6
Physicians and Quality of Medical Care

As we noted in chapter 3, any profession must have as a high priority the provision of the best possible service. Since clients can rarely judge the quality of professional service, professions must police themselves. Indeed, such self-regulation is the hallmark of professions. The highest quality service is a universal professional ethical imperative.

The medical profession meets these standards, or at least seems to do so in the abstract. It places a high premium on excellent and conscientious care. Physicians are recruited from the brightest college students and then rigorously trained in biomedical and clinical sciences. Arduous postgraduate training with rigid supervision by attending physicians reinforces the importance of quality medical care. Physician error is carefully reproached by other members of the profession.

The professional orientation toward quality medical care fits neatly with the traditional ethics of medicine. Since the patient's well-being is the primary objective of the ethical physician, provision of good quality care is a salient concern of the profession.

This rendition of the relationship of traditional medical ethics to quality medical care is an ideal. It also reflects the same attributes of the profession we have noted in previous chapters. Ethics provides a guarantee, this time of good quality care. Yet the patient and the public are unable to evaluate the extent of the physician's fulfillment of her professional duty. The public is in effect again asked to trust the

profession, because the profession is duty bound to provide excellent care. There is no regulation other than decentralized oversight by physicians themselves.

As medicine is drawn into the liberal state, will this societal attitude toward good quality care persist? The answer is no. Indeed, the law concerning patients' reactions to the quality of medical care, medical malpractice law, reveals that over the last thirty years, patients have become less and less willing to accept doctors' offers of trust.1 Injured patients have increasingly sought relief from the courts for care that has been substandard. Under the theory of tort law, this creates incentives for physicians to provide better-quality care. That judges have been increasingly willing to hear liberal theories of negligence indicates that in medical malpractice, as in informed consent, courts are striving to reintegrate medicine into the liberal state by enforcing the law as integrity rather than accepting physicians' self-regulation. Thus a professional perception of a crisis in malpractice litigation can be interpreted, at least in part, as physician resistance to the changes demanded by society.

Medical ethics as just doctoring must consider the role of medical malpractice law in quality assurance. Rather than merely resisting changes in malpractice law, physicians must recognize increased litigation as a given. The liberal state requires evidence of good quality. care. Therefore ethical physicians must think about new methods for providing high quality care as an answer to rising rates of malpractice litigation. More important, this quality assessment must be made public, in keeping with the assumptions of the liberal state. In this chapter I explore physicians' reactions to malpractice and the notion of quality. assurance as a priority for medical ethics today. Of note, in this chapter, as in the last, we move from discussions about a theoretical liberal state, into analysis of our particular American liberal state.

The (Short) History of Quality Assurance

While the professional commitment to good quality care is as old as physicians' codes, the history of organized efforts at quality assurance is quite short. Ernest Codman is now celebrated as the pioneer of quality assurance. As early as 1914, this Boston surgeon pro-

posed evaluating the end results of therapy to improve the quality of care.2 His views were not, however, acceptable to the majority of surgeons at the Massachusetts General Hospital, and he had to leave that institution to start his own hospital. His work was somewhat ignored, as was that of other pioneers such as R. I. Lee and L. W. Jones.3 When A. Donabedian and others began to undertake a systematic assessment of the quality of care in the mid-1960s, they moved in largely uncharted waters.4 Indeed, the last twenty years have seen several false starts in the quality assurance movement.5

A good deal of the blame for the stunted growth of serious and systematic quality assurance must lie with the medical profession. Charles Rosenberg and others have carefully documented the manner in which the medical profession gained authority over the administration of hospitals in the early twentieth century.6 Until that time, hospitals had been controlled, at least to a certain extent, by their lay boards. However, by the 1920s, the overwhelming majority of hospitals in this country were not-for-profit institutions with dual forms of governance. A board of trustees and a hospital administration oversaw the business aspects of the hospital. The medical staff had its own separate governance and controlled the manner in which care was provided.7

This dual form of government was countenanced and indeed required by those who regulated health care. Most hospitals were licensed by state agencies, often the state department of health. However, few actually tried to provide oversight or to evaluate the performance of hospitals. Instead they relied on the Joint Commission for the Accreditation of Hospitals (JCAH) to perform this service. The JCAH had been born out of a power struggle between physician organizations. In 1912, the newly founded American College of Surgeons had undertaken to standardize surgical practice. It established the Hospital Standardization Program (HSP), which stressed self-regulation by the medical staff. Timothy Jost relates that between 1918 and 1935 the percentage of hospitals with an organized medical staff increased from 25 percent to 90 percent.8 By 1950, the complexity of health care had overcome the resources the American College of Surgeons could make available to the HSP. After a great deal of fighting among various professional groups, the American Hospital Association, the American Medical Association, the American College of Surgeons, and the American College of Physicians all coordinated their efforts and the JCAH was organized.

The JCAH adopted the HSP regulations that were in place in most modern teaching institutions in the late 1930s. The JCAH focus, however, was slightly different in its emphasis, for instance on the importance of good medical record keeping. It also stressed that physicians were to control the membership of the hospital staff, and that physicians' were the final authority on decisions regarding quality. The major forums of quality assurance were the clinical conferences, in which physicians gathered together to review pathological findings and assess errors in diagnosis and treatment, and the tissue committee meetings, in which the results of surgical procedures were reviewed by senior surgeons.9

These formal evaluations at weekly conferences remained the cornerstone of physician-based quality assurance. Such methods relied overtly on the commitment of physicians to better quality care. Complications and errors were to be discussed openly, but only for the benefit of the medical staff. There was no requirement to make any of the reports at "morbidity and mortality" rounds public, or even to share them with the hospital.

Physicians who erred frequently could be sanctioned, but again this was completed solely by the medical staff itself. The sanctions took the form of restrictions on admitting privileges, or, in the worst cases, expulsion from the hospital staff.10 Again, the hospital and the state authorities were rarely involved in such cases. Indeed, other hospitals were usually not warned about physicians who had performed poorly enough to be shown the door. Perhaps this was only fair, as many of the expulsions from staff no doubt had less to do with incompetence than they did with perceptions of economic competition.

In any case, these methods of quality assurance were enshrined by the early 1940s, and two generations of physicians matured believing that these were the only real means to achieve high quality care. The JCAH approach was anointed by the federal government when it made JCAH accreditation the major grounds for certifying hospitals as participants in the Medicare program.11

The system of informal sanctions and treatment of physician error is well captured in Charles Bosk's Forgive and Remember.12 Bosk documents the manner in which senior attending surgeons at a teaching hospital discussed and evaluated errors made by their house staff. He reveals that the sanctions were informal, and that after a lesson was learned, the error was forgiven by the other physicians. The trainees were encouraged to do everything possible for the patient, and this

kind of effort made error forgivable. Perhaps most important, the surgeons were trained to refer those cases beyond their own individual level of skill to more senior physicians. Physicians' ethical codes regarding quality of care was to "put the patient first" and to recognize one's own limits.

From Bosk's point of view, this informal system is rational and defensible. His study, however, centered on top surgeons at an academic medical center. One can imagine that not all medical institutions are populated by such conscientious individuals. In hospitals where the skill level is not so high, and the "patient comes first" ethic not quite so established, one can imagine that quality might be ignored completely, for the system puts all the prerogatives in the physicians' hands.

Indeed, this system of informal sanctioning by physicians is another example of their power. Physicians had demanded and hospitals conceded total control over quality assurance. Rather than adopt the statistical methods of Codman, featuring hard endpoints and publicly available information, physicians chose a more private approach. Again, patients were asked to trust physicians, for physicians' paternalism would assure that their colleagues who made mistakes would be sanctioned. The negative freedom of the patient was sharply circumscribed. The patient could not become a rational consumer of medical care because evaluations of physicians by their peers were hidden from view. Thus medical ethics was again to play a critical conceptual role as justification for the lack of social oversight; physicians would provide good quality care because they were devoted to the patient.

The ethics of commitment to good quality care is in many ways rational, for it often brings about better care. Its peculiarity becomes clear only when one considers its lack of openness. An episode from my own training illustrates my point. A patient I had followed in my clinic for several years, and who was very healthy except for a history of diabetes mellitus, was brought to the emergency room in a semi-comatose state. His diabetes appeared to be under control, but many of his blood chemicals were abnormal. He began to suffer from shock while in the emergency room. The physicians on duty called me, and together we puzzled about the cause of the shock. We could not define the etiology, and although we employed all possible supports, the patient died.

As he neared death, a laboratory test, sent off inadvertently, demonstrated that he had taken an aspirin overdose. Indeed, upon receipt

of this information, it became clear that the patient's state was well explained by the aspirin overdose. While he had no history either of aspirin use or of suicidal intention, we all felt we should have made the diagnosis more promptly. We felt this even though a very fast diagnosis probably would not have made any difference, as the patient's state was well advanced by the time he was brought to the emergency room.

The case was reviewed at our weekly morbidity and mortality conference in which we went over the deaths for the previous week. Several of the senior physicians asserted the diagnosis should have been made more quickly and pointed out the salient features of aspirin poisoning. The entire staff could learn from the case. Although no one stated this, all of us involved felt we had made an error in judgment. We wished we had made the diagnosis more quickly. Indeed, those of us involved in the case, and many of our colleagues, learned a great deal about salicylate poisoning, and I doubt I will fail to consider it with similar patients in the future. Therefore, my personal experience, as well as that of other physicians, is that this kind of informal quality control can work.

But does this approach always perform well? Do informal sanctions prevent most substandard care in hospitals? There are few good answers to these questions. The most widely quoted estimates of the incidence of adverse events (defined as injuries suffered by patients as a result of medical management rather than of the process of their disease) and negligent care (defined as care that fails to meet the community standard) were developed over twelve years ago in a California study of 20,000 nonrandomly selected medical records.13 This study revealed that 4.65 percent of hospitalizations resulted in adverse events. Nearly 10 percent of these adverse events resulted in deaths. Negligent care resulted from.79 percent of hospitalizations. These figures would suggest there is a great deal of room for improvement in our approach to quality assurance. Many of these adverse events, and presumably all those arising out of negligence, could theoretically have been prevented. Other reviews of medical records by physicians to identify poor quality care or adverse events have been restricted to nonrandom samples of a few hundred records, but they also suggest there is a great deal of preventable error.14

Above and beyond these statistics, and what they suggest about the efficacy of informal reviews, the current approach to quality assurance seems unsatisfactory because of its private nature. Consider again my

patient who died of salicylate poisoning. He was one of the most solitary figures I have ever met. He was not born in this country, and he had no family here. None of his neighbors knew him. As near as I could tell, he had no estate. Nonetheless, his case has often caused me to think about the work I do. What if he had been brought to the emergency room sooner, and could have been saved? What if our failure to make the diagnosis had clearly led to his death? And what if he had left a family? Would it have then been appropriate to say nothing to the family, to leave our error in the silent world of doctor and patient? Would not the public expect some discussion of the matter? Does it not serve doctors' interest to keep quiet about these issues? If we are really committed to the patient, should that involve informing the patient when errors are made? Suppose I had not been in training. Would there have been the same kind of attention to the care I and others provided? Or would it have simply been possible for us to forgive each other and forget the episode? Do all hospitals pay as careful attention to these issues as the one in which I trained? Would not more formal mechanisms be appropriate to assure that hospitals and medical staffs are conscientious? Do physicians have some ethical duty. to see that this is the case?

Patients might think so. Indeed, just as patients have grown dissatisfied with the traditional handling of information in the doctor-patient relationship, there is good evidence that they have grown skeptical of the profession's reassurances in regard to quality medical care. As with informed consent, the patients have petitioned the courts, this time for supervision of care in the form of medical malpractice litigation.

Tort Law and Quality Assurance

Although the concepts seem foreign to many physicians even today, medical malpractice law has two roles. The first is compensation of victims of accidents. Not all accidents, however, lead to compensation under a tort regime; in fact, only those accidents that were caused by the negligent behavior of another are compensable. Our common law approach ties compensation to negligent behavior. The primary reason for this linkage is the second major function of

tort litigation, that of deterrence. The theory of tort law is that the person causing the tort, or the tortfeasor, is deterred from acting in a negligent manner in the future by the economic penalty of the tort award to the successful plaintiff.

Although many of these concepts regarding torts were discussed by Oliver Wendell Holmes over a century ago, they have undergone a reevaluation in the last twenty years.15 Guido Calabresi's pioneering work on the theory of tort law has spawned a number of economic analyses of tort law, all centering on the value of deterrent and compensation effects.16 This renewed academic interest in tort law is no doubt related to increases in the amount of tort litigation in the 1960s and 1970s.

Trends in jury awards for malpractice claims and product liability suits indicate that there has been a substantial increase in successful tort litigation in this country in the last thirty years.17 George Priest has argued fairly persuasively that, at least with regard to product liability, much of the increase in tort litigation can be traced to the ascendance of the theory of enterprise liability, which provides in its simplest form that businesses should be responsible for the injuries caused by their products.18 Priest suggests that enterprise liability was championed by legal academics who then influenced several powerful courts in the mid-1950s, resulting in a series of landmark legal decisions.19

In Priest's view, one of the most influential of the legal academics, along with Fleming James, was Friedrich Kessler,20 who maintained that the large capitalist enterprises had grown too powerful to be controlled by contract law. Contracts between individuals, the ideal of the nineteenth century liberal state, were no longer possible in the twentieth century. Therefore tort law had to be bolstered by changes in doctrine such that individual plaintiffs could more readily seek redress against concerns like automobile manufacturers or soft drink bottlers. Concepts such as risk spreading and enterprise liability were proposed to restore some balance between litigants and to provide appropriate deterrence signals to industry. Priest argues further that the efforts of academics like Kessler and James were largely successful, and that much of the increase in tort litigation can be traced to the reforms of the common law undertaken at their urging.

This theory of the expansion of tort law bears a great deal of resemblance to our theory about the law as a means to fortify the values of liberalism in relation to those of traditional medical ethics. Just as

I have argued that patients turned to the law when faced with the authoritative claims of physicians, so too can changes in tort law be seen as an attempt by judges, guided by legal academics, to equalize power between plaintiffs and large corporations. Devices such as enterprise liability make defendants fully accountable for injuries sustained by plaintiffs; they reiterate the importance of the individual's negative freedom. Arguments such as those put forth by Kessler are fundamentally liberal.

The increase in tort litigation was not confined to product liability. Medical malpractice became another growth area for personal injury lawyers. The reasons for this are similar to those underlying product liability—judges made it easier for plaintiffs to sue successfully. This resulted, I believe, from the court's perception that the medical profession had grown too powerful, and that the law as integrity in the liberal state demanded change.

There have been a number of explanations for the changes in malpractice litigation that have led to increases in premiums.21 The overall increase must, however, be laid at the bench of the judges who made it easier to win suits. In the early 1950s plaintiffs faced intolerable barriers to winning a suit against physicians because physicians assumed a code of silence and refused to testify against one another. A plaintiff, who had been unconscious, found it was impossible to develop evidence of operating room mismanagement. Locality rules required that experts testifying against local physicians must be familiar with local standards. This made it all the more difficult to identify. appropriate experts. Moreover, courts slavishly applied the rule of medical custom whereby doctors alone were qualified to determine whether care was negligently performed. Courts also refused to find both doctors and hospitals liable simultaneously, nor would they hold a hospital accountable for physicians practicing there.

In essence, the barriers to tort suits kept doctors insulated from outside regulation. Since doctors largely controlled the quality assurance efforts at the hospital, and since there was little state oversight, tort litigation was the only way that patients could demand good quality care. The difficulty of tort litigation and the physicians' control over hospitals consolidated the positive freedom of physicians and removed yet another aspect of medicine from the values of liberalism.

This began to change in the mid-1940s with the California Supreme Court's ruling in Y barra v. Spangard.22 Mr. Ybarra was admitted to a hospital for an appendectomy during an acute attack of ap-

pendicitis. After awakening from the anesthesia, he noticed for the first time pain and decreased neurological function in his right arm and shoulder. Any physician would have recognized his symptoms as a brachial plexus injury resulting from poor positioning of the arm during anesthesia. The defendants asserted that there was no evidence of any act of a particular defendant that had caused this problem. None of the defendants would testify as to what happened in the operating room.

To break this deadlock, the plaintiff asserted the doctrine of res ipsa loquitur. This ancient doctrine had long stood for a common sense inference from plain facts. The California Supreme Court was willing to adjust this meaning, converting it into a burden-shifting device. Henceforth, the court reasoned, that when a healthy portion of the body was injured during medical care, "such circumstances raise the inference of negligence and call upon the defendant to explain the unusual condition."23

The defendants argued vociferously that res ipsa should not be used in medical cases because the doctrine tends to implicate all involved in the care of the patient, without showing that any one in particular was negligent. The court was unyielding. The decision notes that if the court accepted defendants' contentions, there would never be any compensation in medical cases, as patients rarely have access to the evidence of negligence. Thus the Y barra decision was a milestone in malpractice litigation, representing radical use of an old doctrine to empower patients.24 The code of silence among health professionals was effectively broken and courts proved increasingly willing to allow lay judgments of the quality of care.

The doctrine gained even wider use when the California court decided nearly two decades later that res ipsa loquitur could apply even to cases concerning complex medical issues if an expert testified that a poor outcome would not have happened in the absence of negligence on the part of the physician. In the decision of Quintal v. Laurel Grove Hospital, the California court ruled that an instruction on conditional use of res ipsa loquitur was appropriate when cardiac arrest occurred during administration of anesthetic agents, even though arrest is a known, but rare complication of such agents.25 In this case, the expert for the plaintiff testified that arrest rarely occurs in the absence of negligence, The use of res ipsa in such situations further shifted the burden onto defendants to prove they were not negligent.26 Courts were thus using the doctrine of res ipsa to overcome the patient's lack

of medical knowledge and to address "adverse events" that might only tangentially be related to poor quality, care. The burden was placed on physicians to prove the high quality, of their care.

Courts have not, however, taken the further step of eliminating their reliance on medical custom. Unlike the Canterbury decision in the area of informed consent, a layperson standard has not gained ascendance in medical malpractice.27 At least one court, the Washington Supreme Court has approached this threshold. In its decision of Helling v. Carey, the court ruled that even though expert testimony revealed that the standard of practice in Washington was not to test patients under forty, for glaucoma, a patient who was less than forty who became blind from glaucoma that went undiagnosed for several months, could recover damages.28 In effect, the court substituted its own lay judgment that the simple test to detect glaucoma should have been employed in this case.

Strong pressure from physician groups caused the Washington legislature to overturn this decision, and no other courts have attempted to abolish medical custom, even though the expert testimony in Helling had come under attack.29 Nonetheless, other changes in doctrine have undercut the rule of medical custom in ways similar to that of the changes in res ipsa loquitur.

For example, courts have largely abandoned the locality rule. As noted above, the locality, rule required that experts testify regarding the standard of care in the particular geographic area wherein the injury arose. In cases such as Brune v. Belinkoff, courts overturned this rule, easing the burden on plaintiffs trying to find experts to testify on their behalf.30Brune v. Belinkoff involved an injury to a patient's nervous system after he received too much anesthetic agent into his spine. The defense claimed the standard of care was different in New Bedford, Massachusetts, than it was in Boston fifty miles away. The Massachusetts Supreme Court protested that there was little sense in "Balkanizing" the medical profession; moreover, the court implied that the burden on plaintiffs in finding an expert within a small geographic area to testify against his peers was too great. Therefore, they overturned the locality rule and allowed plaintiffs to hire experts from other communities. This change in doctrine allowed plaintiffs to overcome a conspiracy of silence by physicians within one community. Again the courts created conditions that led to greater scrutiny of the care provided by physicians.31

Yet another blow was dealt to physicians' prerogatives when courts

began to apply the notion of corporate liability, for care rendered in hospitals. Corporate liability, was a relatively old notion in tort law that held that an entire enterprise could be held liable for the negligence of one employee if the corporation was responsible for supervising that employee. It had not been used widely in medical malpractice because most physicians were not employed by hospitals. In 1965, the Illinois Supreme Court ruled that the doctrine of corporate liability, could be applied to hospitals, holding them liable for physicians practicing there.32 This ruling and others like it created incentives for hospitals to supervise their physician staffs more closely, and to consider more thoroughgoing quality, assurance than simple reliance on physicians' ethical commitment to patients.33 While physicians continued to rely on informal review processes and an ethical code that "put the patient first" to regulate the quality of care, changes in doctrine wrought by courts in the two decades between 1944 and 1966 removed quality. assurance from the sole control of the physicians.

I believe this greater scrutiny, which can be interpreted as modernity dawning on the medical profession, is perfectly consistent with the theme of reintegration of medicine into the liberal state. As policymakers have become more concerned about injuries arising out of medical care, so has demand for greater accountability, increased. Changes in the res ipsa loquitur doctrine and the locality rule have made it far easier for plaintiffs to find experts to review the quality of care rendered and thus to win suits. Changes in corporate liability of hospitals have impelled hospital administrations to provide oversight. These changes in doctrine have stimulated a general social interest in malpractice litigation, creating greater incentives for better quality, assurance. In essence, the liberal state has asserted through the courts that physicians will be subject to the oversight of liberal tort law.

Changes in tort doctrine by courts have contributed to an overall increase in the amount and severity of malpractice litigation. This increase in litigation presumably increases the deterrent effect of the law and thus increases the quality of care. Whether or not that can be empirically demonstrated, it seems clear that the motives underlying changes in malpractice doctrine were quite similar to those underlying changes in the doctrine of informed consent.34 Courts had become convinced that physicians' power was too unrestricted, that the negative freedom of the patients needed better protection. Thus doctrinal changes reintegrate medicine with the public morality of our state.

It is interesting to note that other liberal states have seen the same

kind of trends in growth of malpractice claims. Both Britain and Canada have experienced similar increases.35 These increases have occurred without the same doctrinal reforms as in the United States, but likely, I would speculate, with the same judicial perception that physician prerogatives demand close judicial oversight.

The Profession's Reaction

There are a number of ways in which the medical profession could have reacted to this crisis of tort litigation. It might have engaged in productive self-examination and asked if the informal mechanisms of quality assurance on which it relied were really effective. Alternatively, it could have avoided the quality question and simply tried to decrease the amount of litigation through political means.36 Indeed, it is arguable that the profession selected the latter strategy.

There are several elements to this "frivolous litigation" strategy, which I will argue has been largely unsuccessful. First, it reflects implicit trust that traditional ethics will ensure that quality care will be rendered by physicians and that the current mechanisms for quality assurance are appropriate. As noted above, throughout the period of increasing litigation since the mid-1970s, there was little evidence that physicians became more interested in quality assurance. Little innovation had taken place in that direction and what was undertaken emanated more frequently from schools of public health than from medical schools.37 This absence of activity on the part of the medical profession suggests that it believed that quality was being well assured by the commitment of individual physicians to quality (the patient comes first) and by existing informal review mechanisms such as tissue committees and clinicopathological conferences.

Second, the professions' frivolous, litigation strategy implies that most suits are without merit, and that rather than being based on errors in medical management and poor quality care, they are prompted by misguided patients and influenced by greedy lawyers. These contentions have been cited time and again by representatives of the organized profession. For example, in an editorial in the Journal of the American Medical Association, the general counsel of the American

Medical Association noted that "many patients who file claims have suffered no injury as result of negligence. Even for those patients who do file claims as a result of negligence, there are wide, irrational variations in both findings of liability, and the amount of damages."38 In their report Professional Liability in the 1980's, the American Medical Association noted:

Some of the problems lie with the courts; changes in the laws relating to professional liability actions and in some hales of the court could be beneficial. A multitude of other factors also contribute to the professional liability problem—higher public expectations of what medicine can offer, a more litigious society, a greater number of lawyers willing to file suits that may not have merit, and the lure of possible huge awards or settlements.39

Nowhere is there consideration that quality of care may be the big problem.

This thesis on the causes of increased liability, appears to represent the consensus of the profession. Repeated polls of physicians by the American Medical Association and American College of Obstetrics and Gynecology reveal that physicians find the leading causes of liability to be the litigious society and greedy lawyers. In depth structured interviews recently conducted in the state of New York feature the same themes. Physicians feel victimized by a society, that allows so much misdirected litigation.40 These issues were also reflected in a bill sponsored by the AMA and introduced in the United States Senate in 1985, which included under the section entitled "Findings and Purpose," the following statement: "The number of professional liability claims against health care professionals and health care providers is increasing at disproportionate rates, beyond any relationship to the quality of care provided."41

Third, the physicians' frivolous litigation strategy, posits that the result of malpractice litigation is not better care, but simply more expensive care, as doctors perform useless tests to decrease the possibility of suits. Thus fear of liability burdens not only doctors, but also society, with the costs of "defensive medicine," which is defined as those medical practices that "are employed explicitly for the purposes either of averting a possible lass, suit or of providing appropriate documentation that a wide range of tests and treatments has been used in the patient's care."42 There have been various estimates of these costs, ranging from $3 to $6 billion in 1975 to $15 to $40 billion in 1983.43 Again, in structured interviews as well as in the official pronouncements of the profession, physicians reiterate that the major effect of litigation pressure is defensive medicine, not better quality care.44

Unfortunately, the rhetoric with regard to defensive medicine has not been matched by good empirical research on the subject. Indeed, there are no reliable studies of defensive medicine, and for good reason. The concept itself is very. difficult to understand. The provision of medical care lies on a spectrum from care that is reckless to care that is conservative and reduces uncertainty to an appropriate level. We tend to think that the best practitioners leave little to chance and pursue a conservative approach to medical problems. Thus conservative care is often good care, and institutions that provide conservative care are to be encouraged. Defensive medicine is care that is too conservative because it attempts to reduce uncertainty to an inappropriate level, and it is motivated only by fear of litigation. It can also involve not doing things, such as giving up obstetrics to avoid liability. Not all inappropriately conservative care is defensive medicine. Moreover, it is often difficult to define the difference, in particular episodes of care, between defensive medicine and careful medicine. This difficulty, is not unexpected; malpractice is defined as substandard practice, and the increase in conservative practices it provokes should, in a perfect system, only bring that practice up to the standard of care.

The existing literature on defensive medicine reflects these conceptual difficulties, and indeed questions the impact of litigation on clinical practice. Nathan Hershey's early interview study suggested that doctors were using some extra diagnostic tests as defensive medicine, but the total cost of these tests was quite small.45 Another study by the editors of the Duke Law Journal also questioned the widespread existence of defensive medicine.46 L. Tancredi and J. Barondess reached much the same conclusion in their influential review in 1978.47 Evidence suggests that doctors themselves do not think that defensive medicine plays a large role in their practices. S. Williams and colleagues studied physicians' perceptions about the reasons unnecessary diagnostic tests were ordered and found that less than three percent of practicing internists thought malpractice played a major role.48 Others have suggested that a major area of defensive medicine is in obstetrics, citing high C-section rates and use of electronic fetal monitoring (EFM). However, these kinds of claims underestimate the complexity. of the factors that go into the introduction of technology, like EFM and the increases in the use of procedures like C-sections.49 Indeed, most physicians would probably have a hard time coming up with a list of procedures that are highly influenced by defensive medicine, aside from some uses of radiology, especially in emergency rooms.

In the past few years, estimates of the impact of defensive medicine

have been developed through econometric models, most often based on survey information. In a recent study published in the Journal of the American Medical Association, R. Reynolds and colleagues used two methods to estimate costs of professional liability.50 Both methods relied on physicians' responses to mailed questionnaires. There was no way to correlate or evaluate the potential bias in physicians' responses. Moreover, the study defined increased record keeping, increased time spent with patients, and increased follow-up visits as examples of costs of professional liability. It is arguable that these are not defensive medicine costs, but rather examples of better care provoked by the deterrent function of tort law. Thus, as J. E. Harris stated in an editorial that accompanied this study, "We need to probe behind the survey responses analyzed by Reynolds and colleagues to uncover the microanatomic detail of defensive medicine."51 Nonetheless, defensive medicine remains a potent rallying cry for the profession even though it is very difficult to assess the size of the problem it represents.

Finally, the frivolous litigation strategy of the medical profession has focused on increased legislative involvement. To counteract the rise in suits, physicians feel compelled to seek relief from state legislatures, arguing that doctors are committed to high quality, care, that litigation is often unfounded, and that it only causes defensive medicine.52 In general, relief has been sought in one of three ways: (1) "raising the threshold" for getting into courts; (2) limiting the amount of damages one can gain; and (3) recasting liability rules.53

There are several ways for legislatures to make it more difficult to get into court. One is to modify, the existing American contingency fee system, which allows lawyers to accept as payment a portion of the compensation of the plaintiff. Indeed, some states have taken to regulating fees directly with percentage caps on the amount of the award that can be taken as payment.54 This lower payment creates incentives for plaintiffs' attorneys to shift from malpractice to other torts. Another way to keep plaintiffs out of courts is to decrease the statute of limitations, giving the injured party less time to discover and prove that the injury was a result of a tort. Yet another method is to force the plaintiff to be screened by a panel of experts before being able to go to trial. This delays the litigation process greatly.

Another effective method for discouraging litigation is to institute damage caps. Legislatures can put ceilings on both economic damages and compensation for pain and suffering, in the hopes that such caps will decrease the amount any one case is worth to a plaintiff's attor-

ney.55 Other simpler ways of reaching the same endpoint are changing the collateral source rule, which has traditionally provided that facts concerning the plaintiff's other sources of income are not to be put in as evidence, and instituting a periodic payment regime for damages.56 All of these changes would tend to decrease the initial lump sum payment owed the attorney, who has no doubt spent many hours bringing the case to trial.

Finally, and perhaps most directly, legislatures can change the liability rules that common law courts have modified. Some legislatures, for instance, have attempted to reinstate the locality rule.57 Others have tried to limit the use of res ipsa loquitur.58 All of these statutory modifications can potentially decrease the amount of malpractice litigation.

Attorneys and others argue that tort reform is nothing other than a strategy of powerful medical and insurance lobbies.59 Perhaps more important, many have challenged such statutory intervention on behalf of defendants as unconstitutional.60 While some patient advocates have tried to enlist the aid of federal courts on this question, most have been loath to find state statutory modifications unconstitutional.61 However, state courts have been willing to accept the due process "quid pro quo" argument that legislation is unconstitutional if it offers relief only to one side of an adversarial dispute.62 As Paul Weiler notes, "The courts, moreover, have perceived that powerful physician organizations have pushed for reforms, knowing that they would cut back on the rights of injured patients without offering them anything like the quid pro quo of limited but guaranteed no-fault benefits which had been a part of almost all prior tort reform legislation."63

In essence, the courts have again shown themselves willing to reiterate that changes in common law were meant to reinstate the public morality, of the liberal state, and that statutory modifications meant to block such reforms are an affront to the liberal constitution. Attorneys, challenging the constitutionality, of statutory reforms that make it more difficult for plaintiffs to sue, argue that such reforms decreased the compensation and deterrent functions of malpractice common law, harming the individual patient and favoring the medical profession. Of course, they reinforce this argument with rhetoric concerning the number of injuries in hospitals and the relatively small number of suits.

But there is reason to believe this contention is more than rhetoric. To address the need for more empirical information on these issues a team of researchers at Harvard University has undertaken a study of

medical injury, and litigation. One of our primal, goals was to develop up-to-date and reliable estimates of the incidence of adverse events and negligence among hospitalized patients. To accomplish this, we reviewed a random sample of 31,000 medical records from New York State using techniques we have previously described.64

Using weighted population figures, we estimated that of the 2,671,863 discharges in New York State in 1984, there were over 98,000 adverse events of which 27,177 were negligent. Using these figures, we estimated the statewide incidence rate of adverse events to be 3.7 percent and the negligent adverse rate to be 1.0 percent. Although most adverse events resulted in minimal (56.8 percent) or temporal-partial (13.6) disabilities, 2.6 percent caused permanent total disability, and 13.5 percent resulted in death. Moreover, the negligence rate was greatest among patients who suffered the more severe adverse events. We estimated that there were over 500 permanent total disabilities and almost 7,000 deaths due to negligent adverse events in New York State in 1984.

From our sample, we found that less than 5 percent of the cases we defined as negligent led to litigation. Thus it appears that only a small fraction of negligent cases were taken to courts. Of course, there were some suits arising out of hospitalizations in which we could find no evidence of negligence.

These results will likely fuel contentions that malpractice reform that provides only relief for physicians and hospitals from suits does not serve society's interest. In view of this and other data, state legislatures should begin to incorporate broader requirements concerning quality, of care into malpractice reform packages. The authors and advocates of such bills could argue that these requirements provide the quid pro quo missing in earlier tort reform packages: physicians get relief from suits but in return must allow more supervision of the work they do. Some of these reforms have already taken place at the federal level and provide relief from antitrust scrutiny for staff physician oversight of other physicians, a subject we will return to in chapter 10.65 Other such reforms are concerned with enhancing peer review by providing protection from defamation suits for physician reviewers in hospitals.66

By far the most important of the initiatives to improve the quality of care at the state level has been the effort to improve state disciplining mechanisms.67 Increasingly over the last five years, state attention to discipline of physicians has accompanied reforms in malpractice law.68 States have traditionally taken a leading role in licensing and

evaluating physicians, although much of their function has been formal and involved little real oversight. The state boards have been inadequately funded, are limited in their authority, and are often dominated by physician members.

The malpractice crisis has spurred changes. At the federal level, the Health Care Quality Improvement Act of 1986 has required creation of a national data bank for reporting disciplinary and malpractice actions against all physicians; this resource should tighten up accrediting and licensing efforts.69 States have gone several steps beyond this. In Massachusetts, for instance, the new Medical Malpractice Reform Act of 1986 requires physicians themselves, as well as hospitals, insurers, and medical societies, to report disciplinary actions and malpractice claims to the Board of Registration in Medicine.70 Each health care facility must develop patient care assessment plans and file these with the board. The board also received new subpoena power under the act and its staff grew from six to twenty-eight in one year. Other states, including Florida, California, New York, and Texas, have instituted similar reforms.71

The upshot of all this legislation is greater oversight of physicians. Chances are, given physicians' posture toward medical malpractice suits and quality, assurance, this supervision will continue to be necessary. A frivolous litigation strategy blinds physicians to any perception that society's concerns about malpractice may be concerns about quality of care. Physicians of course continue to resist the deterrent effect of malpractice litigation by attempting to minimize the amount of such litigation, but they demonstrate little evidence that they are committed to quality assurance programs that might make malpractice suits redundant. The result of the professions' strategy in the courts is that society will attempt to "police" physicians, an outcome that all must feel is unfortunate.

The Ethical Approach to Malpractice and Quality Assurance

As outlined in chapter 4, a medical ethics for the liberal state, just doctoring, must reject paternalism, further patient autonomy, and endorse the altruistic role of physicians in a relationship based on their concern and respect for their patients. It must also

include an honest appraisal by physicians of the social circumstances of the practice of medicine, and a renewed deference to the public morality of the liberal state. These factors come together in a health care sphere that respects liberalism, while evincing a spirit of cooperation and egalitarianism.

The medical profession's attitude toward malpractice and quality assurance has failed to meet this model. Although the profession is committed to providing the best care possible, it has guaranteed the quality of that care through informal, often hidden sanctioning. The profession has made no effort to demonstrate the efficacy of its review programs, as should be expected in the open liberal state. Instead, patients have had to be satisfied with paternalistic reassurances of the quality of the care they received. This approach does not demonstrate respect for patient autonomy.

The medical profession has resisted every judicial effort to empower patients through changes in malpractice doctrine. In their attempt to defend their prerogatives, physicians have again failed to demonstrate any respect for the values of liberalism. Their response has engendered state efforts to oversee the quality of medical practice. This has created an atmosphere more akin to police regulation than to cooperative health care, which I have argued should be the goal of medical ethics as just doctoring.

What, then, is the place of medical ethics with regard to malpractice and quality assurance? The answer seems fairly simple, at least at first cut: physicians must develop methods of quality assurance measures that are both effective and open to public review. If we are to demonstrate that the "patient comes first," we must be able to show that we are giving great thought to, and helping to bring about, better care. Indeed, I would argue that this should be a major priority of medical ethics, although it is rarely discussed in such terms.

Ethical quality assurance must be public, for secretive or hidden efforts do not respect the public morality of the liberal state or the negative freedom of the patient. Moreover, the results of quality, reviews must also be made public, just as they are in other industries. This allows the individual to make informed choices, another tenet of the liberal state. It also means that in addition to tissue committee and morbidity and mortality conferences, physicians must undertake other forms of quality assurance that will bring about better care.

There are many new quality assurance techniques taking shape which can be included in a just doctoring approach to the problem of

substandard care. In the evolving literature on this subject, measures of quality, are usually subdivided into process and outcome measures. Process measures evaluate care through reliance on criteria or elements of care that professionals agree are relevant and measurable.72 Outcome measures are perhaps more straightforward in that they focus on the short-term and long-term results of care. Outcome measures are, however, more difficult to develop. For instance, those outcomes now used as measurements include mortality and readmission rates. Both of these measures are heavily influenced by the mixture and the severity of cases that go into the class of outcomes being measured. For example, the sickest patients might refer themselves to the best hospital. That hospital may have a rather high mortality rate overall, but that high mortality would presumably be related to the higher incidence of very sick patients, not to the quality of the care delivered. Another way to assess outcomes is simply to have patients rate the care they have received.73 Again, however, there are problems with the validity of this kind of measurement.

Given these troubling aspects of outcome measures, there is a great deal of interest in process measures that can be shown to affect quality, of care. For example, anesthesiologists at Harvard University have recently adopted standards for anesthesia for common surgical procedures. The standards specify simple monitoring techniques, or practice guidelines, that are based on clinical experience as well as claims exPerience.74 The standards themselves can be specified on one-half of a page, and they appear to have significantly decreased the number of claims brought against Harvard anesthesiologists.75 However, anesthesia may prove to be an atypically easy area in which to develop such criteria; similar criteria for obstetrics require dozens of pages, for example, and thus may be much less easy to adopt. Further research into practice guidelines must be a priority for the medical profession, just as it has become for the federal government.76

In addition to exploring methods for quality assurance, physicians must also ensure that an appropriate institutional context exists for review initiatives. For example, fears of litigation can disrupt efforts to monitor the quality of care. Physicians involved in quality review can become embroiled in litigation in two ways.77 First, patient-plaintiffs seeking to establish that the care they received was substandard can petition the records of quality assurance offices concerning the care they were provided.78 Those physicians carrying out the review might be asked by the court to testify, about such cases. Second, quality

assurance can lead to litigation when embittered physicians who have been cited by quality, monitors sue the hospital and the physician reviewers for libel, defamation, or antitrust violations.79 Most states provide some protection for physicians against such suits, but there are loopholes in state laws, and tort lawyers have continually sought to expand the scope of litigation in both areas.80 Both types of suits make physicians uncomfortable about participating in review programs. If quality assurance is to become more rigorous, presumably physicians will have to be insulated to some degree from such suits.81

One form of quality, assurance obviates the need to focus on particular physicians or patients. Some health care policy analysts have begun to advocate managerial methods of quality assurance.82 Taking cues from modern management techniques that have been widely adopted in Japan, these analysts argue that physicians are only part of the whole process of medical care.83 Mistakes, oversights, or inefficiencies at any step can lead to bad outcomes. The analysts suggest that the entire process of medical care must be examined if we are to bring about better quality care. The attractiveness of this approach is that it moves the focus from the physician to the health care team, and it thus overcomes the perception of quality assurance as an incrimination of physicians.

In summary, the response of physicians, motivated by a sense of medical ethics as just doctoring, to the malpractice crisis should be a renewed commitment to better quality, care, as evinced by greater participation in quality assurance programs and scrupulously honest efforts to develop efficacious and open methods for improving medical care. Physicians must prove to society, that they can regulate themselves and that society will benefit from this regulation. The Liberal state will not at this point accept simple reassurances from physicians. We must prove that new approaches to quality assurance will work. If we do so, society may be willing to maintain protection for physicians who participate in such programs. In lieu of this demonstration by the profession of its commitment to quality, physicians will undoubtedly face stronger and stronger external regulation and oversight. This threat can only strengthen the appeal for ethical quality assurance efforts by all physicians.

The Ethics of Compensation for Iatrogenic Injury

I have argued that medical ethics must be cognizant of policy issues and physicians should openly address these issues. Some might argue that given my orientation, it seems peculiar to argue that physicians should concentrate on quality assurance alone. My critics might ask, should not patient compensation be a part of the profession's concerns about malpractice? If tort law does not provide appropriate and equitable compensation, should not physicians work for a better system?

My answer is a qualified yes. It is apparent from the empirical studies discussed above that the current system of compensation through tort litigation has serious drawbacks. As I have argued at length, one element of medical ethics, including medical ethics as just doctoring, is the physician's altruistic commitment to the patient. This commitment must not only include respect for the public morality of the liberal state but also dutiful action on behalf of patients—action not expected in other relationships in the liberal state. The key to medical ethics in the liberal state is that the physician retains a "patient comes first" attitude of other-regardness, while respecting fundamental rights.

Given this basis for medical ethics, the reliance on tort law as the sole source for compensation for an injured patient seems peculiar, and perhaps counterproductive. The tort law has several functions. Of these, we have discussed deterrence and compensation. Another function, which has been advocated by several legal scholars over the past decade, is the corrective justice aspect of tort law.84 The fact that tort law requires the injurer to pay the injured is not coincidence according to these scholars. Rather, they believe that tort law plays a special justice function in the liberal state, allowing the injured party to "even" himself with the injurer. This corrective justice is said to play an important symbolic role in the liberal state. It is not, however, a necessary part of modern liberalism as discussed in the first few chapters of this book.

Corrective justice has moderate appeal in some instances in the liberal state. Certainly the individualistic and egalitarian aspects of liber-

alism are reinforced when plaintiffs seek redress from impersonal corporations in product liability claims. Yet corrective justice makes little sense when considering a close relationship between doctor and patient that has now been complicated by medical injury. Thus medical ethics in the liberal state argues for alternatives.

Indeed, given the principles of just doctoring, administrative compensation for medical injury is preferable to tort law for several reasons.85 First of all, a no-fault compensation system means that the injured party only needs to show that she was injured during treatment to gain compensation. This means that both negligent and nonnegligent injury are compensated. Our research suggests that about one-quarter of all adverse events arc negligent. Since only negligently caused injuries arc compensated in a tort regime, this means that 75 percent of all injuries in our country go uncompensated. Thus a no-fault approach provides broader coverage for patients, a goal of our "patient comes first" ethic.

Second, the no-fault system docs away with the corrective justice aspects of the tort system. In Sweden and New Zealand, where no-fault compensation plans now exist, injuries are considered for compensation once they arc reported to the authorities. Increasingly, injuries are reported by the physicians themselves. The physicians, as part of their ethical commitment to the patient, try to help procure compensation for the patient.

Physicians arc so inclined for several reasons. Compensation is available for all injuries, not just those caused by negligence. Thus in helping the patient procure compensation, the physician is not admitting error. In addition, the physician in this system can accept that errors do occur. She knows, as do all of us, that mistakes are made by everyone in every line of work. She is not expected to be perfect. She knows that the authorities supervising physicians are keeping statistics on errors made by physicians, and she can thus be assured that an isolated mistake will not lead to discipline.

In the tort regime, the courts increasingly hold physicians liable for injuries suffered by patients so as to provide both broad compensation and deterrence signals for a profession that does not appear actively to police itself. This makes it well nigh impossible to admit error. The possibility of a physician's ethical commitment to patient compensation evaporates in the face of the corrective justice of the tort system. In the no-fault regimes of Sweden and New Zealand, the assumption is that accidents will occur, and both patient and physician have reason

to cooperate in the effort to gain compensation for the patient. The "patient comes first" ethic can flourish,

This is not to say that the deterrence function is lost. On the contrary, there is the possibility, of administrative systems for monitoring physician performance. Much of the information for these monitoring committees would come From episodes that lead to compensation. Physicians and other health care workers could participate in oversight committees and seek to uncover patterns of substandard care. If the patterns were the result of a single practitioner, steps could be taken to limit this person's practice. If the patterns were a result of a "systems" error, this information could be shared with all health authorities.

In this regime, physician altruism can flourish, and yet there is no infringement on the liberal rights of the patient. The physician can help the patient gain compensation for an injury. In addition, physicians can work selflessly to help improve medical care by pursuing quality assurance programs without fear of tort litigation.

Consider for example, the case I outlined at the beginning of this chapter. My patient died of salicylate poisoning. If he had a family or heirs, I and others associated with his care would have faced a series of very difficult questions regarding compensation for his family. While we did not believe we had provided substandard care, we all questioned whether we could have given better care. Yet we likely would not have recommended that the family sue us to gain compensation.

If we had been in a no-fault system, we could have recommended an application for compensation and forwarded the case to the appropriate authorities. They could have then considered whether compensation was due the patient's survivors. They could have also made a separate finding on the standard of care provided the patient. If it was substandard, we would have been on notice. If any one of us had continued to display questionable judgment, his practice would then be restricted. Perhaps new systems for alerting physicians to potential diagnoses could have come out of discussion engendered by the case. Thus the whole incident could have been bandied in an honest manner that allowed ethical behavior on the part of the physicians and reiterated the profession's commitment to high quality care.

Of course, the fear of those who oppose no-fault compensation is that the deterrence function of tort law will not be replaced by sincere efforts to bring about better quality care by physicians. They believe

that physicians will simply try to maintain authority over the practice of medicine and will not be amenable to disciplining themselves. They argue that the courts are necessary to interest the doctors in quality, assurance, and that tort litigation is the sine qua non of better quality care.

Those who advocate tort litigation to deal with medical error and patient injury have history on their side. Physicians committed to ethical care of the patient may agree with me that a no-fault approach has much promise for the reinforcement of ethical behavior by physicians. However, the only way they can convince the skeptics is by renewed efforts to develop quality assurance programs and to discourage substandard practices and police substandard providers. Quality assurance efforts are critical to convincing society that the ethical commitment to the patient can provide the kind of motivation that heretofore has only been provided by threat of court sanctions. Thus the answer to medical ethics as just doctoring, for the present and future malpractice crises, must be better quality oversight by the medical profession. Once this commitment is clear, physicians' efforts to develop no-fault approaches will be recognized as an ethical commitment to the patient, not a self-serving ploy.