Originally Published October 24, 1993; Section: NATIONAL; Edition: TWO STAR; Page: A-1; Memo: MEDICATION ERRORS; Day 1: THE OVERVIEW; TOMORROW: Many hands, many mistakes.
Contributing to this series were Tim Dunham, Post-Gazette technology systems editor, and Clark Hallas, who served as a special projects reporter for The Pittsburgh Press until his death in December 1992.
Steve Twedt, Post-Gazette Staff Writer
Every day, patients in U.S. hospitals are killed or crippled because they are given the wrong medication or the wrong dose of medication.
And because hospitals aren’t required to share that information and don’t often voluntarily do it, the same mistakes are made over and over again.
An investigation by the Pittsburgh Post-Gazette shows the same medication errors are repeated, day after day, in one hospital after the other, all over the country. The mistakes often involve the same medicines, the same devices, the same misreadings and the same disastrous results. For example:
— The same feeding tube mix-up that killed 75-year-old retired steelworker Joseph Orlinski at Mercy Hospital Jan. 20 has killed at least nine other patients from Virginia to Arizona in the past eight years.
— The same confusion between dilute and concentrate lidocaine syringes that killed Edward D. Wolf Jr., 61, at St. Clair Memorial Hospital in Mt. Lebanon in 1990 has caused at least 42 hospital deaths nationwide since the late 1970s.
— Similar mix-ups between potassium chloride and other clear liquid medications that killed two elderly women patients — one at North Hills Passavant Hospital in 1988 and one at St. John’s Mercy Health and Hospital Center in 1989 — have been recorded for as long as the last 30 years, resulting in at least five other deaths in Pennsylvania alone in the past six years.
Why don’t hospital personnel learn from each other’s mistakes? Because they usually don’t hear about them. Only a fraction of the errors are ever published in professional journals and even when they are, it’s hard to know where to look, since they may appear in any of the thousands of such journals now published. In rare instances, the local media learn of a serious medication error and disseminate the information. But then, usually only in the local area. Otherwise, the errors remain the hospital’s deadly secret.
“My impression is that hospitals do a pretty good job of detecting errors,” said Don Ruwe, a hospital pharmacist in Edgewood, Ky., and member of the Kentucky Board of Pharmacy. “They just don’t tell anybody.”
And they don’t have to. There are no requirements for hospitals in most states — including Pennsylvania — to share or analyze information on errors. Sometimes a medication error might be referred to a licensing board for discipline against a pharmacist, doctor or nurse. Or a coroner’s office will be told about a death. But those agencies have narrow responsibilities, such as the revocation of a license or determination of criminality. It isn’t their job to look for patterns among errors. And even when mistakes are reported there’s no central agency — governmental or private — that systematically seeks out and analyzes the information to alert other hospitals.
Without formal oversight, the hospitals are left on their own to share information about errors. But they’re not likely to. “It’s something we don’t want to talk about, and I’ve always been troubled by that,” said R. David Anderson, former president of the American Society of Hospital Pharmacists.
“It’s just been sort of a sin committed by hospital administrators to hide that type of information to avoid any potential suit. It’s a defensive measure to prevent it from getting into court, which then becomes public knowledge.” Such secrecy contributes to a growing number of horror stories, experts believe. And mounting evidence gleaned through anecdotes and research is revealing a staggering picture. In a nationwide study of hospital pharmacists conducted by the Post- Gazette, 250 pharmacists estimated there were 16,000 medication errors in their institutions last year alone, 106 of them resulting in patient deaths. And while the typical response from individual pharmacists was 12 errors that caused death, disability or harm in 1992, one pharmacist estimated 3,400 errors in one hospital that year.
Kenneth N. Barker, head of the Department of Pharmacy Care Systems at Auburn University and one of the country’s pre-eminent medication error experts, estimates there is one error per patient per day in every hospital. If his calculation is correct, that would mean more than 200 million medication errors occur yearly.
Relatively few of those errors may carry truly serious consequences. But if the recent experience of New York state, one of the few states that track the outcome of medication errors, holds true for the nation, more than 830 patients in the United States may be dying each year from such mistakes.
Even in the face of those statistics, hospitals don’t share their information. Said Barker: “Hospitals spend a lot more time and money worrying about whether you got properly charged than whether you got the wrong drug.” And they’re not the only ones not paying attention.Who’s watching? As the agency that oversees the safety of drugs and medical devices, the U.S. Food and Drug Administration would seem a likely choice to monitor medication errors.
Indeed, drug manufacturers must report all “adverse events” involving their products to the FDA when they learn of them, but FDA officials acknowledge their reporting system was not designed to track medication errors.
“I don’t think people had thought much about it. People looked at FDA’s responsibility as being that a product be accurately labeled. And that the responsibility for reading and interpreting and properly acting on the label was that of the health professional,” said Dr. Peter H. Rheinstein, director of the medicine staff in FDA’s Office of Health Affairs.
The Post-Gazette also found that key state and federal agencies that do have critical information about the worst medication errors have added layers to the veil of secrecy. For example, a “no publicity” clause is routinely included in settlement agreements handled by the Pennsylvania Medical Professional Liability Catastrophe Loss Fund, a state agency set up by the Legislature to cover the largest medical malpractice claims. That provision means patient families may jeopardize their settlement if they talk. “I’ve seen too many cases where clients who wanted to speak out were put in a very awkward position of being told they had to choose between a very valuable benefit or forgoing publicity,” said Pittsburgh malpractice attorney Jerry Meyers. Such publicity could prompt other hospitals to make sure the same tragedy doesn’t happen to one of their patients, Meyers said. If the details are kept secret, however, “that means it can happen again and again and again, with no one thinking it’s a sufficiently serious problem to adjust procedures to deal with it.”
Details on many medication errors also are locked away in the federal National Practitioner Data Bank, part of the U.S. Department of Health and Human Services. The databank tracks malpractice lawsuits, settlements and licensing board actions against doctors and other practitioners nationwide. Yet it threatens a $10,000 fine for anyone who publicly discloses the information. The courts are one place that may force a medication mistake into public view, if the patient or the patient’s family find out about the mistake and file a negligence suit. But if the case is settled out of court, as many are, the details can be permanently sealed at the hospital’s request, or at least kept out of public view for years. In the absence of any official requirements to share when and how mistakes are made in hospitals, some health-care professionals voluntarily do so. In Allegheny County, Coroner Dr. Joshua Perper has convened local hospital officials to look at medication errors and other mishaps at their centers. ”The idea is to take groups of similar cases and to offer suggestions” on how to prevent similar accidents, Perper said.
Officials from 26 county hospitals attended the closed-door March 9 meeting. The group’s second meeting is set for Nov. 18. Nationally, the U.S. Pharmacopeia, a non-profit organization that sets manufacturing standards enforced by FDA, runs a voluntary medication errors hotline. The hotline was founded by eastern Pennsylvania pharmacists Neil M. Davis and Michael R. Cohen, who have done extensive research into the prevention of medication errors. Reports to the hotline are confidential and Cohen and Davis would not provide identifying information on any specific medication errors for this report.
“We know that we don’t get a very large percentage of the actual incidents because it’s not required,” Cohen said. “These (callers) go above and beyond what they’re required to do. … Unfortunately, there are not a lot of people who take that time.”
Some experts believe voluntary reporting isn’t enough. It’s “a failed approach,” said Dr. David C. Classen, who helps run one of the nation’s most sophisticated computerized patient information systems at Salt Lake City’s Latter Day Saints Hospital. Classen speaks from experience: After his hospital installed its computerized system, it picked up 731 patient adverse reactions — 185 of them medication errors — in 18 months. During the same period, the traditional system of incident reports recorded only nine.
While no comprehensive source on medication errors exists, certain patterns emerged when the Post-Gazette reviewed public documents, professional journals and newspaper accounts around the country. Some of Pittsburgh’s and Pennsylvania’s most nightmarish medication mix-ups during the past four years involved drugs or devices that had been cutting a trail of tragedy for years.
Sometimes, as in Joseph Orlinski’s death at Mercy, the error involved equipment as common as an IV line, and a mixture a child could safely drink. Orlinski, a 75-year-old retired steelworker, was killed in January when a first-year nurse inadvertently attached a feeding tube to an intravenous or IV line at Orlinski’s wrist. The milkshake-like fluid meant to nourish Orlinski instead surged into his bloodstream, clogging his lungs and killing him within minutes. The script sounds hauntingly familiar to the family of Cynthia Tigner. Tigner died in 1985 at age 22 at Ohio State University Hospital in Columbus after a nurse confused her feeding tube and IV lines and put Maalox directly into her bloodstream. Tigner, who had been under treatment for a brain disease, left behind a husband and 15-month-old son. That same year, according to Allegheny County’s Perper, a 77-year-old man being treated for cancer of the esophagus at a Pittsburgh hospital was killed when liquid nutriment was put into his bloodstream. Perper wouldn’t name the hospital or the patient.
And less than six months before that, in the fall of 1984, a similar mistake in Cleveland involving a 64-year-old patient was caught in time and the man wasn’t harmed. Surgeon Thomas Stellato, later reporting the case in a medical journal, recommended special connectors for feeding lines to prevent them from being connected to intravenous lines. More than eight years after his article appeared, Stellato said in a phone interview that “I can’t really say there’s been any obvious changes.” Similar mix-ups surfaced almost every year since Tigner’s death in 1985, including:
— On April 23, 1986, a San Diego nurse administered a mineral supplement into a woman’s central line catheter. The patient died.
— In December 1988, a young boy at Phoenix Children’s Hospital was given cold medication meant for his gastric line in his central IV line and died.
— In 1989, a Richmond, Va., Memorial Hospital nurse inadvertently put liquid nourishment in a patient’s central line and Dilantin, used to control seizure disorders, in the feeding line. No information was available about what happened to the patient.
— On Jan. 14, 1992, a 49-year-old woman at University of Wisconsin Hospital and Clinics in Madison died minutes after a pharmacy technician administered Metamucil — a laxative — in her IV line. Nurses say these mix-ups should not occur. Feeding tubes are wider than IV tubing, and some are coffee-colored to distinguish them from the clear IV lines. In the Mercy Hospital incident, however, a nurse had jury-rigged the feeding tube line for Orlinski with IV tubing before the second nurse made the fatal infusion. In the Tigner death, the two distinctive catheters still had identical stopcocks, or valves.
Usually, such accidents spawned investigations and reforms within the individual hospital to ensure the accident wasn’t repeated there. Following Orlinski’s death, Mercy standardized tube feedings using a pump designed solely for that purpose. Phoenix Children’s began using separate, clearly labeled pumps for IV’s and gastrointestinal feedings. Their feeding tubes now have no port that would allow for mistaken injections. But the changes these individual hospitals have made should be made by all hospitals, said John Senders, a Maine-based university professor and specialist in the study of human error. Senders, who has appeared as an expert witness for doctors and nurses sued for medication errors, said the first such mix-up should have tipped manufacturers and hospital officials to redesign the tubing so that a feeding tube wouldn’t fit into an IV connection.
“This is a trivial engineering issue. What we’re asking the manufacturers to do is what the gasoline people did (after the introduction of unleaded gasoline): Have a small nozzle to fit a small hole,” he said.
“If the bastards can do it there, they can do it here.” But because there’s no central reporting for such errors, the persuasive drumbeat of accident after accident was not heard.
In June 1990, three newborns died at Albert Einstein Medical Center in Philadelphia after a hospital pharmacist accidentally dispensed potassium chloride instead of a 5 percent dextrose solution. The dextrose was to be used to flush the tiny babies’ catheters, a routine chore in the unit. Potassium chloride, used to replenish life-sustaining potassium levels in patients, is among the most commonly dispensed medications in hospitals. But its ability to upset the heart’s rhythm makes it dangerous. And it has a history of being confused with other colorless, odorless fluids:
— Two Allegheny County hospitals, St. John’s and North Hills Passavant, each had a fatal potassium chloride error within four months of each other in 1988 and 1989. Perper would not release the identities of the victims.
— On Jan. 6, 1992, Clifford “Doc” Watson, 93, of Englewood, Colo., died after a nurse with 17 years’ experience mistakenly injected him with potassium chloride instead of the diuretic Lasix. Watson was in Swedish Medical Center in Englewood for pneumonia.
— On Dec. 24, 1991, Richard J. Borden, 58, of Saukville, Wis., died after an emergency room nurse inadvertently gave him potassium chloride. Borden had arrived at the hospital with no heartbeat, but staff members had managed to stabilize him before the potassium killed him.
— In California, three patients in nine months died in separate potassium chloride mix-ups in 1986-87.
— South Carolina had separate potassium chloride deaths in 1988 and 1990, the second one killing a 3-week-old baby. — York Hospital, in York County, had two patients die from potassium chloride mishaps, one in 1987 and another in 1989.
To prevent confusing potassium chloride bottles with bottles of similar- looking liquids with similar caps, U.S. Pharmacopeia told drug manufacturers that starting this year they must use only black caps on potassium chloride. The caps must have the words “Must Be Diluted” imprinted on them to distinguish it from other drugs.
But it took more than 30 years of scattered reports of potassium chloride accidents before the pattern was spotted and “black capping” was ordered.
Even after a U.S. Pharmacopeia committee publicly recognized the problem in 1990, it took nearly three more years to implement the black cap standard. In that time, at least two more people died from similar mix-ups. With the black caps, Davis predicted, potassium chloride poisoning may be reduced by 80 percent. But while all newly manufactured potassium chloride will have the black cap, already-existing stock — much of it with red caps — is still on the shelves, noted U.S. Pharmacopeia consultant Joseph Belson. “The scary thing from my perspective is that there are going to be both labels on the shelves,” Belson said. “You’re probably talking about a couple of years that this stuff could be around.”
Some Deaths Spur Action
With some drugs, there are numerous possibilities for calamitous error.
In March 1989, Michael Sosnoskie of Middletown, Dauphin County, a 3-year- old Down syndrome child with leukemia and heart problems, died at Hershey Medical Center after a mix-up of two cancer-fighting drugs. One drug, vincristine sulfate, is administered by intravenous injection. The other, methotrexate, is injected directly into the spinal fluid. The two similar syringes were on the operating room tray and the doctor mistakenly injected the vincristine into Michael’s spine. “I heard Michael screaming,” recalled his father, Leon, who was in an adjoining room waiting to take the boy home. “I ran back there and said, ‘What the heck happened?’ She said, ‘We had a terrible accident.’ ” Michael lingered for 15 days before he died. He wasn’t the first victim.
Similar reports can be found as early as 1968. And just four years before Michael’s death, a mix-up in Albany, N.Y., between vincristine and vinblastine, another cancer drug that is supposed to be injected into the spine, killed 21-year-old Lillian Cedeno and forced the premature delivery of her daughter, who died a few weeks later. And Michael wasn’t the last. Two years after his death, a mix-up killed Brandon Quintero, 5, when he was given vincristine intravenously, but at the higher vinblastine dose. Quintero was being treated at Duke University in Durham, N.C., for a tumor on his arm. The vincristine mix-ups illustrate the different ways health authorities respond to medication errors.
The public outcry after the death of Lillian Cedeno and her child in 1985 helped push New York lawmakers to adopt a stringent mandatory reporting law for hospital patient mishaps. By contrast, following the death of Michael Sosnoskie in 1989, Pennsylvania officials limited their inquiry to the doctor’s actions. No negligence was found and she returned to practice. And that’s where it ended.
Under Pennsylvania law, hospitals must report to the state only in cases of an “interruption in service,” such as a fire or labor strike. When St. Clair Memorial Hospital had three unusual overdose deaths in 13 months, William White, head of the hospital division for the Pennsylvania Department of Health, found out only after reading newspaper accounts. White’s agency then launched two separate investigations. “If I thought there was a need for it (mandatory reporting) I would push it and start talking it up. At this point, I don’t see it. But I understand why some people do,” White said.
In North Carolina, Quintero’s death in 1991 and three earlier deaths prompted the North Carolina Board of Pharmacy to adopt mandatory reporting for fatal medication errors caused by pharmacists.
“From a public protection standpoint, I don’t see how you can argue against it,” said David R. Work, the North Carolina board’s executive director and president of the National Association of Boards of Pharmacy. Work was instrumental in getting the national association to adopt mandatory reporting of serious errors as part of its Model Pharmacy Act in May 1992. A year later, no other state has adopted it. “It’s just disappointing that no one has taken this up,” Work said. “We certainly can’t be unique. These things don’t just happen between Virginia and South Carolina. They’ve got to happen at other places, too.”
They are happening other places, say medication error experts. But trends in medication errors are not spotted, said Henri R. Manasse Jr., a University of Illinois at Chicago pharmacist who has written about medication errors. “What we need in this country is a national repository on that kind of data, so you know where the hot spots are. If you have data on medication errors in a central repository, then you’re able to see patterns of certain kinds of errors and send up a national alert.” In the meantime, patients will continue to pay with their lives.