Handling Hospital Errors: Is Disclosure the Best Defense?

Why do patients sue? Bad outcomes and errors in care are obvious factors, but some of the available evidence implicates deficient communication. One attorney explained it to me this way: In over 25 years of representing both physicians and patients, it became apparent that a large percentage of patient dissatisfaction was generated by physician attitude and denial, rather than the negligence itself. In fact, my experience has been that close to half of malpractice cases could have been avoided through disclosure or apology but instead were relegated to litigation. What the majority of patients really wanted was simply an honest explanation of what happened, and if appropriate, an apology. Unfortunately, when they were not only offered neither but were rejected as well, they felt doubly wronged and then sought legal counsel. Reviewing his experience as a malpractice defense attorney, Green (1) estimated that although less than 20% of medical malpractice cases involve negligence, almost all involve a breakdown in the physicianpatient relationship. Patients often form unrealistic expectations because their physicians fail to discuss treatment alternatives. In a review of closed claims in Florida, Hickson and colleagues (2) found that almost 50% of perinatal injury lawsuits were motivated by suspicion of a cover-up or by the desire for revenge. Levinson and coworkers (3) found that primary care physicians were less likely to be sued if they told patients what to expect, encouraged them to talk, used humor, and spent more time with them. An estimated 75% of all malpractice lawsuits involve inpatient or emergency department care; in these care settings, the physician and the patient usually do not have an established relationship (4). Malpractice litigation aims to gain compensation for patients injured by negligent care and to discourage the practice of such care. How well does the current system achieve these goals? A review of 30 195 hospitalizations in New York State showed that in 1984, injury caused by medical management occurred in 4% of hospitalizations; about 25% of these injuries could be attributed to negligence (5). Forty-seven malpractice suits resulted, but only 8 were associated with negligence (6). In a prospective study at a Chicago hospital, Andrews and colleagues (7) found that almost 18% of patients (and almost 50% of surgical patients) had a serious adverse event related to inappropriate care, but only 1% of patients filed a claim for compensation (7). These data suggest that medical errors are common but rarely lead to litigation and even more rarely result in compensation for the patient. Nevertheless, the prevailing legal climate inhibits discussion of medical errors. In a study of residents, Wu and coworkers (8) found that only 50% informed their attending physician of errors and less than 25% informed the patient or the patient's family. Andrews and colleagues (7) found that problems in care that are identified at rounds and at clinical meetings are rarely reported to the risk management department. Although some researchers have argued for disclosure of mistakes (9, 10), many strenuously disagree with this idea (11). Full disclosure could certainly provide an otherwise uninformed patient with a basis for litigation. Would disclosure mollify potential litigants and reduce the amount of awards, or would it lead to a torrent of lawsuits? In this issue, Kraman and coworkers (12) provide encouraging data on an institutional policy that includes early review of injuries, full disclosure of findings to the patient, and fair compensation. This policy, which is currently in use at a Veterans Affairs medical center in Lexington, Kentucky, was developed after the hospital experienced two costly, unsuccessful defenses of malpractice cases. The original policy called for proactive investigation of cases that could result in litigation. However, during such an investigation, the hospital's risk management committee discovered an error of which the patient and family were unaware. The committee took the courageous position that in such cases the hospital was ethically bound to act in the best interest of the patient and to disclose pertinent findings even if they implicated the hospital or staff. Since its adoption in 1987, this practice has led to many settlements. This is not surprising, especially in light of the fact that the hospital offers assistance with filing claims. Of note, five settlements involved serious outcomes that would probably not have resulted in claims if the hospital had not informed patients and families of errors. Unexpectedly, however, this policy has not caused an onslaught of litigation. Early case review has allowed the risk management committee to negotiate reasonable settlements, and payments have been moderate. Overall, the institution has realized a cost savings, in part because of reduced legal expenses. Compared with 35 other Veterans Affairs medical centers in the eastern United States, the Lexington center has an average workload and is in the top quartile for number of claims filed and the bottom quartile for payments. The authors acknowledged that the case for disclosure is not airtight. Their analysis was simple and did not adjust for case mix. However, for the study results to be altered, the Lexington center and the other centers would need to differ greatly in severity of cases. In addition, we do not know how many injuries occurred as a result of negligent care but were not disclosed to the hospital. The authors do not mention the potential benefit of decreasing defensive medicine (changes in medical practice included by the threat of liability) which by some estimates may account for as much as 16% of physician services (13). It is also important to note that to many, the Veterans Affairs system is not representative of the world of medical practice in the United States. Its patient population consists mainly of older men of limited means, a group that may have finite expectations and a low level of litigiousness (14). In addition, government physicians are protected from personal liability and are not named in malpractice lawsuits, although they are subject to reporting in cases of negligence. In the Veterans Affairs system, patients can qualify for compensation without a finding of negligence if their injuries are considered to be service connected. For all of these reasons, we cannot be sure that a disclosure policy similar to the one used by the Lexington center would yield the same results in the world of managed care and indemnity insurance. In the policy described by Kraman and colleagues, the risk management committee is responsible for disclosure. I view responsibility for disclosure as belonging more naturally to the physician. When a physician recognizes an error in the care of a patient, I recommend the following steps. If the patient had an adverse outcome, corrective action should be taken. Hospital risk management personnel should be informed as soon as possible so that they can help examine the causes and circumstances of the mistake. Immediate disclosure should be made if it is needed to secure the patient's consent for additional treatment; otherwise, disclosure may best be made in conjunction with risk management staff. Disclosing an error to a patient or family is similar to breaking bad news. It should involve a straightforward description of the nature of the mistake, its consequences, and the corrective actions that were taken. Expression of remorse and an apology are important. Ample opportunity should be allowed for questions. The physician should recognize that disclosure may upset the patient and should guard against acting defensively. Even if the mistake did not result in an adverse outcome, I still recommend that the near miss be disclosed. Physicians may continue to be reticent about disclosing mistakes because they fear that it will damage their reputations and livelihoods. What measures would make disclosure seem safer? One simple suggestion is that apologies be excluded from admission in evidence (15). However, severing the current link between perceived negligence and compensation for the injured patient seems desirable. The proposed system of enterprise liability would hold the hospital liable instead of the individual physician (16). Changes in methods of patient compensation may also be possible; for example, in Denmark, patients can receive payment from the state if they are injured in the course of medical care (17). Demonstration projects should be authorized and funded to test various no-fault and enterprise liability systems (18, 19). The model used by the Lexington center, which dictates honestly notifying patients of medical errors and offering timely, comprehensive help in filing claims, seems designed to maximize malpractice lawsuits. Instead, it seems to have maximized only the number of patients who are justly compensated for injuries. This seems to be the rare solution that is both ethically correct and cost-effective. However, organizational and cultural changes will be required to realize the benefits of this policy. Hospitals and individual departments must encourage an atmosphere that allows open discussion of errors and recognition of their inevitability. Mechanisms will be needed to ensure that reports are made to the risk management department, and quality improvement efforts should include prospective tracking of error reporting. Educators should promote shared decision making between patients and physicians. Only by changing the expectations of both patients and physicians can we achieve the solutions that will decrease medical errors and their devastating consequences.