIG). According to the OIG, the report, Hospital Incident Reporting Systems Do Not Capture Most Patient Harm, is the first truly national adverse event study documenting hospital-based adverse events among Medicare beneficiaries.
Previous studies assessing errors and adverse events had tracked the number of events reported but had no denominator; in other words, one hospital might have 1,000 reports of adverse events and errors in a given month, while another might have 50, but there was no way to know how many adverse events had been missed and how much harm had actually occurred. But this OIG study dug deeper, conducting in-depth year-long reviews led by independent physicians to first identify the rates of harm, then returning to the hospitals in the study where they knew harm had occurred to find out if incidents had been reported or not.
Overall, the inspector general estimated that more than 130,000 Medicare beneficiaries experienced one or more adverse events in hospitals in a single month. Perhaps one of the most surprising findings: Even some of the most serious, egregious adverse events were not reported. Of the 12 deaths in the sample that were iatrogenic, only two were actually reported as errors. More serious errors were no more likely to be captured as errors than mild reactions like breaking out in hives caused by a medication.
Why aren’t more hospital-based errors and adverse events being reported, and how can this situation be changed? The OIG’s report provides some insight into the underlying problem. The dominant reason that events weren’t reported was that they weren’t perceived as adverse events. (Not all adverse events are medical errors: A medical error is a preventable adverse event.) A urinary catheter infection, while not necessarily a medical error, is always an adverse event, yet only one of the 17 urinary catheter infections identified by doctors in medical review was reported.
Lack of recognition. Seeing adverse events and errors as just part of routine care is only one part of the problem, said Rosemary Gibson, MSc, a former senior program officer at the Robert Wood Johnson Foundation and coauthor of The Treatment Trap: How the Overuse of Medical Care Is Wrecking Your Health and What You Can Do to Prevent It (Lifeline Press, 2010) and Wall of Silence: The Untold Story of the Medical Mistakes That Kill and Injure Millions of Americans (Lifeline Press, 2003). Other issues include:
Fear of retribution. “You have to feel safe to report someone else’s error, or your own,” Gibson said. “Most hospitals have yet to create a safety culture.”
A sense of disbelief. “I remember talking to a hospital CEO once right after the IOM report,” Gibson said. “He said that after a significant error, he would get up in the morning and look in the mirror and think, ‘Did this really happen here?’ There’s almost a sense that if you don’t see it, it didn’t exist.”
Competing pressures. “The requirements of health care reform have taken up so much time and energy that I fear safety has moved to the back burner,” Gibson said. “Someone in a quality and safety leadership role at one hospital said to me, ‘Safety was just a fad. We’re not doing that anymore.’”
Productivity demands. “Health care’s mantra today has become volume, volume, volume. If you already have an environment that’s not as safe as we would like, and you ramp up the volume so people have to do more in the name of productivity, what’s going to happen?” Gibson said.
Adding to these problems is the fact that most hospital systems, like the legal system, contain no incentives to report errors—if anything, the incentives are stacked in the other direction, said Rahul Shah, MD, associate surgeon-in-chief and associate professor of otolaryngology and pediatrics at Children’s National Medical Center in Washington, DC. “On a hospital level, the medical staff will scrutinize you. There’s professional liability. There’s your reputation. Currently, there’s no positive incentive to report other than doing the right thing.”
Dr. Shah is the author of the only study to date to have collected data on errors in otolaryngology and classified them into the realms of pre-operative, operative, and post-operative care. Published in 2004, this study suggested that there may be as many as 2,600 episodes of major morbidity and approximately 165 deaths annually attributable to preventable errors in otolaryngology (Laryngoscope. 114(8):1322-1335).
“In pre-operative care, we find that the biggest cause of preventable errors is not getting all your consults ready before proceeding with surgery,” Dr. Shah says. “During the operative phase, technical errors are predominant. That makes perfect sense: We’re a surgical specialty. You need redundancy of equipment when possible, as well as technical representatives for such equipment.”
Post-operative errors in otolaryngology, Dr. Shah said, largely involve medications. “I’m a pediatric otolaryngologist and I prescribe only five or six medications, but even with that you’ll get dosage errors. I pause and take a moment to reorient myself before I write a prescription.”
Dr. Shah uses the phrase “zones of risk” to describe the areas of vulnerability for otolaryngologists. “We use sophisticated equipment, we do myriad surgeries, we work in the airway with oxygen and electrocautery, which entails a potential risk of fire, and we operate on many sites of surgery in a closed cavity,” he said. “We’re uniquely vulnerable.” (See also, “Adverse Events in the Medical Office Setting.”)
One hospital that is far ahead of the curve in error reporting is the University of Michigan Health System. In 2001, just two years after the release of To Err Is Human, the health system adopted a process of full disclosure of medical errors that involves multiple components, including an online incident reporting system, a multidisciplinary claims review committee, open and honest communication with patients and families and quality improvement initiatives guided by reported errors.
“They’ll get 7,000 or 8,000 reports a year from staff,” Gibson said. “That’s what you want. Most of them are minor or near misses, and you want to know about those so you can learn from them and prevent the larger errors. But you have to create a space where it’s safe for people to report.”
She said hospitals must adhere to the “Just Culture” principle in error reporting, a philosophy that recognizes that even competent professionals make mistakes. Quality improvement leader and Harvard School of Public Health Professor Lucian Leape, MD, has said that the biggest impediment to error reporting and prevention in medicine is “that we punish people for making mistakes.” A Just Culture model strikes a balance between personal accountability and the complex systems health care professionals work in and, when errors are made, establishes mechanisms to appropriately attribute responsibility to both individuals and systems.
“If I made a mistake and anybody else in my position could have made that same mistake, the chances are the system contributed to my making that mistake, and Just Culture would suggest I shouldn’t bear all the blame for it,” Gibson said. “That’s different from a case in which I did something that was truly negligent. Hospitals often do not make this distinction and lack a Just Culture.”
But traditional reporting systems in health care often do not support a Just Culture model or a truly patient safety-focused approach, said Brian Nussenbaum, MD, Christy J. and Richard S. Hawes III Associate Professor of Otolaryngology and Head and Neck Surgery at Washington University School of Medicine in St. Louis. “Paper or online reports are sent to hospital risk management departments, whose concerns are primarily to limit the potential legal risk. There is little emphasis on systems improvement or prevention, dissemination of incidents to others in the organization is unusual and the impact on clinical care is often not felt by providers.” (See “Hospital-Based Safety Programs: Making Them Work.”)
Whose Responsibility Is It?
Who should be reporting errors? Ideally, the responsibility lies with everyone involved in patient care, Dr. Nussenbaum said. In practice, physicians are more likely to report events that caused permanent harm or the death or near death of a patient, but overall they report only 1 percent of adverse events (Jt Comm J Qual Patient Saf. 2008;34(9):537-545). Nurses, who are more likely to report events that cause no harm or harm that’s only temporary, report about 45 percent of events. In the same study, residents reported only a little more than half of adverse events to attending physicians.
Dr. Nussenbaum said he tries to make sure that his residents do better. “If important adverse events are not being reported, then a discussion about these events does not happen and we can’t learn how to improve,” he said.
All his residents have a small card that fits in a shirt pocket. “It lists a set of criteria for reporting that all otolaryngology attendants agreed upon. If any of these events happen, they should be reported,” he said. The indicators were derived from recommendations by the Institute for Healthcare Improvement, the Agency for Healthcare Research and Quality and the American College of Surgeons, along with suggestions by attendings in his department.
The system usually yields about 15 to 20 adverse events reports per month. “At our mortality and morbidity conference, we discuss all mortalities, all reviewable events and all near misses,” Dr. Nussenbaum said. “The residents are particularly encouraged to discuss the near misses; those are the ones we feel that we learn a lot from. We analyze cases in a Just Culture manner, balancing personal accountability with system factors, and come up with a plan of what to do to try to proactively prevent these events from happening again.”
Over the past two years, this approach has led to numerous changes in patient care, including elimination of the use of unguarded bipolar cautery devices, placement of pre-mixed calcium gluconate on the floor for stat orders and implementation of airway tattoos on hospitalized laryngectomy patients.
But not all otolaryngologists are convinced that simply increasing the number of adverse events and errors reported is a goal to be pursued.
“No matter how precisely you try to define it, there are fuzzy boundaries,” said David Roberson, MD, associate in otolaryngology and director of implementation for the Program for Patient Safety and Quality at Children’s Hospital in Boston. “It would be easier to crack down on adverse event reporting if there were 100 percent agreement on what constituted a reportable adverse event.”
Children’s Hospital implemented an adverse event reporting system about five or six years ago, Dr. Roberson said. For the first three years, the number of events reported rose steadily, and then it plateaued. Now, the 400-bed hospital has about 15,000 adverse event reports per year.
“Several hundred of them are IV infiltrates,” Dr. Roberson said. “It’s nice to be able to count those and say we’d like that number to be smaller than it was last year, but is the world a better place because for every IV infiltrate, we have the nurse fill out a three-page form? With 15,000 reports a year, we can’t thoughtfully analyze them all. We do a serious analysis on the level three, four and five errors and those judged preventable. Let’s say the one in seven figure from the OIG report is correct and we ought to have 100,000 errors reported every year. How would we handle that number?”
—Rosemary Gibson, MSc
The science of error reporting in medicine, and what to do with the information gathered, is still evolving, Dr. Roberson said. “It’s relatively new for most hospitals to have all this information and to know what to do with it. Most take it as a public health ideal that every adverse event be reported, but I don’t think there’s any proof that it’s better to have all events reported than half.”
Gibson said she believes that what’s needed is a public system for the reporting of adverse events in hospitals, like a Consumer Product Safety Commission for health care. “If your child gets injured by a defective crib, there’s a system for you to report that,” she said. “We need that for hospitals. There should be an open capability for reporting events online, in person, any way you want, whether you’re on the hospital staff, a patient or a family member. Such a system should acknowledge the report, provide for an immediate action step in urgent situations and offer assistance and support to the person doing the reporting.”
But overall, the response in the healthcare industry to the OIG’s report has been “we need to do more research,” Gibson said. “We don’t need more research, we need to apply what is known. There are institutions that have made pockets of progress, and patients are alive today and home with their families because we’ve made care safer, but safety still hasn’t become a system property in health care.”