EDITORIALS
Comparing Cost/Benefit Ratios for AMI Care In this issue of Annals, Wears and colleagues report on a probabilistic model to compare the relative cost/benefit ratios of different strategies for managing patients with the potential for acute myocardial infarction (AMI) as a function of the physician's probability estimate of AMI. The authors use the economic concept of marginal cost/ life saved to provide a concrete interpretation of their results and a means of comparing m a n a g e m e n t strategies. Their article, in addition to being timely and thought provoking, introduces emergency physicians to a sophisticated m e t h o d of decision analysis. While the tools are powerful, the applicability of the theoretical results depends on the underlying assumptions. Some of these assumptions need additional clarification and discussion. See related article, p 953 A m o n g the factors the authors associate with the cost of a missed AMI is the cost of litigation. While this is commendable from the standpoint of thoroughness, it is unclear if litigation cost is the basis on which we should decide medical careJ Litigation costs seen by the health care industry have been modeled by the authors as a function of the clinical presentation (probability estimate of AMI), action taken by the physician (largely the disposition), and the outcome from that action. Other factors, including the physician's d o c u m e n t a t i o n of low risk for AMI and the patient's potential lost earnings, are excluded from this model. If the authors' model is designed to weigh the cost/ benefit ratio of AMI care for society, then their model is too simplistic. Legal expenses for the health care industry represent income for the legal industry and do not directly equate with an expense to society as a whole. While m a n y would argue that these services provide no true benefit to society, such services e c o n o m i c a l l y c o n t r i b u t e to the gross national product. Hence, from an economic position, society m a y view the cost of litigation as s i m p l y the m e a n s by w h i c h a p o r t i o n of the legal profession is funded. Assuming society desires to fund the legal industry, the authors' model greatly overestimates the "cost" of litigation to society. If the authors' model is designed to weigh the cost/ benefit ratio of AMI care for the health care industry in isolation, then a different measure of benefit should be used. Contrary to the individual clinician's desire, the health care industry e c o n o m i c a l l y does not specifically care if the patient's life is saved. The health care industry merely wishes to provide a given service at the lowest cost (including litigation cost) and obtain the m a x i m u m reimbursement. The cost/patient treated m a y be a more relevant economic outcome m e a s u r e for the health care industry. Obviously the health care i n d u s t r y does care about the patient as a consumer, but not on the personal level practiced by the clinician. 172/1014
The authors also assume that admission alone will not increase mortality (or significantly reduce the n u m b e r of productive days of the patient's life). This assumption requires supportive evidence. One can certainly anticipate increased m o r b i d i t y w h e n patients are admitted needlessly. This m a y occur when the patient receives diagnostic cardiac catheterization or similar invasive procedures after being ruled out. Furthermore, the enforced bed rest associated with admission m a y result in increased thrombophlebitis and other morbidity. With a more lax admission policy, some patients m a y be observed solely in the emergency department due to a shortage of monitored beds. This m a y adversely affect quality of care and litigation costs in other patient problem areas. There m a y also be h i d d e n p s y c h o l o g i c a l costs to individuals a d m i t t e d falsely for rule-outs; they m a y not return to work immediately after discharge. We can only speculate on the true cost of false rule-outs that should be added to the authors' model. Additionally, to make the authors' model workable, the physician m u s t be able to make a reasonable estimate of the probability of AMI. This m a y either be done by the physician's gestalt estimateZ, 3 or by using a predictive tool.3, 4 Unfortunately, we have little information regarding h o w well any of these estimates function in the low range of probability for AMI where the authors' model suggests we should operate. Finally, t h e r e are a d d i t i o n a l difficulties w h e n one wishes to change focus from a population model (as outlined by the authors) to the patient before us. As clinicians we m a y be swayed in our decision making by specific features that are more certain for the patient at hand. Despite these limitations, the authors' model is of value for focusing our attention on the issues that affect chest pain patient admission practices. While we should continue to practice with a high index of suspicion and be cognizant of the authors' model, we should not accept the authors' thresholds as gospel. Jerris R Hedges, MD, FACEP Division of Emergency Medicine The Oregon Health Sciences University Portland
1. Rouan GW, Hedges JR, Toltzis R, et al: Chest pain clinic to improve follow-up? {letter]. Ann Ernerg Med 1988;17:868-869. 2. Tierney WM, Fitzgerald ], McHenryR, et al: Physicians estimates of the probability of myocardial infarction in emergency room patients with chest pain. Med Decis Making 1986;6:12-17. 3. Goldman L, Cook EF, Brand DA, et al: A computer protocol to predict myocardial infarction in emergencydepartment patients with chest pain. N Engl J Med 1988;318:797-803. 4. Pozen MW, D'AgostinoRB, Selker HP, et al: A predictive instrument to improve coronary care unit admission practices in acute ischemic heart disease. N Engl J Med 1984;310:1273-1278.
Annals of Emergency Medicine
18:9 September 1989