No End to Errors
Three Years After a Landmark Report Found Pervasive Medical Mistakes In American Hospitals, Little Has Been Done to Reduce Death and InjuryDec 3, 2002 | Washington Post
When the august Institute of Medicine (IOM) issued its blunt assessment of medical errors three years ago, Lucian L. Leape, the pioneering Harvard physician researcher who helped write the report figured it would meet a fate common to such documents: The initial flurry of media accounts would be followed by a swift descent into obscurity.
Instead, the report's conclusion that as many as 98,000 hospitalized Americans die every year and 1 million more are injured as a result of preventable medical errors that cost the nation an estimated $29 billion commanded attention in a way Leape and his co-authors never imagined.
Shortly after its release, Congress held hearings and promptly earmarked $50 million for research into the causes and prevention of medical mistakes. President Clinton announced his support for a key recommendation mandatory reporting of serious errors an idea borrowed from aviation.
Four bills that would establish error reporting systems have been introduced on Capitol Hill. A consortium of Fortune 500 companies launched a year after the report's release is pressing hospitals to make specific changes in clinical practice known to reduce mistakes. "Patient safety" has become a mantra of the nation's hospitals. Beginning in January hospitals will be required by accreditors to show they meet six basic standards that reduce errors, which the IOM said kill more Americans than breast cancer, traffic accidents or AIDS.
"Frankly we were all very surprised," recalled Leape, a former pediatric surgeon and the author of several earlier groundbreaking studies of medical errors. "Before the IOM report, nobody was doing diddly squat. Now there are a lot of good people involved and a tremendous amount of activity," he said. "Of course, activity is not the same as progress."
The distinction drawn by Leape underscores the reality of the nascent movement to reduce medical mistakes: There's a lot of talk, but no significant progress. The reasons, observers say, include fierce resistance by doctors and hospitals to mandatory reporting and other IOM recommendations, a lack of oversight by the federal government and the absence of an effective consumer lobby.
As a result, experts contend, it's doubtful that patients checking into most of America's 5,200 hospitals today are any less likely to be killed or injured than they were on November 29, 1999, when the report was issued. With the conspicuous exception of the Department of Veterans Affairs (VA) medical system, whose hospitals have embraced the ethic and many of the methods that have made aviation and other industries safer, most hospitals have taken few new steps to protect patients from errors.
"I'd say patients are safer today in some hospitals, and certainly in the VA, but it's still a pretty small minority," said physician Don Berwick, a member of the IOM panel who is president of the Boston-based Institute for Healthcare Improvement, a nonprofit group dedicated to bettering the quality of health care. "Safety is a very hard thing to accomplish and it has to be pushed way up to the top of the list, and that still hasn't happened" in most places.
The vast majority of hospitals still rely on paper charts that often can't be located and are difficult to decipher, rather than more accessible and legible computerized medical records. Fewer than 3 percent have fully implemented computerized drug ordering systems, which have consistently shown dramatic reductions in drug errors. Operations performed on the wrong body part or the wrong patient have increased, according to the Joint Commission on the Accreditation of Healthcare Organizations (JCAHO), which inspects hospitals.
The nation's most exhausted and inexperienced doctors the 100,000 interns and residents who staff teaching hospitals continue to work as many as 130 hours a week, often with little or no supervision. Hospital-acquired infections, which kill about 90,000 patients annually, have increased 36 percent since 1980, a rise that coincides with the proliferation of bacteria capable of resisting the most potent antibiotics, according to the Centers for Disease Control and Prevention (CDC).
Although the estimates of death and injury contained in the report are huge, the actual number of deaths is undoubtedly higher. The IOM considered only errors committed in hospitals, and not in other medical settings where they undoubtedly also abound: clinics, outpatient surgery centers and doctors' offices.
Experts say the nationwide shortage of registered nurses as well as the unprecedented demands on emergency rooms have exacerbated an already bad situation.
The incessant talk about safety and the paucity of concrete action "illustrates the difference between meeting-room reality and in-the-trenches reality," said Michael L. Millenson, a visiting scholar at Northwestern University and author of the 1997 book "Demanding Medical Excellence." Millenson said that while the IOM report contained nothing new, it did succeed in finally generating serious attention to medical errors, a problem first mentioned in the medical literature in 1955.
Yet many doctors, he added, refute the report's central thesis that mistakes are numerous and affect all players in the increasingly dysfunctional health care system. And most resist the notion that hospitals' faulty systems need to be overhauled to guard against errors that can result from anything short of perfect performance by individuals.
"You won't believe the number of times I've heard a doctor say, with a straight face, 'I don't make mistakes,' " said Millenson. "There's an old saying in aviation: The pilot is the first one at the scene of an accident. Well, in medicine, if someone makes a mistake, who gets hurt? It's not the doctor. Who pays? It's not the hospital. Nobody's doing this on purpose, but they're not losing money on it, either."
Hospital CEOs, Millenson added, are loath to risk the wrath of their medical staffs by focusing on safety. Why spend millions of dollars on computerized drug ordering systems when it's easier and more popular to buy a new piece of equipment that doctors want and that could also attract patients, increase revenue and mollify the board of trustees?
Nor is there any pressure from the federal government to reduce mistakes. Hospitals are paid the same amount if their error rates are abysmal or stellar. "What's the competitive advantage for a hospital that tackles this problem?" Millenson asked. "What would they advertise: 'Now killing fewer patients'?"
Because only a minority of states require that serious errors be reported, it's impossible for experts to figure out how best to prevent their repetition or for consumers to know where they are most likely to occur.
"We can't even count errors, so we don't have any more real information than we had when we wrote the report," said IOM panelist Arthur Levin, director of the Center for Medical Consumers in New York. "And we don't have public information that could help guide people through this maze."
Don Nielsen, a physician who is senior vice president for quality leadership at the American Hospital Association (AHA), said his organization has encouraged its members to create "a culture of safety" in their institutions.
"There's a much greater awareness that errors can occur, and from a hospital perspective there is increased attention and greater resources being devoted to safety," he said.
"I think we've scratched the surface. But more change needs to occur," Nielsen added.
This story will examine five areas that directly affect hospital patients to see what's happened in the three years since the IOM issued its report.
Medication errors are among the most common preventable mistakes, the IOM report found, and they remain rampant in hospitals; experts have estimated that more than one million serious drug errors occur annually in hospitals alone. A recent report in the Archives of Internal Medicine found that one in five doses of medication dispensed at 36 hospitals and nursing homes around the country was either the wrong drug or the wrong dose, or given at the wrong time or to the wrong patient.
The more drugs a patient is taking and the more people involved in the delivery of a medicine, the greater the chance of a mistake. The explosion in the number of drugs on the market there are now more than 10,000 has increased the chances of error. So has the similarity of names that can be easily confused, such as Lamisil, a drug prescribed for fungal nail infections, and Lamictal, an epilepsy drug.
While many drug errors don't injure patients, others are lethal. The most notorious of these, cited in the IOM report, is the massive chemotherapy overdose that killed Boston Globe health reporter Betsy Lehman, 39, in 1994 and gravely injured another woman. The circumstances of Lehman's death, which was the subject of a front page story in the Globe, rocked the Boston medical establishment as well as cancer treatment centers around the country. It is widely regarded as a watershed event that led to the birth of the fledgling patient safety movement.
The Lehman overdose, which a dozen doctors, nurses and pharmacists failed to notice, was caused by an initial miscalculation of a toxic breast cancer drug, a mistake compounded by a cascade of other errors. They include the failure of doctors at the prestigious Dana-Farber Cancer Institute in Boston, a federally designated comprehensive cancer center, to investigate complaints by Lehman's husband, a scientist at Dana-Farber, or lab tests that indicated something was terribly wrong.
Other fatal medication errors have resulted from the accidental overdose of concentrated drugs, particularly potassium chloride. For decades these drugs were stored on hospital wards where they were administered to critically ill patients as an additive to intravenous solutions to restore electryolyte balance.
Sometimes harried or distracted nurses forgot to dilute them and mistakenly administered a lethal injection. (Potassium chloride instantly stops the heart and is used for this purpose in states that administer the death penalty by lethal injection.)
To prevent accidental overdoses, the IOM recommended that undiluted potassium chloride and similarly hazardous drugs be removed from patient floors. "We advocated that as far back as the 1980s," said Michael R. Cohen, a pharmacist who is president of the Institute for Safe Medication Practices, a nonprofit group in suburban Philadelphia. "For years we would publish reports of actual injuries. And they would recur because people and organizations were not held accountable for addressing them."
Starting in January, the JCAHO will require that hospitals remove potassium chloride and other hazardous concentrated drugs from patient floors. While most have done so, "occasionally we still see holdouts on specific nursing units," Cohen said.
Other common causes of drug errors include the use of abbreviations such as "u" (short for units) which can be mistaken for a zero, misplaced decimal points and doctors' legendarily illegible handwriting. Three years ago a Texas jury awarded $450,000 to the family of a 42-year-old man who sued a cardiologist and a pharmacist after he was given a massive overdose of the wrong drug and died. The pharmacist said he had trouble reading the doctor's writing.
To address the legibility problem, some hospitals have sent doctors to remedial penmanship classes. Most experts consider this a poor substitute for the more effective and expensive remedy endorsed by the IOM: computerized drug ordering systems linked to a hospital pharmacy, which one study found reduced medication errors by 86 percent. These systems also can help prevent accidental overdoses, like the one that killed Lehman, and make it virtually impossible for a doctor to prescribe a drug to which the patient has a known allergy, another common mistake.
But so far, according to Suzanne Delbanco, executive director of the Leapfrog Group, an organization of Fortune 500 companies that is pressuring hospitals to improve quality and reduce errors, only 2.5 percent of hospitals have fully implemented computerized drug ordering systems.
The AHA's Nielsen said that cost, the poor quality of some systems and physician resistance have been barriers to computerized drug ordering.
Persuading doctors to use computerized drug prescribing systems, which take longer than scribbling a note in a chart, is a formidable challenge. "Even in hospitals that have them, [some] doctors refuse to use them," said Charles Inlander, director of the People's Medical Society, a consumer group in Allentown, Pa.
Nielsen agreed that physician resistance remains an obstacle. "Doctors will give the nurse a verbal medication order and will write it down and tell her" to enter it in the computer, thereby subverting the system, he said.
Dennis S. O'Leary, the internist who directs the JCAHO, is adamant: Surgeries in which doctors perform the wrong operation or operate on the wrong side of the body or on the wrong patient "should never happen."
Yet every year since 1995, O'Leary said, the commission has seen an increase in voluntary, confidential reports of what is known as wrong-site surgery. Since 1998, the JCAHO has issued two warnings to hospitals about the problem.
"There has been no diminution," said O'Leary, who noted that the five to 12 such reports received each month involve outpatient surgery centers as well as hospitals.
The lack of a national error-reporting system means there is no way to tell how often the problem occurs. Although cases reported to JCAHO include the removal of the wrong breast or kidney or a biopsy on the wrong side of the brain, the problem is believed to be most common in orthopedic surgery.
"This is not quite Dick and Jane, but it's pretty close," O'Leary said. "I think there are a lot of errors around very simple stuff like whether the X-rays are up on the view box correctly."
In the case of Kevin Walsh, a 41-year-old construction worker from Staten Island, N.Y., they were not. Last year, neurosurgeons at Long Island College Hospital in Brooklyn operated on the wrong side of his brain because a CT scan they were working from was reversed.
While it seems elementary, eliminating wrong-site surgery won't be easy. Five years ago, Memphis surgeon S. Terry Canale, immediate past president of the American Academy of Orthopedic Surgery, said, his group launched a campaign called "Sign Your Site," modeled after successful programs in Britain and Canada. The goal is to get orthopedic surgeons, who have a 25 percent chance of performing wrong-site surgery at some point in a 25-year career, to sign the correct site with an indelible pen just before surgery, after verifying the procedure with the patient.
Canale said he was surprised at the opposition he encountered. "All these famous prima donna orthopedic surgeons said, 'I don't have time to go talk to a patient before surgery' or 'I've never made a mistake, why should I do this?' And I said, 'You're exactly the kind of guy who's going to get into trouble.' "
In the past year, Canale noted, resistance has diminished; he estimates that between 60 percent and 80 percent of orthopedic surgeons nationwide are now complying. Canale said he expects the JCAHO's new requirement that hospitals devise specific procedures to eradicate wrong-site surgery to further boost compliance.
But the issue underscores what experts say is a leading cause of error: the lack of uniform procedures in the medical profession, which has fiercely resisted standardization and continues to prize physician autonomy.
"For one thing, there's still some discussion about whether to mark the right side or the wrong side or both sides," said consumer advocate Art Levin. "There's not even a uniform standard about which side to mark."
That's like allowing each airline or every pilot to determine the plane's route and which preflight procedures to follow, experts contend.
"In an airplane, the pilot and co-pilot go through a checklist every time before they take off," O'Leary noted. "We don't do that in a hospital."
After years of mostly futile attempts to persuade doctors, nurses and other health care workers to wash their hands between patients, the CDC recently unveiled its new "hand hygiene campaign." This is the agency's latest effort to reduce the estimated 2 million infections and 90,000 deaths annually caused by infections that patients pick up in hospitals. Half of these infections could be prevented by proper hand washing, according to the CDC.
Officials hope the new campaign will be more successful than its predecessors. For the first time the CDC is advocating the use of alcohol-based hand rubs, which it says are faster, easier, equally effective and less irritating than soap and water.
But there's no way of predicting how many doctors, nurses and other health care workers will use them. "It's possible these handrubs will increase compliance," said the AHA's Nielsen, "but I wouldn't generalize, given the history of poor compliance."
Studies have found that hand washing is the exception rather than the rule and is inversely related to status: Doctors are less likely to wash than nurses' aides. Although gloves are ubiquitous in hospitals, they mostly protect the wearer, not the patient and can spread infections if they become contaminated.
Patients have no way of determining the infection rate at a particular hospital. The CDC tracks infections through a system of voluntary reporting by about 400 hospitals, Nielsen said and the agency does not disclose their identities or infection rates.
"There is no evidence that hospitals are doing anything about this problem," said Inlander of the People's Medical Society. "This is one of the most common errors and one of the biggest problems confronting patients. And there's no pressure on hospitals to institute vigorous hand washing programs."
Fatigue and Supervision
Despite numerous studies from aviation, aerospace, the military and other industries linking fatigue with mistakes, sometimes fatal ones, most of the nation's 100,000 interns and residents continue to work 80 to 120 hours per week. At night and on weekends, when senior doctors are largely absent, these neophyte physicians are expected to make life-and-death decisions with minimal guidance. Defenders of the current system say that no studies have linked fatigue or inadequate supervision to medical errors; some have said that doctors are different than other professionals and learn to transcend exhaustion, in defiance of the laws of human physiology.
That may change. The federal Agency for Healthcare Research and Quality is funding eight studies examining the relationship of fatigue, stress and sleep deprivation to mistakes made by doctors and nurses in hospitals.
Long work hours by doctors "especially residents are incompatible with a safe, high quality health care system," warned Stanford anesthesiologists David M. Gaba and Steven K. Howard in a recent article in the New England Journal of Medicine. If organized medicine doesn't reduce these excessive hours, they warn, "change may be ultimately forced on us."
In the past year, bills have been introduced in Congress by Sen. Jon Corzine (D-N.J.) and Rep. John Conyers (D-Mich.) which would limit the maximum workweek for residents to 80 hours and penalize programs that violate these rules up to $200,000 annually. In addition the group that accredits residency programs in the nation's hospitals has announced it will restrict residents' hours to a maximum of 80 per week averaged over four weeks beginning in July 2003; critics say the group's guidelines are riddled with loopholes and lack tough penalties.
In New York, the only state that limits residents to 80-hour workweeks, these regulations are routinely violated because the fines for noncompliance are so low.
It's hard to know what role long hours might have played in the death of Mike Hurewitz last January at New York's Mount Sinai Hospital, but inadequate supervision was a factor, according to state investigators. The 57-year-old Albany resident died after choking on his own vomit, three days after he donated part of his liver to his brother, a physician. Hurewitz was being cared for by an intern who told officials she felt "overwhelmed" after being left alone in charge of 34 patients.
There is a glimmer of change in the traditional "see one, do one, teach one" ethos that has been a hallmark of medical training, in which neophyte doctors are supervised by those a year or two ahead of them who may be equally exhausted.
Six months ago, Beth Israel-Deaconness Medical Center in Boston, a teaching hospital affiliated with Harvard Medical School, became one of the first in the nation to staff its intensive care units around the clock with a specialist in critical care.
"It just made sense that the sickest patients in the hospital at 2 a.m. should have attendings and not just residents looking after them," said pulmonologist Richard Schwartzstein, who devised the program, which costs $300,000 annually. Schwartzstein said that when he trained at Beth Israel 20 years ago, an intern and a second-year resident ran the ICU at night, a situation that still prevails at many hospitals.
"Now the residents don't have the anxiety of flying alone," he said, noting that trainees are more likely to seek guidance from a senior doctor who is in the hospital rather than risk calling a sleeping one at home.
While it's hard to quantify, Schwartzstein said he's certain the presence of a seasoned physician has improved patient safety. Among the errors that have been detected early, he said, are misinterpreted electrocardiograms, a missed heart attack and ventilators that were set too high could have damaged lung tissue.
"How many lives have been saved? I don't have numbers, " Schwartzstein said. "But it's clear we've had an impact."
Experts say that the rapid turnover of registered nurses and their increasingly large caseloads contribute to errors; so far there are few studies that prove this. Two reports published in the past six months in the New England Journal and the Journal of the American Medical Association concluded that patients in hospitals where nurses had heavier workloads had a higher risk of dying.
"Anyone who thinks that the nursing shortage and medical errors are not causally related is not in this planetary system," said O'Leary, who notes that an analysis of 1,609 serious errors reported to the JCAHO over the last five years involved nurse understaffing. The problem was also cited in New York's investigation of Hurewitz's death.
According to the AHA, 126,000 nursing jobs are vacant at American hospitals, sometimes as the result of poor working conditions, and 56 percent of hospitals are using temporary nurses who are far less likely to be familiar with a hospital, its staff and its machinery than permanent staff. These factors are all potential sources of error: Studies in aviation show that people trained to work in teams make fewer mistakes than those with no such experience.
Reformer Don Berwick said he remains hopeful that the awareness raised by the IOM report will translate into programs that demonstrably reducte errors. "I don't know why the public isn't more pissed off about this. Imagine what the reaction would be if we had a similar mortality in aviation."