The Decision-Analytic Approach to Medical Malpractice Law
Formal Proposals and Informal Syntheses

Harold Bursztajn, M.D.,
Robert M. Hamm, Ph.D.,
Thomas G. Gutheil, M.D.,
and Archie Brodsky, B.A.

From the Program in Psychiatry and the Law, Massachusetts Mental Health Center, Boston, Massachusetts, and the Center for Research on Judgment and Policy, University of Colorado, Boulder, Colorado. Please address requests for reprints to Dr. Bursztajn, Massachusetts Mental Health Center, 74 Fenwood Road, Boston, Massachusetts 02115, USA.

The close relationship between difficult medical decisions with potentially tragic outcomes and feared malpractice litigation has been an impetus to seeking analytic tools to aid decision makers. Although psychiatric decision making has been subject to well-publicized legal scrutiny, such scrutiny extends to all areas of medicine. During the past decade a decision-analytic reckoning of costs and benefits has been proposed as an improvement over traditional legal standards (particularly those of professional custom) for determining negligence in medical malpractice actions. For the most part, however, the courts have not adopted the mathematical approach to negligence determination used in Helling v. Carey in 1974. Nonetheless, the courts today do weigh costs and benefits, though in a less precisely numerical way, in deciding whether a physician has acted as a "reasonable and prudent" practitioner. Careful assessment of risks is specifically mandated by the informed-consent requirement, by which the physician is held responsible for disclosing the risks of alternative treatments to the patient. The synthesis of accepted professional standards and cost-benefit calculation in the "reasonable and prudent" criterion of negligence law corresponds, in cognitive terms, to the synthesis of the intuitive and analytic modes of thought in "quasi-rational" thinking. As such, it offers a useful model of decision making for physicians and patients. (Med Decis Making 4:401-414, 1984)

The use of decision analysis in medical decision making has been extensively explored and debated [1,2]. Simultaneously, the decision-analytic perspective has made its way into medical malpractice law, where it has been strenuously advocated if only tentatively applied [3-5]. The retrospective use of decision-analytic tools by the courts in determining malpractice liability is not only an area of great interest and importance in itself, but may also shed light on the potential application of these tools to before-the-fact decision making by physicians and patients.

Climo's1 report that even psychiatrists resort to decision rules motivated by the perceived pressure to practice legally defensible medicine is indicative of the impact that the fear of malpractice has had on medical decision making. Malpractice law operates from hindsight in that it seeks to compensate patients or their families for harm actually suffered. At the same time, it attempts to correct for hindsight by determining whether a physician was negligent on the basis of the information available at the time the physician's decisions were made. Decision analysis, introduced into legal reasoning as a refinement of traditional legal considerations, has had some success in protecting both plaintiffs and defendants from injustice due to hindsight. It has also come up against limitations that have tempered the courts' (if not decision theorists') enthusiasm for its application to malpractice law. A look at both the strengths and the limitations of decision analysis in curbing judicial hindsight may suggest guidelines for using it effectively to enhance medical foresight.

Antecedents of Decision Analysis in the Legal Tradition

The law has a long tradition of appealing to implicitly decision-analytic reasoning to decide cases of negligence. In 1871 a railroad was held liable in the death of a man who ran in front of an oncoming train in order to throw a small child to safety [6]. The verdict turned on the reasonableness of the man's decision (of necessity made in haste) to place himself in danger, which the railroad claimed constituted contributory negligence. A commentary by Terry in the Harvard Law Review in 1915 [7] listed five factors to be considered in assessing the reasonableness of such a risk:

  1. The magnitude of the risk. ...
  2. The value or importance of that which is exposed to the risk, which ... may be called the principal object. ...
  3. ... the value or importance of the collateral object [that which is to be attained by taking the risk]. ...
  4. The probability that the collateral object will be attained by the conduct which involves risk to the principal; the utility of the risk.
  5. The probability that the collateral object would not have been, attained without taking the risk; the necessity of the risk. ...

These elements were then listed for the case at hand:

  1. The magnitude of the risk was the probability that [the man] would be killed or hurt. That was very great.
  2. The principal object was his own life, which was very valuable.
  3. The collateral object was the child's life which was also very valuable.
  4. The utility of the risk was the probability that he could save the child. That must have been fairly great, since he in fact succeeded. ...
  5. The necessity of the risk was the probability that the child would not have saved himself by getting off the track in time.

The commentary did not resort to mathematical calculations, but did make informal estimates of probability ("very great," "fairly great") and value ("very valuable"). Its conclusion, although not stated in the language of a modern decision analyst, represented an intuitive attempt at decision analysis: "Here, although the magnitude of the risk was very great and the principal object very valuable, yet the value of the collateral object and the great utility and necessity of the risk counterbalanced those considerations, and made the risk reasonable" [7].

In cases of professional negligence, however, the courts for a long time saw little need for this common-sense version of decision analysis, since there was a ready-made "community standard" of due care. By this standard a physician was expected "to possess the degree of skill and learning which is possessed by the average member of the medical profession in good standing in the community in which he practices, and to apply that skill and learning, with ordinary and reasonable care" [8]. By this criterion of professional custom, a physician's acts were measured against what other physicians would have done in similar circumstances.

The first rock on which this criterion foundered was the inconsistency of local community standards in the face of advances in medical knowledge. Improved communications and a growing interdependence among physicians led to an expansion of the range of reference to "the same or similar communities" [9] and finally, in the case of nationally certified medical specialties, to a nationwide community of practitioners [10]. A more vexing issue has been the lack of a consensus among physicians on many clinical questions. The courts have been compelled to recognize that a physician may exercise due care through the use of methods approved by a "respectable minority" of the profession [11,12]. With all of these qualifications, the difficulty of identifying either a legitimate professional consensus or a respectable minority has led more than one judge to exclaim, "The test for malpractice is not to be determined by a plebiscite" [13,14].

Finally, what if the standard of the profession itself permits negligent practice? In the famous T. J. Hooper case of 1932, Judge Learned Hand held that "a whole calling [tug boat owners] may have unduly lagged in the adoption of new and available devices [radio receivers]. ...Courts must in the end say what is required; there are precautions so imperative that even their universal disregard will not excuse their omission" [15]. By this precedent the courts could overrule the customary practices of an entire profession on the basis of a higher standard of due care. Judge Hand enunciated such a standard in 1947, in language which for the first time brought mathematical calculation into negligence law:

Since there are occasions when every vessel will break from her moorings, and since, if she does, she becomes a menace to those about her, the owner's duty, as in other similar situations, to provide against resulting injuries is a function of three variables: (1) The probability that she will break away; (2) the gravity of the resulting injury, if she does; (3) the burden of adequate precautions. Possibly it serves to bring this notion into relief to state it in algebraic terms: if the probability be called P; the injury, L; and the burden, B; liability depends upon whether B is less than L multiplied by P: i.e., whether B [is less than] PL. [16]

With this decision, handed down just when decision analysis was under development, an "expected value" criterion for determining negligence (known as the Learned Hand rule) was established.

Decision Analysis in Medical Malpractice Law

There are numerous cases of medical malpractice on record in which the professional customs appealed to by the physician defendants have been held not to provide adequate safeguards for patients [17,18]. The rationale for these decisions was stated in 1972 in Canterbury v. Spence (a case better known for establishing the strongest legal precedent for the informed-consent requirement). Echoing Hooper, the court in Canterbury stated, "Prevailing medical practice...has evidentiary value in determinations as to what the specific criteria measuring challenged professional conduct are and whether they have been met, but does not itself define the standard" [19]. By analogy with the nonmedical negligence cases addressed by Judge Hand himself, an application of the Learned Hand rule to medical malpractice law appeared to be called for. In the same year Posner [20] advocated a thoroughgoing calculus of risks and benefits in which decision-analytic techniques would be used to allocate damages according to economic principles.

The stage was set for the controversial Helling v. Carey decision of 1974, in which an ophthalmologist was held negligent by the Supreme Court of Washington for failing (over a nine-year course of treatment) to administer routine tonometry screening to a young woman who then was found to have suffered permanent eye damage from glaucoma. The court acknowledged that it was the customary practice of ophthalmologists nationwide not to administer this test to patients under 40 because glaucoma occurs in only one in 25,000 patients in this age group. Nonetheless, the court ruled as follows:

The precaution of giving this test to detect the incidence of glaucoma to patients under 40 years of age is so imperative that irrespective of its disregard by the standards of the ophthalmology profession, it is the duty of the courts to say what is required to protect patients under 40 from the damaging results of glaucoma. [21]

It was the duty of the courts, then, to perform their own cost-benefit analyses rather than to rely on those implicitly embodied in the customs of the profession. The outlines of a simple decision analysis are contained in a concurring opinion in the case:

I believe a greater duty of care could be imposed on the defendants than was established by their profession. The duty could be imposed when a disease, such as glaucoma, can be detected by a simple, well-known harmless test whose results are definitive and the disease can be successfully arrested by early detection, but where the effects of the the disease are irreversible if undetected over a substantial period of time. [21]

In a later analysis Schwartz and Komesar [4] carried out the calculations to show that, as they structured the choices, the court's decision was consistent with the Learned Hand rule.

In the same year that Helling v. Carey was decided, Forst [3] independently proposed that the expected-value criterion of decision analysis be substituted for the standards of evaluation set by the legal tradition, which he characterized as vague, inconsistent, and not conducive to good medical care. By this argument the poor definition of the legal principles used to determine negligence had the effect of keeping physicians guessing. As a result, physicians unsure of their legal ground would tend to play it safe, i.e., practice defensive medicine. Forst's paper expressed a growing belief that some form of retrospective decision analysis would provide a better corrective to judicial hindsight than the traditional "community standards" criterion. Whereas the latter was biased toward the established, conservative practices of the profession (even if complacently negligent or overly attuned to the prevention of rare calamities), the mathematical approach to negligence appeared to have no bias at all [22]. Although the decision in Helling v. Carey favored greater use of technological safeguards, a cost-benefit analysis might just as well argue against the routine use of technology, as in the case of electronic fetal monitoring [23]. It is from this hopeful viewpoint that enthusiasm for the decision-analytic resolution of medical malpractice cases persists to the present day among some decision theorists [4,5].

The Reaction Against Decision Analysis

Helling v. Carey has generated considerable controversy within both the medical and legal professions. [24]. Physicians and legal experts have rallied to the defense of professional custom as a fairer and more reliable standard for determining negligence [25]. In response to such protests the state of Washington passed a statute specifying that medical malpractice cases be decided on the basis of the standards of the profession [26]. On the whole, the courts have not followed the Helling precedent in any sustained way [27].

Why the quantitative approach to negligence determination has not been compelling to judges and juries (let alone physicians) may be inferred from the language of Forst's proposal:

We then assert that, for any negligent treatment, the difference in expected disutility between the negligent treatment and the treatment demonstrated to be superior can be converted into a dollar amount such that the plaintiff would have been indifferent between having the inferior treatment along with this dollar compensation versus having the superior treatment with no compensation. [3]

The confidence of the author's tone and the good sense of this statement as an abstract proposition do nothing to eliminate either the technical difficulties of utility assessment [28] or the larger question of whether the various disutilities of negligent treatment can be made commensurable with one another or with dollars. The formulation quoted is, at any rate, rather distant from the complex and changing realities of pain, disability, and death.

Thus, the scepticism of judges and legislatures (as in the state of Washington) may reflect an intuitive appreciation of underlying methodological dilemmas. As a proponent of this application of decision analysis notes, "The exact forms of personal utility functions are not known, measures of health status and risk aversion are subject to error, and the phrasing of questions designed to clarify utility functions can have startling impact on the answers received" [5,29]. Learned Hand himself acknowledged the pitfalls of quantification:

The difficulties are in applying the rule,...they arise from the necessity of applying a quantitative test to an incommensurable subject-matter. ... It is indeed possible to state an equation for negligence in the form, C = P x D, in which the C is the care required to avoid risk, D, the possible injuries, and P, the probability that the injuries will occur, if the requisite care is not taken. But of these factors care is the only one ever susceptible of quantitative estimate, and often that is not. The injuries are always a variable within limits, which do not admit of even approximate ascertainment; and, although probability might theoretically be estimated, if any statistics were available, they never are. ...It follows that all such attempts are illusory, and if serviceable at all, are so only to center attention upon which one of the factors may be determinative in any given situation. [30]

The elements of the equation are not all equally difficult to quantify, however. Economic losses are more easily quantified than noneconomic ones; laboratory test results fit more readily into numerical calculations than subjectively experienced symptoms. The danger here is that a formulaic approach will bias the user toward overemphasizing "hard" information. In Tribe's words, "Even the most sophisticated user is subject to an overwhelming temptation to feed his pet the food it can most comfortably digest" [31]. With this bias in mind, some observers have turned Forst's argument on its head by predicting that a mathematical risk-benefit criterion for negligence will result in more, not less, defensive practice. It has been argued, for example, that psychiatrists who feel compelled to justify their actions in terms of risks and benefits in order to avoid malpractice liability (especially in sensitive areas such as the prediction of violence toward self or others) will be risk-averse even when the riskier choice has the higher expected value [32]. This may happen if risks and benefits are too narrowly defined, e.g., if the knowledge gained from attempting a treatment and observing its effects is not included as a benefit. It will happen, of course, only if the courts are unsophisticated in their use of the risk-benefit criterion or if practitioners base their actions on a fantasy of simplistic judicial retribution rather than on actual court decisions that do recognize the complexities of psychiatric treatment [33]. Nonetheless, the effects of such errors on legal judgments represent a legitimate concern in the evaluation of a method which must, after all, be carried out by fallible human beings. The principle established by Helling v. Carey is unassailable, but in practice it may still contribute to an indiscriminate use of diagnostic technology.

Decision analysis is employed (whether by physicians and patients or by attorneys, judges, and juries) in the hope of controlling the effects of various distorting influences on unaided human judgment. The latter include personality, emotional, cognitive, and philosophical factors. Each of these can, however, just as readily bias the three main judgmental tasks required in decision analysis: the structuring of the decision tree, the assessment of probabilities, and the assessment of values [22]. Whether or not formal decision-making methods are used, emotional regression during illness (intensifying the desire for certainty) and mistrust between patient and physician can have a devastating impact on decision making [2]. In a malpractice trial the jurors' empathic identification with the patient's suffering may precipitate similar reactions. The special danger of the formal decision-analytic approach is not that it creates these problems, but that it may lull decision makers (in either a medical or legal setting) into a state of lessened vigilance about them. In addition, misunderstanding of the implications of mathematical operations on the part of attorneys, judges, and juries may result in intentional or unintentional abuses. It was for these reasons that, even before Helling v. Carey, Tribe [31] expressed strong reservations about the consequences of introducing mathematical reasoning into legal proceedings.

Legal and Methodological Syntheses

The courts have to some extent resolved the opposition between the custom-based and cost-benefit criteria by failing back on another tradition of negligence law, that of appealing to the hypothetical conduct of a "reasonable and prudent practitioner." One recent court decision phrased this standard as follows: "A physician who undertakes a mode or form of treatment which a reasonable and prudent member of the medical profession would undertake under the same or similar circumstances shall not be subject to liability for harm caused thereby to the patient" [14]. This criterion, while anchored in the accustomed practices of the profession, also allows for the critical perspective added by the words "reasonable and prudent." To determine whether a physician's actions have been reasonable and prudent, an informal or intuitive analysis of costs and benefits is called for.

Legal precedents supporting both halves of the "reasonable and prudent" synthesis have accumulated in recent case law. The "community standard" lives on in the growing emphasis on the second opinion, which was given new legal significance by a recent decision holding an orthopedic surgeon liable for performing an operation against the advice of a consultant [34]. As a result of this decision, physicians contemplating major therapeutic interventions can be expected to seek supportive consultations more readily and to proceed more cautiously in their absence. The cost-benefit approach, on the other hand, is instantiated in an even more central legal doctrine, that of informed consent. The doctrine of informed consent has been defined as "the duty to warn a patient of hazards, possible complications, and expected and unexpected results of...treatment. ... As the probability or severity of risk to the patient increases, the duty to inform him of it increases" [27]. Thus, the physician must enter into a consideration of risks both in deciding how much the patient must be told and in actually presenting the proposed treatment for the patient's consent. Significantly, the dominant (though not unanimous) legal trend has been to uphold the patient's right to informed consent irrespective of professional custom [19,35,36]. Even the legislative reaction against the courts' imposition of a cost-benefit criterion for negligence has had the effect of legitimizing a framework for cost-benefit analysis. Thus the Washington state statute that reestablished the professional standard of due care in the wake of Helling v. Carey specifically exempted from its provisions those cases in which informed consent was at issue [26].

The disclosure of risks to the patient is part of a larger assessment of costs and benefits which the physician needs to perform and (by the informed-consent requirement) share with the patient, since the courts will attempt a similar analysis retrospectively if a bad outcome leads to a malpractice suit. The courts have been clear and consistent in ruling that a bad outcome in itself does not prove negligence [37,38]. Nor will a physician be held liable for a mere error of judgment [39]. A reasonable and prudent physician, however, is expected to undertake a course of action that is defensible (relative to alternate courses of action) in terms of expected costs and benefits.

Contrary to the fantasies of putative legal reasoning that motivate a particular specialty to practice defensive medicine, as described by Climo1 for psychiatry, examination of actual court decisions shows legal reasoning to be consistent with the decision-analytic normative standards for clinical reasoning [33,40].2 For example, a psychiatrist who is treating a suicidal patient is responsible for weighing

the possible benefits of close monitoring and increased protective control ...against the risks of regressive loss of ego-strength, stigma, reduced capacity to function autonomously, decreased liberty with compromise of some rights, and recurrent suicidality. The therapist who reduces restraints on a patient is taking a calculated risk of a suicide attempt, in return for therapeutic benefits that increase the likelihood of the patient's returning to normal life. [33]

Here, where the relevant factors are probably less amenable to quantification than in any other area of medicine, the courts have nonetheless been willing to scrutinize the costs and benefits in each case rather than simply hold psychiatrists liable for either all suicides by patients or none [41,42]. A good example of this informal cost-benefit reasoning (presented as an alternative to judicial hindsight) is found in an appeals court decision overturning a malpractice judgment against a psychiatrist:

[The patient's] reaction to constant surveillance, the possibility that his heart condition would be aggravated by continuing such surveillance, the gesture-like nature of his prior suicidal inclinations, the rehabilitative aspects of "open-ward" treatment and the enhanced possibility of obtaining [the patient's] consent to electroshock therapy in the more relaxed open-ward atmosphere were all factors which defendant doctor could properly consider in reaching the judgment whether, on balance, the prescribed program was worth the risk involved. [43]

Similar language is used in determining negligence in other contexts of medical decision making. A leading textbook of malpractice law gives the following summary of liability with respect to diagnostic testing:

...the more complicated and dangerous the test may be, the less likely it is that a court will find that a physician was negligent in failing to perform it. If, however, the patient's condition is quite serious, of course a physician would be much more likely to be held liable for failing to have ordered a fairly dangerous test than if the condition was no more serious, at worst, than the test itself. [27]

The same informal calculus is used in judicial review of the risks and benefits of treatment with drugs:

Of course, good medical practice frequently requires administration of a drug in the face of known side effects, and the matter usually is one of calculated risk. If the patient's condition warrants use of the drug and if no other drug is as effective, the physician is not negligent even if side effects do occur. If another drug would, however, be as effective and less potentially hazardous he is legally obliged to prescribe it. [27]

Such language takes us back full circle to Terry's 1915 law-review commentary on the 1871 railroad negligence case. It uses a decision-analytic framework, one that allows for, but does not insist upon (and in practice usually does not entail) mathematical calculation. Within this framework an action that has as its consequence risk of a bad outcome does not subject the physician to liability for malpractice. However, for the patient legally to assume the inevitable risks of treatment — whether in the case of x rays [44], elective surgery for diabetes [45], diagnostic testing such as angiograms or aortograms [46], or even release from the psychiatric hospital against medical advice—the patient must be fully informed of those risks. The informed-consent procedure, if properly carried out, thus provided both a legal defense for the physician and an opportunity to clarify costs and benefits both for the patient and for the physician [33,40,47,48]. In turn, decision analysis, employed with greater or lesser formality at the physician's discretion, may be as much a useful tool in structuring the choices presented for the patient's consent as it is in the physician's ongoing review (alone or with colleagues) of the reasonableness and prudence of the clinical decisions made.

The "reasonable and prudent" standard of due care, interpreted in this way, does justice to the uncertainties of medical decision making as neither of the other two standards does alone. It is more flexible and context-sensitive than either [22]. Instead of subjecting the individual case to the dictates of either custom or mathematics, it respects the unique conditions surrounding a given case and allows the physician to justify his or her actions in the light of those conditions. Moreover, it more readily accommodates consideration of difficult-to-quantify subjective factors than either of the other two standards [33]. Finally, it is more consistent with twentieth-century, probabilistic science, which (as distinct from the earlier, mechanistic paradigm) provides principles and decision rules (maxims or rules of thumb) suitable for coping with uncertainty, making critical use of diagnostic and therapeutic technology, and replacing hindsight with realism about the limits of prediction [2].

Counterparts of the legal reasoning described here (in terms of a synthesis of formal and informal methods) range from a decision strategy outlined by Benjamin Franklin in 1772 to a theory of cognition proposed two hundred years later by psychologist Kenneth Hammond. Franklin wrote to the British scientist Joseph Priestley:

When these difficult cases occur, they are difficult, chiefly because while we have them under consideration, all the reasons pro and con are not present to the mind at the same time; but sometimes one set present themselves, and at other times another, the first being out of sight. Hence the various purposes or inclinations that alternatively prevail, and the uncertainty that perplexes us. To get over this, my way is to divide half a sheet of paper by a line into two columns; writing over the one Pro, and over the other Con. Then, during three or four days consideration, I put down under the different heads short hints of the different motives, that at different times occur to me, for or against the measure. When I have thus got them all together in one view, I endeavor to estimate their respective weights; and where I find two, one on each side, that seem equal, I strike them both out. If I find a reason pro equal to some two reasons con, I strike out the three...and thus proceeding I find at length where the balance lies; and if, after a day or two of further consideration, nothing new that is of importance occurs on either side, I come to a determination accordingly. And, though the weight of reasons cannot be taken with the precision of algebraic qualities, yet when each is thus considered, separately and comparatively, and the whole lies before me, I think I can judge better, and am less liable to make a rash step, and in fact I have found great advantage from this kind of equation, in what may be called moral or prudential algebra. [49]

Hammond [50] believes that the intuitive and analytic modes of thought are the two extremes of a continuum, with the ideal synthesis (for many purposes) represented by what he calls "quasi-rational" thinking. Among the three traditions of negligence law, the perception of what others in the professional community would do is largely an intuitive process, while the calculation of expected values to satisfy the Learned Hand rule is an analytic one. The "reasonable and prudent" standard, which combines the two modes of thought, might be said to have a "quasi-rational" character. It has been argued, for example, that the acknowledged difficulty of predicting suicidal or violent behavior makes it both unfair and futile to hold any psychiatrist liable for failure to prevent such behavior [32]. This "defendant's brief" assumes the psychiatrist's task is so uncertain that one can only proceed intuitively. At the other extreme, plaintiffs who claim that the mere fact of a tragic outcome (e.g., a patient's committing suicide while under a psychiatrist's care) is, ipso facto, proof of negligence are assuming that the psychiatrist's environment would allow total certainty if only one would practice analytically. The courts have been more sophisticated than either side in recognizing that prediction is a matter neither of total certainty nor of total uncertainty, but rather of informal judgments of probability and value. The courts (especially at the appellate level), by applying the "reasonable and prudent" standard with the help of the language of risks and benefits, have, in effect, been engaging in as well as endorsing quasi-rational thinking [33].

In contrast to Forst [3], who claims that the principles of the law are internally contradictory, Tribe [31] maintains that the rich traditions of the law, with their intricate rituals and elaborate symbolism, enable the law to tolerate and work with uncertainty. Elsewhere, Tribe has rejected the "elusive ideal of wholly objective, impersonal, and detached instrumental analysis" as "not only unattainable but destructive" and called for "a subtler ...and more complex style of problem solving" [51]. Our review of recent case law and Hammond's description of the modes of human cognition converge to show how people meet this need.


Even articles advocating the wider use of decision-analytic thinking in malpractice law contain caveats such as: "Decision analysis has not been developed to the point of enabling many-decimal-place precision of optimal regulations. It does, however, provide general insights and policy guidance" [5]. This statement, as applied to malpractice law, echoes a legal commentary published in 1927, which states, "We must not assume that we can rely upon any formula in regard to 'balancing interests' to solve negligence cases. The phrase is only a convenient one to indicate factors which may be considered and should not connote any mathematical correspondence" [52], After a brief flirtation with a more exact calculation of costs and benefits, the courts have, in the main, returned to this traditional viewpoint.

The law's "quasi-rational" style of assessing medical decisions after the fact may be a good model for making such decisions before the fact, in that it recognizes the limitations of more formal methods besides those of human beings. Although no method can completely avoid the distortions and biases of decision making, judicious use of formal methods may at least inhibit self-deception on this score. Decision analysis has been compared with the "training wheels" of a bicycle as a provisional substitute for internal balance gained from experience, and with the psychoanalyst's couch as a prop for learning self-analysis [22]. Such aids are best used in a time-limited or selective, rather than rigid and routinized, way. For example, a physician might draw decision trees to document difficult decisions, but without necessarily filling in the numbers. Overall, physicians and patients would do well to emulate the courts in mastering informal uses of the decision-analytic method.


  1. Climo LH: Some thorny medical judgments and their outcomes. The view from the public mental hospital. Med Decis Making 4:415-424, 1984
  2. Hamm RM, Clark JA, Bursztajn H: Psychiatrists' thorny judgments. Describing and improving decision-making processes. Med Decis Making 4:425-447, 1984


Thanks to Ms. Audrey Bleakely for her help in the preparation of this manuscript.


  1. Weinstein MC, Fineberg HV: Clinical Decision Analysis. Philadelphia: Saunders, 1980
  2. Bursztajn H, Feinbloom RI, Hamm RM, Brodsky A: Medical Choices, Medical Chances. How Patients, Families, and Physicians Can Cope with Uncertainty. New York: Delacorte, 1981
  3. Forst BE: Decision analysis and medical malpractice. Oper Res Jan-Feb: 1974, 1-12
  4. Schwartz WB, Komesar NK: Doctors, damages and deterrence. An economic view of medical malpractice. N Engl J Med 298:1282-1289, 1978
  5. Thompson MS: Health versus money. Value judgments in the perspective of decision analysis. Med Decis Making 3:285-297, 1983
  6. Eckert v. Long Island R.R., 43 N.Y. 502 (1871)
  7. Terry HT: Negligence. Harvard Law Rev 29:40-54, 1915
  8. Boyce v. Brown, 51 Ariz. 416, 77 P. 2d 455 (1938)
  9. Tallbull v. Whitney, 564 P. 2d 162 (Mont., 1977)
  10. Robbins v Footer, F. 2d 123 (D.C. Cir., 1977)
  11. Bruce v. United States, 167 F. Supp. 579, D.C. Cal. 1958
  12. Kortus v. Jensen, 237 NW 2d 845, Neb. 1976
  13. Chamler v. McClure, 505 F. 2d 489, CCA 6, 1974
  14. Hood v. Phillips, 554 S.W. 2d 160, S. Ct. Tex 1977
  15. The T.J. Hooper, 60 F. 2d 737 (2d Cir. 1932)
  16. U.S. v. Carroll Towing Co., 159 F. 2d 169 (2d Cir. 1947)
  17. Incollingo v. Ewing, 282 A 2d 206, Pa. 1971
  18. Guilbeau v. St. Paul Fire and Marine Insurance Co., 325 So. 2d 395, La. 1975
  19. Canterbury v. Spence, 464 F. 2d 772 (D.C. Cir. 1972)
  20. Posner RA: A theory of negligence. J Legal Stud 1:29-96, 1972
  21. Helling v. Carey, 83 Wash. 2d 514, 519 P. 2d 981 (1974)
  22. Bursztajn H, Hamm RM, Gutheil TG: The technological target. Involving the patient in clinical choices. In, Reiser SJ, Anbar M, eds: The Machine at the Bedside. Strategies for Using Technology in Patient Care. Cambridge, England: Cambridge Univ Press, 1984, pp 177-191
  23. Thompson MS, Cohen AB, Palmer RH: Decision making on the clinical use of electronic fetal monitors. Semin Fam Med 3:89-94, 1982
  24. King JH: In search of a standard of care for the medical profession. The "accepted practice" formula. Vanderbilt Law Rev 28:1213-1276, 1975
  25. Epstein RA: The limits of medical malpractice. N Engl J Med 298:1311-1312, 1978
  26. Washington Revised Code (Supp. 1975), Section 4.24.290
  27. Holder AR: Medical Malpractice Law (2nd ed.). New York: Wiley, 1978
  28. Bursztajn H, Hamm RM: The clinical utility of utility assessment. Med Decis Making 2:161-165, 1982
  29. Bursztajn H, Chanowitz B, Gutheil TG, Hamm RM: Context specific language effects in the decision to prescribe neuroleptics. Presented at the Fifth Annual Meeting of the Society for Medical Decision Making, Toronto, Canada, October 2-5, 1983
  30. Moisan v. Loftus, 178 F. 2d 143 (2d Cir. 1949)
  31. Tribe LH: Trial by mathematics. Precision and ritual in the legal process. Harvard Law Rev 84:1329-1393, 1971
  32. Klein JI, Glover SI: Psychiatric malpractice. Int J Law Psychiatry 6:131-157, 1983
  33. Gutheil TG, Bursztajn H, Hamm RM, Brodsky A: Subjective data and suicide assessment in the light of recent legal developments. I. Malpractice prevention and the use of subjective data. Int J Law Psychiatry, in press
  34. Fritter v. Cassidy, NWC 60672, C232092 (Cal., 1983)
  35. Scaria v. St. Paul Fire and Marine Insurance Co., 227 NW 2d 647 (Wisc, 1975)
  36. Zeleznik v. Jewish Chronic Disease Hospital, 366 NYS 2d 163 (1975)
  37. Staloch v. Holm, 111 NW 264 (Minn., 1907)
  38. Teig v. St. John's Hospital, 387 P. 2d 527 (Wash., 1963)
  39. Eckleberry v. Kaiser Foundation Northern Hospital, 359 P. 2d 1090 (Ore., 1961)
  40. Bursztajn H, Gutheil TG, Hamm RM, Brodsky A: Subjective data and suicide assessment in the light of recent legal developments. II.Clinical uses of legal standards in the interpretation of subjective data. Int J Law Psychiatry, in press
  41. Fiederlein v. City of New York Health and Hospitals Corp., 80 App. Div. 2d 821, 437 N.Y. 2d 321 (1981)
  42. Johnson v. United States, 409 F. Sup. 1283 (D. Fla. 1981)
  43. Topel v. Long Island Jewish Medical Center, 431 N.E. 2d 293, 55 N.Y. 2d 682 (1981)
  44. Natanson v. Kline, 350 P. 2d 1093, 354 P. 2d 760 (Kans., 1960)
  45. Mainfort v. Giannestras, 111 NE 2d 692 (Ohio, 1951)
  46. Bowers v. Talmage, 159 So. 2d 888 (Fla. 1963)
  47. Mazur DJ: Translating court decisions into ethical-legal guidelines for patient-physician communication about medical risk. Presented at the Fifth Annual Meeting of the Society for Medical Decision Making, Toronto, Canada, October 2-5, 1983
  48. Gutheil TG, Bursztajn H, Brodsky A: Malpractice prevention through the sharing of uncertainty. Informed consent and the therapeutic alliance. N Engl J Med 311:49-51, 1984
  49. Bigelow J, ed.: The Complete Works of Benjamin Franklin (Vol. 4). New York: Putnam, 1887, p 522. Cited in Hogarth RM: Judgment and Choice. The Psychology of Decision. Chichester, England: Wiley, 1980, p 142
  50. Hammond KR: Principles of organization in intuitive and analytical cognition. University of Colorado, Institute of Behavioral Science, Center for Research on Judgment and Policy, Report No. 231, 1981
  51. Tribe LH: Policy science. Analysis or ideology. Phil Pub Affairs 2:66-110,1972
  52. Seavey WA: Negligence-subjective or objective? Harvard Law Rev 41:1-28, 1927