Ben Kolb was 7 when he went into the hospital on Dec. 13, 1995, to have a minor ear operation. The boy didn't want to have the surgery, but his mother and the nurses assured him everything would be fine.
It was not fine. Ben went into cardiac arrest on the operating table and died a day later.
What went wrong? It took the hospital three weeks of detective work to figure that out. A laboratory a thousand miles away finally confirmed the awful truth: A syringe that was supposed to contain lidocaine instead contained adrenaline. The adrenaline stopped Ben's heart.
The hospital reached a financial settlement with the boy's family immediately. But the family also insisted that the hospital go public with the case and change its procedures to reduce the chances of such an accident happening ever again -- even though the hospital had been using standard industry practices. Now, at that hospital and many others, intermediary containers are no longer used in transferring medications from a nonsterile to a sterile field.
We know all this because the doctors, nurses, administrators and risk managers at this hospital looked very hard at what occurred and told the world what they learned. They even made a video about it. No one was fired. No one was blamed for this tragic death.
The hospital is Martin Memorial Health System in Stuart, Fla. And nurse Doni Haas, Martin Memorial's director of risk management, will tell you all about it if you ask her.
Don't point; do tell. This is the model that patient safety advocates want to encourage for all U.S. hospitals and healthcare facilities. The idea is to learn from the experience of other industries, most notably aviation. Don't point the finger; instead, dig out the systemic failure. And the Joint Commission on Accreditation of Healthcare Organizations is signing up hospitals for this program, ready or not.
"The underlying concept here is nobody makes a mistake on purpose," says Lucian Leape, M.D., a professor at the Harvard School of Public Health and one of the godfathers of the patient safety movement. "Most errors are made by good and careful people who made a dumb mistake."
The problem is, when people and organizations know they'll be punished for making errors, they're less likely to report them. But fear of punishment, when you examine it, "is a very ineffective way for reducing errors," Leape says.
"Aviation, nuclear power, other hazardous industries that have been successful at getting the error rate down to a very low level, have done it by trying to redesign systems instead of punishing people," he says. "They look at it as an error in the system. How do we do this in healthcare, is the question."
That's what the Joint Commission is trying to figure out.
It's borrowed some nomenclature from those other industries. The term to describe such an accident is a "sentinel event." The Joint Commission defines it as "an unexpected occurrence involving death or serious physical or psychological injury, or the risk thereof."
If one happens at your hospital, the Joint Commission wants you to tear apart all the elements of your procedures, systems and habits to find out where and how the mistake was made, and write it up in a "root-cause analysis." Ask why and then peel the onion back another layer, and ask why again.
The Joint Commission, meanwhile, will add an asterisk to your accreditation status, telling anyone who asks that your organization is on "accreditation watch" until satisfactory completion of the root-cause analysis. (But that policy is changing shortly to remove the public stigma.)
"This whole thing basically emerged because the Joint Commission was caught with their pants down," says Opal Reinbold, a senior associate with BDO Seidman, an accounting and consulting firm in Orange County, Calif. In 1995 and 1996, a series of horrible patient deaths and traumas flared up in newspapers and on television. They happened at hospitals that had good accreditation histories. "They sat down and said, We have to do something about this," Reinbold says. "These bad events in the news are making us look bad, making hospitals look bad." Accreditation watch was the result.
Resistance. People in the accredited healthcare industry are not exactly overjoyed at the prospect of reporting their mistakes to the Joint Commission and performing a root-cause analysis. Hospital lawyers and risk-management experts fear that the peer-review confidentiality protections they enjoy when doing internal quality assurance don't extend to reports they submit to the JCAHO. They think malpractice lawyers could gain access to these reports through the discovery process.
The Joint Commission's board of commissioners considered these complaints in February but reaffirmed its existing policy. It is convening a task force to examine state laws on confidentiality of medical records. In the meantime, it has asked hospitals that submit data on sentinel events to omit the patient's name and promised it will return the root-cause analysis without keeping copies on file (Feb. 2, p. 2; Feb. 9, p. 28; Feb. 23, p. 2).
And healthcare organizations complain that the Joint Commission's policies are unclear, its response is tardy and disorganized, and its reaction is sometimes punitive.
"It was kind of after the fact," says Leslie Lovejoy, quality improvement coordinator at Redbud Community Hospital in Clearlake, Calif., which had a patient death sentinel event in March 1996. "We had already done our process improvement. Then the Joint Commission came along, and we had to translate it to a root-cause analysis."
"They're not real good on following through," says Sue Saydak, quality manager of Veterans Affairs Medical Center in Detroit. Her hospital was placed on accreditation watch just six months after the program's inception. "We still have not brought resolution. We were under the impression they'd come back for a follow-up inspection. But since then we've had our regular triennial survey. We're still somewhat up in the air."
At Lakeland Medical Center-St. Joseph (Mich.), where a patient died after her husband took apart a feeding pump, "the home-care surveyor lost our report. It took a year to find out we were on accreditation watch," says Ruth Hall, director of quality management. "And I have a real problem with their definition of a sentinel event. . . . Do they want us to report every little thing?"
Hermann Hospital in Houston found the process of root-cause analysis highly worthwhile and reduced its medication errors 50% from 1996 to 1997. In August 1996 an infant died after a digoxin overdose.
But the hospital objects to having its accreditation watch status made public. This JCAHO policy is insensitive to or ignorant of the implications of public disclosure on the hospital's mission, public image and market initiatives, says Joanne Turnbull, Hermann's chief quality and utilization officer.
Complaints and reservations notwithstanding, many who participated in this process with the Joint Commission found it of great value. Roland Abellera, director of quality management at St. Bernard Hospital in Chicago, declined to talk about the sentinel event ("We don't want to air out any dirty laundry") but described the root-cause analysis in glowing terms: "It causes you to look at why something happened. And it makes the (hospital) leaders involved. It's more than just a couple of managers talking." His hospital made changes in procedures as a consequence of what it learned.
Steve Watters, director of Mendota Mental Health Institute in Madison, Wis., says the root-cause analysis was "very helpful to our organization. It increased our sophistication in reviewing significant incidents and has given us a useful tool that we continue to use." In Mendota's case, the root-cause analysis didn't establish any cause-and-effect relationship between the two patient deaths and existing procedures. But it still allowed them to identify and implement a number of improvements.
Roseland Community Hospital in Chicago, where two people died in a fire set by a patient in restraints, had an excellent experience with the process "because of the person who did it," says Gloria Harden, vice president for quality management at the hospital.
Sensitive subject. Hospitals and other accredited organizations want to do a good job of patient care, but they almost instinctively resist the pokes and prods from their accrediting overlords. The Joint Commission is about as popular in hospitals as Kenneth Starr is at the White House. You never know what the JCAHO inspectors are going to stick their noses into next.
And while administrators moan and groan in private, they are loath to air their disgruntlements in public. MODERN HEALTHCARE tried to contact all 36 of the organizations that have been placed on accreditation watch. Most organizations didn't want to talk about it, which probably reflects how sensitive this issue is.
And others are apparently in denial. "Well, we weren't really on (accreditation watch). It was very short-lived," says Nancy Baker, quality assurance manager at Presbyterian St. Luke's Medical Center in Denver. "I don't think we have very much to contribute," she says, declining to speak further on the subject.
At UMass Health System-Marlborough Hospital, vice president of nonclinical and support services Katherine Lively sent a letter to MODERN HEALTHCARE Jan. 28 saying, "It is critical that your records be corrected that UMass Health System-Marlborough Hospital is not on JCAHO accreditation watch."
Technically, she is correct. The Joint Commission's accreditation committee removed Marlborough from the accreditation watch list Jan. 22. Specific information on what led to the watch was not available.
John Barr, manager of marketing and public relations at Hillside Children's Center in Rochester, N.Y., says his organization declines to talk to news organizations except "on issues pertinent to Hillside."
Dawn Rice, Hillside's liaison to the JCAHO, is more forthcoming. "The Joint Commission's policy changed because of us, you know. We reported ourselves, and we were placed on accreditation watch. We felt they shouldn't have. Under the new policy, we wouldn't be on accreditation watch."
Avoiding the stigma. This reluctance to be associated with a sentinel event is probably why the program hasn't worked very well so far, the Joint Commission thinks.
"It was clear fairly early on that we were having some problems with it," says Dennis Barry, president and chief executive officer of Moses Cone Health System in Greensboro, N.C., and an AHA commissioner on the board of JCAHO commissioners. "It had to make a transition from being a policy that took a punitive perspective as opposed to a positive, collegial perspective."
So in November the board of commissioners changed the policy to encourage self-reporting of errors. Now if your organization notifies the Joint Commission of a sentinel event within five days and completes a proper root-cause analysis within 30, the JCAHO won't put you on accreditation watch. You're in a safe harbor.
Consumer advocates and Joint Commission critics argue that this makes hospitals less publicly accountable for their actions. But patient-safety researchers argue that the public shame approach just doesn't yield the desired end.
"I totally understand the concern of critics of healthcare, who worry that this could result in inappropriate forms of secrecy or exculpation of people," says Donald Berwick, M.D., president of the Institute for Healthcare Improvement. Keeping the material under lock and key seems counterintuitive to basic American values of openness and accessibility.
However, he says, "once you set up an environment where people feel comfortable bringing errors forward, you get a chance to grapple with patterns of error that you don't get in systems where you affix blame."
In other complex industries where the penalties for error are high, it has been found that the best method of continuously increasing safety is to create a system of blame-free reporting, he says. "There just are no strong counterexamples to that. And there are very strong success stories."
Aviation model. In 1975 the Federal Aviation Administration created the Aviation Safety Reporting System. It's a database of reports sent in by anybody who observes or is involved in an aviation incident that represents a threat to safety.
"For every accident, there are large numbers of incidents of similar type," explains Charles Billings, formerly chief scientist at NASA's Ames Research Center, who helped devise this system. Billings defines an incident as "something went wrong, but the airplane didn't get bent, and the people didn't get bent. . . . It has been our belief for a considerable time, the way to prevent accidents is to understand why incidents occur and to understand that in some depth."
And it's a lot easier to harvest knowledge from an aggregation of incidents rather than accidents. For one thing, the participants are still alive. "Pilots are usually the first at the scene of an aircraft accident," Billings notes wryly. "They are often very hard to talk to after that."
Key to the success of the Aviation Safety Reporting System is its promise of confidentiality to the reporter. This is so critical an element in the program that the FAA isn't even allowed to receive the incident reports. They are sent to NASA, which codes them and strips them of all identifiers before entering the information into a database. Samples of information gleaned from these reports can be found on the World Wide Web at http: olias.arc.nasa.gov/asrs.
More than 300,000 reports have been made to the database in 23 years. Many of the patterns detected in reported incidents have led to safety reforms.
"Whether that can be made to work in medicine is another question," Billings says.
Patterns. The Joint Commission's database of sentinel events is small -- 194 events reviewed -- but a few patterns are emerging.
The most common category is medication errors. Second is inpatient suicide. "That was a surprise to me," says Richard Croteau, vice president for accreditation services.
The JCAHO also is breaking down the incidents by type of facility. At least seven of the organizations on the list are mental health facilities. "That may be a reflection of the incidents we've noted of suicides," Croteau says. "About half of the suicides are in psych hospitals. The remaining half are in psych units of regular hospitals."
One of the troublesome issues still to be worked out is the vagueness of the definition of sentinel event. "The definition is, `major loss of life and limb, or risk thereof,' " says Reinbold. "I don't know how you can identify what the `risk thereof ' is, unless it's a near-miss."
So, are hospitals supposed to report all their near-misses within five days or risk the wrath of the accreditation committee if it comes up in a survey that they didn't?
People in the industry suggest the Joint Commission has given inconsistent responses when pressed to clarify this question.
In Croteau's view, it should all be reported. "There's so high a risk of an adverse outcome if it happened again that you treat them as sentinel events."
Yet in so doing, the Joint Commission contravenes one of the critical principles of the aviation reporting system, the distinction between accidents and incidents. Accidents in aviation are handled by the National Transportation Safety Board, and they very quickly become public knowledge.
In aviation incidents, by contrast, as Billings observes, "the issue of liability is moot. Nobody has suffered harm, aside from loss of face."
The Joint Commission is putting them both in the same database. And the accidents are being covered with the same cloak of confidentiality as the incidents.
Croteau says the Joint Commission does not look at it that way. "Reducing risk can be achieved by looking at both -- events that have in fact a bad outcome, as well as those that by luck didn't have a bad outcome."
The Joint Commission also has considered using the hands-off, third-party model pioneered by the FAA and NASA, "more as a research activity to gather information about the frequency and nature of events." Croteau says it's possible that people would more willingly report adverse events if they talked to a neutral group, not their accrediting agency. "It's still on the table," he says. "It's just that as this thing is evolving, we haven't developed a system of that nature."