Ethics of error disclosure: When to say, “I'm sorry”


  • Bryn Nelson PhD

Part 1 of a 2-part series grappling with the ethics of disclosure amid a push toward more transparency.

A 33-year-old woman tests positive for cervical cancer, and when her physicians review slides of previous Papanicolaou tests, they discover abnormalities that were missed. What should they do?

The Institute of Medicine's 1999 report “To Err Is Human: Building a Safer Health System” acknowledged the reality that everyone makes mistakes. Nevertheless, the report stressed, up to 98,000 people die in US hospitals every year due to preventable medical errors. The eye-opening numbers and focus on preventable harm has helped to train more attention on errors and error disclosure, a trend that has accelerated with the more recent emphasis on health care quality.

“Transparency is getting much more airtime,” says Denise Dudzinski, PhD, a bioethicist at the University of Washington in Seattle, who studies error disclosure. Counterarguments such as, “What patients don't know won't hurt them,” are losing currency, she says. “The trend is now more toward, ‗We owe them an explanation.‘ I think that's really, really good.”

Disclosure is often a difficult process, and experts say greater transparency is still hampered by systemic barriers such as the fear of incurring a lawsuit, damaging a provider's professional status or reputation, or harming the physician-patient relationship. Furthermore, not everyone agrees about when an adverse event rises to the level of a reportable error. One traditional school of thought still contends that only those discovered errors that have a bearing on patient management should be reported. However, the view increasingly pushed by ethicists, insurers, and quality experts argues that all errors should be openly disclosed to patients to identify system failures and improve patient safety. In the long run, they say, honesty is the best policy.

Jay Kalra, MD, PhD, a pathologist at the University of Saskatchewan in Saskatoon, Saskatchewan, Canada, and an expert on medical errors, disclosure, and quality management, believes physicians have a duty to disclose errors as a standard of care. “For me, it's a very big ethical issue,” he says. “I'm a physician and I have to make sure that I treat my patients in an ethical manner so I provide them the best possible health care.”

Errors tend to fall into 2 main categories. An error of execution refers to “the failure of a planned action to be completed as intended,” whereas an error of planning involves “the use of a wrong plan to achieve an aim,” according to the Institute of Medicine report. Health care institutions once viewed errors mainly as blunders that might be corrected by retraining or replacing individuals. Now, Dr. Kalra says, researchers widely believe that most errors reflect systemic flaws that may require broader institutional changes and the creation of a culture of safety, a view he has embraced in a recent book.[1]

Correcting a mistake and ensuring it isn't repeated, however, first requires recognizing it. “If a patient was going to go for surgery on the right knee and ended up having surgery on the left knee, you would know immediately,” Dr. Kalra says. However, in a medical laboratory, the likelihood of error recognition is usually much lower, and interpreting whether a cytology slide truly shows cancerous cells can sometimes be a judgment call.

Shades of Gray

In the United Kingdom, National Health Service policy mandates the disclosure of any discovered Papanicolaou test abnormality. The policy also directs pathologists to review past cytology slides and the biopsies of any patient who develops cervical cancer, even if it is the earliest stage possible. If pathologists attach a “negative” finding to those historical slides, they are subsequently reviewed by a quality assurance team of experts. “So, if anything, people tend to overreport rather than underreport them on review,” says Amanda Herbert, MB, BS, a histopathologist and cytopathologist at St. Thomas' Hospital in London, UK.

That system, she says, can lead to problems: “A bad review of cytology slides showing ‗missed‘ abnormalities can be an invitation to litigation unless written carefully,” she says. Therefore, putting the slide review in context is critical when writing an audit report, a delicate balance that Dr. Herbert highlights in an editorial on disclosing cervical cancer audits.[2]

Varying shades of gray can likewise fill the space between a clear misinterpretation and a failure due to the inherent sensitivity and specificity of a test, or between a diagnosis and the subsequent course of action. Elliott Foucar, MD, a pathologist in Albuquerque, New Mexico, recalls seeing a pigmented lesion that had been diagnosed as early melanoma by another pathologist. Consequently, the patient was treated very aggressively, “which I would argue was a treatment screw-up,” he says.

When an expert reviewed the case years later, the patient's diagnosis was reversed to benign. “Of course, years later, with no evidence of subsequent disease and very aggressive initial therapy, it is a lot easier to make a benign diagnosis on these problematic lesions,” Dr. Foucar says. “The original pathologist got sued.”

That kind of scenario can make a retrospective review of older slides a particularly fraught exercise. How do clinicians avoid retrospective bias, for example? Is the exercise part of a well-organized quality-control activity that will improve practice by productively bringing errors to light? Or is the intent to single out individuals for blame?

The United States, Canada, and many other countries lack uniform disclosure policies. Nevertheless, experts are crafting more effective error prevention and response strategies. Proactive in-house plans, for example, can promote clear lines of communication, fairness, and a mutually agreed-on roadmap for how to promptly respond without resorting to unproductive finger pointing. Dr. Kalra's group, for example, has developed a comprehensive “no-fault model” that he believes is helping to advance a culture of safety.

The key is to be prepared. “Don't think it's never going to happen. It's going to happen sometime,” says Fred Rodriguez Jr, MD, professor of pathology at the Louisiana State University School of Medicine in New Orleans (Dr. Rodriguez emphasizes that his views are his own and do not reflect the official viewpoint of Louisiana State University).

A Near Miss

Carefully constructed quality control and assessment plans can pay big dividends. Among some groups, one pathologist has the primary reporting responsibility, whereas any primary diagnosis of malignancy requires a second board-certified pathologist to review and confirm the diagnosis. Other groups use a multiheaded microscope to perform a combined sign-out with multiple colleagues, whereas the practice of having pathologists periodically review their peers' cases at random is likewise catching on.

Such peer review may also lead to significantly lower error rates. Dr. Herbert and her colleagues view “rapid review” as one good method for detecting near misses. In her experience, she says, the technique detects high-grade abnormalities in approximately 1 in 1000 slides initially classified as “negative” or “inadequate.”

As part of an in-house review, a pathologist may catch a colleague's oversight before any report is issued. Should the discovery be disclosed to the patient anyway? If a patient witnesses a near miss, Dr. Dudzinski says physicians have an obligation to disclose because that patient is owed an explanation. However, if a mistake is caught and corrected before it leaves the laboratory, she says, patient disclosure is unnecessary. “Here is why: the whole purpose of having these checks and balances and these quality measures in our institution is precisely to protect the patient from being on the receiving end of a harmful error,” she says. In other words, the process worked.

Dr. Rodriguez agrees, but cautions that near misses do not necessarily mean that everything is “hunky-dory.” A near miss, he says, may still identify opportunities for quality improvement and require some internal housekeeping. “You've got problems, but it's stuff that you can look at internally,” he says. “You're not altering, changing, or impacting anything outside the lab.”

That calculus changes for a corrected pathology report, even if it does not lead to a change in treatment. In that case, Dr. Rodriguez believes the report should be openly disclosed and closely monitored. There may be a settlement involved, but Dr. Rodriguez and others say forthrightness and honesty, and yes, an apology or expression of regret, can often help head off litigation and mend the physician-patient relationship. “The pathologist shouldn't hide behind the ‗paraffin curtain,‘” he says. “It's better to be upfront. If you've messed up, admit you've messed up.”