A 77-year-old farmer with recurring kidney stones visits his urologist for an annual examination. Prior to seeing the patient, the physician is taken aside by her nurse, who tells her the patient had been in the emergency department the previous night with hematuria. A CAT scan had been done, which indicated that the renal tumor seen on last year’s CAT scan was larger and there were now lung metastases. The physician cannot remember ever seeing the radiology report from last year. To her complete surprise, it is found filed in the patient’s chart. There is no record in the chart that the results were ever shared with the patient. She considers herself extremely meticulous and has never had such an oversight before. The urologist considers what she should tell the patient.
A 12-year-old boy has cataract surgery at a large teaching hospital. At a critical moment the surgeon’s hand slips, severely rupturing the lens capsule. The planned implantation of an intraocular lens has to be abandoned. Instead, the patient will have to use a contact lens. The physician wonders what he should tell the patient and his family about the surgery.
Well-publicized reports of harm occurring to patients as a result of their medical care in the USA (Patient Safety Foundation, 1998), Canada (Sinclair, 1994) and the UK (Smith, 1998) have raised public concerns about the safety of modern healthcare. The US Institute of Medicine report (Kohn et al., 2000). To Err is Human encouraged efforts to prevent patient harm as did the UK Department of Health (2000). In Canada, the release of the Canadian Adverse Events Study (Baker et al., 2004) resulted in the first tangible federal funding for a national Canadian Patient Safety Initiative. Similar initiatives are underway in Spain, Australia, and many other countries.
Adverse medical incidents are by definition injurious or cause a setback to someone’s interests (Davies et al., 2003). Harmful effects of healthcare include recognized natural complications causing some injury to patients, such as a wound infection following an appendectomy. Failure to manage illness according to best practices, such as improper hand-washing techniques or poor instrument sterilization prior to patient contact, are other forms of error, not so much events as harmful processes of care. Errors are considered “preventable” and not primarily a result of the disease process. One definition states that an error occurs when there is “failure to complete a planned action as it was intended, or when an incorrect plan is used in an attempt to achieve a given aim” (Leape, 1994a).
Negligence is established only in a court of law and should be distinguished from “honest mistakes,” the latter being typically errors of judgement (Sharpe, 1987). For example, careful and capable practitioners may make a mistake in diagnosis despite having done everything correctly up to that moment. Such acts may not be considered liable. Perfection is not the standard as “even reasonable doctors make mistakes” (Picard and Robertson, 1996). If a mistake could be made by a reasonably careful and knowledgeable practitioner acting in a similar situation, then the mistake may not be negligent (Kapp, 1997; Wu et al., 1997).
Most obviously, further harm to a patient may be caused by a clinician’s non-disclosure. For example, a patient who is not told that he was given a wrong medication will not know what side effects to look for and might suffer avoidable injury as a consequence.
Failing to disclose errors to patients undermines public trust in medicine because it potentially involves deception (Bok, 1979). Non-disclosure is, potentially, a breach of the physician’s fiduciary responsibilities, a lapse in the commitment to act solely out of concern for the patient’s welfare (Robertson, 1987).
Disclosure of error and adverse events is also consistent with recent trends in healthcare toward more openness with patients and the involvement of patients in their own care (Etchells et al., 1999; Hébert, 2008). Patients are due information about errors out of respect for them as persons. They also cannot properly consent to further treatment unless they know what went wrong.
By the principle of justice (fairness), patients or families, when harmed, should be able to seek appropriate restitution. This may involve seeking monetary recompense or pursuing professional regulatory misconduct hearings. Although some people sue solely for financial reasons, many are disturbed, as Vincent et al. (1994) and others have written, by the absence of explanations, a lack of honesty, the reluctance to apologize, or being treated as a neurotic. Where serious harm occurs to a patient through the egregious and/or willful violation of safe and reliable healthcare practices or through deliberate dishonesty about the incident, families may gain some solace from knowing that the issues are treated seriously and that violators are identified and receive appropriate professional sanction (Shore, 2004).
Finally, non-disclosure of error may also undermine efforts to improve the safety of medical practice in general (Lansky, 2002). If practitioners are unable to be honest with patients or families regarding the untoward event, they are unlikely to be entirely candid in reporting the incident to the appropriate authorities within the healthcare setting. This will retard efforts to identify the faults and weaknesses in our healthcare processes and procedures.
Physicians sometimes argue that non-disclosure may be justified out of concern about needlessly increasing patient anxiety or confusing the patient with complicated information, thus obscuring true choice (Lantos, 1997). This perspective, called “therapeutic privilege,” has not been viewed positively in recent years by Anglo-American courts and should be invoked only in extraordinary circumstances (Kent, 2005). With regards to complex procedures, suggesting that the information would only be confusing to a patient is in direct contradiction to the initial assumption that the patient was capable enough such that the physician accepted his/her informed consent to proceed. Those who support the concept of therapeutic privilege argue most strongly in cases of “harmless” error or “near misses,” and in relation to contentious aspects of the incident, such as “who did it.” It has also been suggested that requiring disclosure of harmless “almost incidents” would threaten to overwhelm the disclosure process with “noise.” Requiring disclosure of “who did it” simplifies what are usually complicated incidents and might result in poisonous finger-pointing, interfering with efforts to improve patient safety. That said, professionals must be prepared to share some of the burden of the incident and not always attribute blame to the “system.”
The law recognizes that physicians may make mistakes without negligence. Indeed it frowns less on the mistakes than on dishonesty and attempts to conceal error – such concealment being incompatible with the doctor’s fiduciary role (Robertson, 1987; Picard and Robertson, 1996). In Stamos v. Davies (1985), a respirologist mistakenly biopsied the patient’s spleen in attempting a lung biopsy. When the patient asked for the results, rather than honestly admitting the error, the doctor replied he had “got something else.” The judge found that the respirologist had breached a duty of disclosure owed to the patient “as a matter of professional relations.”
Punitive damages of $20 000 were awarded in a 1999 British Columbia case in which a surgeon left an abdominal roll in the patient’s abdomen during a laparotomy and presacral neurectomy. The surgeon waited over two months before telling the patient. During that time, he took active steps to cover up the mistake (e.g., by telling the nurses not to make any written record of it). The court described the surgeon’s delay in informing his patient and his deliberate attempts to cover up his mistake as demonstrating “bad faith and unprofessional behavior deserving of punishment” (Shobridge v. Thomas, 1999).
These decisions and similar cases (Vasdani v. Sehmi, 1993; Gerula v. Flores, 1995) suggest that, in Canada at least, a doctor who makes an error in treating a patient has a positive legal duty to inform the patient. We recognize that the adversarial legal climate may be perceived as a disincentive for clinicians to be honest about error. However, the reality of this threat is often exaggerated and misplaced. The Harvard Medical Practice Study found that only 2% of negligent adverse events ever led to actual malpractice claims (Localio et al., 1991). In one study, aggressive error disclosure by a Veteran’s Administration Hospital in Kentucky led to a significant reduction in total dollars devoted to lawsuits (Kraman and Hamm, 1999).
In order to improve patient safety, certain jurisdictions currently have legislation requiring the reporting of incidents to the appropriate health authorities (Oklahoma State Legislature, 2004). More far-reaching examples of such legislation also require disclosure to patients and/or families (Legislative Assembly of Quebec, 2006) and also makes the admissions of error or adverse events safe from legal discovery (Illinois 93rd General Assembly, 2005). Lacking such addendums, legislative efforts to improve the quality of care and encourage reporting may falter.
Since the publication of the earlier version of this chapter in 2001, there has been a proliferation of professional policies addressing disclosure of adverse healthcare events. Disclosure of error is now explicitly addressed in Codes of Ethics for physicians in Canada (Canadian Medical Association, 2004) and the USA (American College of Physicians and Surgeons, 2005). Many professional and regulatory bodies or licensing authorities for physicians have, or are in the process of adopting, policies requiring physicians to disclose adverse events to patients (College of Physicians and Surgeons of Ontario, 1994). In the past, insurers of medical professionals were traditionally too liability-shy to openly advise candor with patients when adverse events occur. This is now changing. For example, the insurance organization for Canadian physicians, the Canadian Medical Protective Association, well aware of new trends in this area, advises honesty, seeing it as the best way to strengthen the doctor–patient relationship and so lessen the sting of any potential malpractice (Canadian Medical Protective Association, 2000).
Hospitals often have policies encouraging the reporting and disclosure of “medical incidents” as part of quality-assurance programs (VA Healthcare Network, 1998; Lamb et al., 2003). Efforts to develop policies of disclosure in the USA have been fueled, in particular, by the requirement of their national hospital accrediting body that all hospitals must have in place a way of disclosing “unanticipated outcomes of care” to patients and/or to families (Leape et al., 1998). Guidelines for and articles addressing the “open disclosure” of untoward medical incidents may be found in the USA (ECRI, 2002), UK (Vincent et al., 1999), Canada (Ontario Hospital Association, 2005), and Australia (Walton and the Clinical Practice Improvement Unit, 2001).
Medical care causing harm is a very significant problem. The Harvard Medical Practice Study from the mid 1980s showed that 3.7% of patients in hospital suffered an adverse event. Approximately half of these events were considered preventable (Brennan et al., 1991; Leape et al., 1991). The Quality in Australian Health Care Study, conducted in the mid 1990s, found that 17% of admissions were associated with an adverse event, with 51% of these considered preventable (Wilson et al., 1995). In Utah and Colorado, 1992 data revealed a rate of injury from medical care of 2.9% (Gawande et al., 1999). Recent studies in the UK, Denmark, France, New Zealand, Canada (Canadian Institute for Health Information, 2004), and Spain (ENEAS, 2005) all show similar rates of adverse events (3–16% of hospitalized patients) and for preventability (20–50% of all such events).
However, inconsistencies in definitions of error and study methodology (Goldman, 1992; Rubin et al., 1992; Smith et al., 1997) should make us question claims that tens of thousands of patient deaths per year result from physician error (Leape, 1994b). The studies cannot say that the patient would be alive today were it not for this or that error. These are not studies of causality.
The real situation might be better (Brennan et al., 1991; Wilson et al., 1992). It also might be worse – hospital records are notorious for underreporting adverse events (Cullen et al., 1995) and there are many barriers for accurate reporting by staff (Vincent et al., 1999). As well, there are other dimensions of error that are not explored by these studies. For example, ambulatory care, the site of many healthcare encounters, remains largely not subjected to monitoring and study. Error studies are limited in that up to half of all US healthcare interventions can be shown to be “inappropriate,” although perhaps not always unsafe (McGlynn et al., 2003).
Regardless of the precise magnitude of harmful medical events, each event raises the issue of disclosure. In one study in which patients of primary care physicians were given hypothetical situations, 98% wanted honest acknowledgement of errors, even if minor (Witman et al., 1996). If not so informed, the surveyed patients indicated that they would be more likely to sue the physician. A recent survey, also of patients in the USA, bears this out: without full disclosure of an error, 34% of those surveyed would seek legal advice; 19.6% stated they would do so even with full disclosure, if the error was life threatening (Mazor et al., 2004).
The Kentucky experiment shows that hospitals can be honest with patients about unanticipated care outcomes and not increase their legal risks. It did not address whether openness either eliminated or reduced lawsuits directed against physicians in particular.
Patients and physicians often differ in what should happen to the erring physician. In a study by Blendon et al. (2002), one-quarter of patient responders favored more lawsuits; 40% wanted the erring doctor to be fined, and 50% favored suspension of licenses. These options were chosen by a very small number of surveyed doctors. In a 2005 study, health plan members were prepared to forgive an erring doctor if the patient withheld critical information from the doctor (93%) but would not do so if the doctor was simply tired (32%), lacked knowledge (24%), or failed to follow up (15%) (Mazor et al., 2005).
One problem with these studies is the assumption of a ready distinction of error – a term suggesting moral failure and fault to some – from other adverse events: if something bad happens, it must be somebody’s fault. Study subjects are asked to respond to a simplified scenario, drained of the complexity the participants experienced at the time of the event. When harm to a patient occurs because of medical care, it is important, where possible, to be sure about what happened, yet it is often difficult to distinguish different classes of event or where fault may or may not lie. Whereas some adverse events are the result of individual error or even negligence, other events arise because of systemic issues (e.g., two medicine bottles with a similar appearance) or the errors and oversights of multiple members of the healthcare team.
Frankly disclosing error can be challenging for practitioners (Hilfiker, 1984; Newman, 1996). Medical professionals have high expectations of themselves and, not surprisingly, find it difficult to acknowledge their errors openly before patients and colleagues (Finkelstein et al., 1997). Clinicians who have observed a colleague disclosing an error to a patient are more likely to do so themselves (Hobgood et al., 2006). This ethical “see-one, do one” is a nice variation on an old medical pedagological technique and clearly needs a little refinement. It is reassuring as it suggests that disclosure is enabled by seeing that your fellow clinicians have done so and have survived. Disclosing such events may be less traumatic if practitioners follow practical guidelines for breaking bad news (Buckman et al., 1998). If uncertain about how to talk to a patient concerning an error, clinicians would be wise to seek advice from their peers, mentors, or skilled hospital representatives before doing so. In any case, early notification of the clinician’s professional insurers is recommended.
In general, it should be assumed that the patient would want full disclosure of errors and adverse events. The professional duty to disclose error is a proportionate duty: as the harm or risk of harm from an event to the patient increases, so increases the duty to disclose (Bogardus et al., 1999). When an error is made but there is no obvious harm, the requirement to disclose error diminishes. For example, minor deviations from a plan of care should not necessarily trigger disclosure; major deviations clearly should. Everything between is subject to discussion and needs to be addressed, but the focus should always be on what is best for the patient as seen from someone in the patient’s position. Clinicians might consider asking themselves: “if you (or a loved one) were the patient, what would you want to be told?”
Disclosure should take place at the right time, when the patient is medically stable enough to absorb the information, and in the right setting. Physicians should take the lead in disclosing error to patients and their families (Levinson et al., 1997). They should try not to act defensively or evasively but, rather, explain what happened in an objective and narrative way, trying to avoid reacting to the charged response that such disclosure might generate. A physician may say “I’m sorry this has happened.” Patients may appreciate this form of acknowledgement and empathy. This may strengthen, rather than undermine, the physician–patient relationship.
If the adverse outcome requires medical attention, practitioners should disclose this and seek prompt help for the patient. Patients may be reassured by knowing that the physician is not only remorseful but also dedicated to rectifying the harm and preventing further harm by a clearly defined course of action. It may be wise to offer to get a second opinion or the option of transferring care to another physician if the physician–patient relationship no longer seems viable.
Meeting with patients, and their families if necessary, in a timely way after an error can help to avoid suspicions about a “cover-up.” Although worrisome and uncomfortable to most clinicians, having lawyers present, if desired by the patient or the family, may help to ensure that all their concerns are expressed and addressed. A healthcare team meeting in advance of a conference with the patient and family should establish that all relevant information regarding the sequence of events leading to the adverse outcome is at hand, mutually understood, and presented as clearly and openly as possible. It will also be important to say what, if anything, will be done to prevent the recurrence of such errors in the future. Patients and families may better accept what has happened to them if they can be reassured that medical care will be improved in the future.
Willful violations of safe healthcare practices and procedures are rare but must not be tolerated (Goldmann, 2006). When practitioners witness errors made by other healthcare providers, they have an ethical, if not legal, obligation to act on that information. Errors causing serious medical harm are ignored to the peril of the profession as well as the public (Irving et al., 1998).
An emphasis on the faults of the system does not abrogate individual responsibility for untoward events. Individuals who exhibit poor judgement and show no signs of insight into their conduct ought to be reported by their peers to the appropriate authorities. Depending on the circumstances and the magnitude of the error, options range from encouraging disclosure by the erring practitioner to discussing the situation with the hospital unit director, department chief, risk manager, a professional insurer, or a representative from a regional health authority. The goal of reporting is not necessarily punitive. Where appropriate, reporting of error may lead to therapeutic attempts to identify treatable causes of error such as substance abuse or psychiatric illness (Shapiro, 2003). Hospitals and healthcare regions should have policies in place to protect such reporters from retaliation.
Medical harm and error are complex and multifaceted problems for practitioners, patients, families, and society. Adequate responses and solutions to them will depend on heightened personal, professional, and cultural commitments to honesty and truthfulness.
Assuming the report of renal cell cancer is accurate and in the right chart, the urologist should first find out what the patient now knows. If the patient is indeed unaware, the urologist needs to tell the patient, at this visit, about the findings on both the initial and current studies. She must be prepared for anger, shock, and disbelief. She should empathically acknowledge that emotion as well as her own upset. The patient may ask about the consequences for his health and his farming business. False reassurances, blame placed on the patient for failed follow-up or on office staff or fellow health professionals (radiology, trainees) will not be helpful. Neither, however, should the urologist shoulder the entire blame as there may be other contributing factors such as systems issues. If the patient wishes, he should be offered immediate referrals that week, if not that day, to another urologist and to an experienced oncologist. The physician should re-evaluate her office procedures and inform the patient of what will be done to prevent similar errors in the future. As serious error is typically emotionally difficult for healthcare professionals, appropriate support should be sought for the consultant through peers, family, friends, and, if needed, helping professionals (Shapiro, 2003).
In the second case, the surgeon should inform the patient and his family about the intraoperative event and the inability to achieve the intended outcome. Although the incident may not have a bad visual outcome for the patient, the surgeon must warn them of the possibility. He should arrange for appropriate follow-up surveillance and tell them what, if anything, can be done should the bad outcome occur. Hopefully, the possibility of the bad outcome was addressed in the initial informed consent. If so, the surgeon is not responsible for contact lens expense. However, he may offer to provide help by making appropriate referrals (e.g., to social work) to address the issues of contact lens cost and management.
An earlier version of this chapter has appeared. Hébert, P. C., Levin, A. V., and Robertson, G. (2001). Disclosure of medical error. CMAJ 164: 509–13.