Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
BRIEF COMMUNICATION Postoperative Death and Malpractice Suits: Is Autopsy Useful? Philippe Juvin, MD*, Frédéric Teissière, and Michel Durigon, MD† MD†, Fabrice Brion, MD†, Jean-Marie Desmonts, MD*, *Service d’Anesthésie et de Réanimation, Hôpital Bichat, Paris; and †Service de Médecine Légale, Hôpital Raymond Poincaré, Garches, France T he number of autopsies has declined in many hospitals in the last few decades (1). One factor in this trend is probably the greater confidence with which physicians can establish a premortem diagnosis. However, when a malpractice suit is filed after a postoperative death, the plaintiff often considers the medical chart data inadequate or untrustworthy. In this situation, an autopsy could have confirmed the cause of death and eventually determine whether a fault occurred. The medical value of such autopsies has not been assessed. The present study was designed to determine whether autopsies performed in the case of postoperative death with suspicion of malpractice help identify the cause of death. Methods This retrospective medical chart review study included all adult (⬎18 yr old) patients who died after a surgical procedure between 1993 and 1998 in three districts of France (Yvelines, Eure et Loir, and Val d’Oise, i.e., approximately 2.8 million inhabitants) and those who had an autopsy performed after the family filed a malpractice suit. All autopsies were performed within 3 days after the death, according to the same procedure, in a single forensic institute. The organs were examined in situ, then removed, examined and dissected ex situ, and subjected to a histological examination. The circumstances and the cause of death indicated by the physician in charge of the patient (primary diagnosis) were abstracted from the patients’ medical charts, as were the results of the autopsy (autopsy diagnosis). The primary and the autopsy diagnoses were compared by a panel of three certified physicians Accepted for publication April 26, 2000. Address correspondence and reprint requests to Philippe Juvin, Service d’Anesthésie et de Réanimation, Hôpital Bichat, 46 Rue Henri Huchard, 75018 Paris, France. Address e-mail to pjuvin @free.fr. 344 Anesth Analg 2000;91:344–6 (one anesthesiologist, one forensic physician, and one surgical intensivist). When the diagnoses were different, the three physicians reviewed the data to confirm or refute the autopsy diagnosis. When the autopsy failed to identify the cause of death, it was determined whether a different autopsy technique would have been more informative. Results Fifteen autopsies were performed during the study. Mean time (range) from surgery to death was 4 days (0 –10 days) (Day 0 ⫽ day of surgery). The clinical history and primary and autopsy diagnoses of the patients are reported in Table 1. A primary diagnosis was recorded in 11 cases (Cases 1 to 11). Among these primary diagnoses, 3 were confirmed (Cases 1–3) and 7 (Cases 4 –10) were invalidated by the autopsy. Among these 7 cases, the autopsy corrected the primary diagnosis in 5 cases (Cases 4 to 8) and disproved the primary diagnosis without finding the cause of death in Cases 9 and 10. An autopsy diagnosis was obtained in 11 cases (all the cases except Cases 9, 10, 11, and 15). Among the 4 cases with no primary diagnosis, the autopsy identified the cause of death in 3 cases (Cases 12–14). The technique used during the autopsy was not appropriate to perform a diagnosis in 2 cases (Cases 11 and 15). Discussion This report demonstrates that most early postmortem autopsies performed on patients who died postoperatively provide new and often unexpected information of great assistance in identifying the cause of death. Our first result was that most of the diagnoses performed by the physician in charge of the patient at the time of death were wrong. Our second result was that the autopsy yielded new information (providing or ©2000 by the International Anesthesia Research Society 0003-2999/00 ANESTH ANALG 2000;91:344 –6 BRIEF COMMUNICATION JUVIN ET AL. POSTOPERATIVE DEATH AND AUTOPSY 345 Table 1. Patient Characteristics with Primary and Autopsy Diagnoses Case 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Clinical history Medical chart with a primary diagnosis of the death Cause of death confirmed by the autopsy Male, 70 yr old. Printzmetal angina. Two-vessel coronary bypass procedure. Died on D1. Male, 57 yr old. Uninvestigated chest pain 3 wk earlier. Surgery for fracture of the humerus. Arterial hypotension intraoperatively and acute heart failure postoperatively. Died on the same day with shock Female, 35 yr old. Cardiorespiratory arrest during labor. Cause of death invalidated by the autopsy Cause of death identified by autopsy Male, 77 yr old. Coronary heart disease and left ventricular failure. Transurethral resection of the prostate. Seemed to have shown evidence of chest pain intraoperatively. Was not investigated postoperatively. Found dead in his bed on D1. Female, 92 yr old. Patient on aspirin therapy. Uninvestigated fall 2 mo before surgery for an acetabular fracture. Found dead on D2. Female, 30 yr old. Laparoscopic surgery for uterine myoma. Circulatory collapse followed by cardiorespiratory arrest during the procedure. Died a few hours later. Female, 20 yr old. Substance abuse. Appendectomy. Found dead on D10. Male, 73 yr old. Surgery for an acetabular fracture. Inadequate perioperative anticoagulant therapy. Died suddenly on D1. Cause of death not identified by autopsy Female, 59 yr old. Surgery for L4-5 disk herniation. Circulatory collapse followed by circulatory arrest when turned to the prone position. Glasgow 3 coma after resuscitation and until death on D15. Male, 84 yr old. Colectomy. Fell from his bed on D7. Found dead on D8. Female, 47 yr old. Liposuction. Was given 2 g of amoxicillin intraoperatively. Left the operating room at 5 pm. At 10 pm, loss of consciousness followed by circulatory arrest during an amoxicillin injection. Died on D1. Medical chart with no primary diagnosis of the death Cause of death identified by autopsy Male, 78 yr old. Total gastrectomy for cancer. Fever on D4. Negative laparotomy on D5. Died from septic shock on D10. Male, 87 yr old. Continuity reconstitution after removal of the transverse colon. Shock at the end of the procedure. Death on D3 with MOF. Male, 35 yr old. Surgery for diastasis of the rectus abdominis muscles. Intestinal obstruction requiring resection of necrotized bowel loops on D8. Died on D11 with MOF. No cause of death identified by autopsy Male, 68 yr old. History of diabetes, asthma, and coronary heart disease. Surgery for a fracture of a toe. Unexplained agitation the next day requiring injection of droperidol. Found dead 15 min after the injection. D ⫽ postoperative day, MOF ⫽ multiple organ failure. Primary diagnosis Autopsy diagnosis Myocardial infarction Myocardial infarction Myocardial infarction Myocardial infarction Amniotic fluid embolism Amniotic fluid embolism Myocardial infarction Old myocardial infarction, with no evidence of recent cardiac ischemia. Pleural effusion (1.3 L) caused by decompensated heart failure and responsible for the death Intracerebral bleeding (right fronto-parietal area and brainstem) Congenital abnormality of the coronary arteries explaining the recent ischemic lesions responsible for the death Massive pulmonary embolism Pulmonary embolism Pulmonary embolism Overdose Pulmonary embolism Cancer of the esophagus. Massive bleeding caused by fistulization of the esophagus into the aorta Myocardial infarction No myocardial infarction. No findings explaining the death Cerebral hematoma Anaphylactic shock No cerebral hematoma. No findings explaining the death No findings explaining the death None (septic shock of unknown cause) Intraperitoneal abscess opposite the esoduodenal suture None (MOF of unknown cause) Large intraperitoneal hematoma (⬎5 L) None (septic shock of unknown cause) Necrosis of the entire small bowel. Massive bleeding from a gastric ulcer None No findings explaining the death 346 BRIEF COMMUNICATION JUVIN ET AL. POSTOPERATIVE DEATH AND AUTOPSY correcting the diagnosis) in two of three of the cases. These proportions are at variance with those of earlier studies performed in intensive care units (2,3), in which most primary diagnoses were confirmed by the autopsy. However, there was no suspicion of malpractice in these previous studies. These discrepancies may be explained by the fact that autopsies performed in case of suspicion of malpractice may be more likely to disprove the primary diagnosis than autopsies overall. It is indeed reasonable to assume that family members who did file charges may have done so because they noticed something that raised their suspicions. It follows that autopsies may have been performed more often in cases characterized by egregious mistakes; in this situation, the yield of the autopsies would be high, as was the case in our study. This recruitment bias in our study is similar to that in the American Society of Anesthesiologists Closed Claim Project. In agreement with previous studies (3,4), the new information provided by the autopsy could have influenced treatment decisions in some cases. In Cases 12 (unrecognized intraperitoneal abscess), 13 (unrecognized intraperitoneal hematoma), and 14 (unrecognized necrosis of the small bowel), early surgery could have modified the prognosis. In Case 4, early recognition of decompensated heart failure could have allowed an appropriate treatment. In Case 5, the discontinuation of aspirin therapy after a fall could have reduced the intracerebral bleeding. In Case 6, an intensive perioperative management of the coronary disease could have prevented the myocardial infarction. In Case 7, an anticoagulant therapy could have changed the prognosis. However, Patients 6 and 7 appear to have suffered from rare events, and the desirable changes in management are not supported by data on cost-effectiveness. In Case 8, outcome would not have been different even if the fistula had been identified. Thus, among the eight cases (Cases 4 – 8 and 12–14) in which the autopsy provided a diagnosis which had not been suspected premortem, changes in perioperative management would have modified the prognosis in five cases (Cases 4, 5, and 12–14), suggesting that faulty or negligent practice may have occurred. More generally, one cannot be certain that that knowing the true diagnosis would have prevented the death. Nevertheless, it is clear that failure to make the diagnosis has reduced the likelihood of patient survival. In two cases, the autopsy technique was not appropriate for performing a diagnosis. In Case 11, the primary diagnosis was anaphylactic shock, and the postmortem specimens for anaphylaxis mediator testing were not obtained. In Case 15, death occurred after ANESTH ANALG 2000;91:344 –6 injection of a neuroleptic agent, but specimens were not obtained to look for a drug overdosage or an anaphylactic reaction. These cases suggest that good practice recommendations should be developed to determine the specific characteristics of the autopsy procedure to be used in the event of a perioperative death. Limitations of this report include the small sample size and the fact that autopsies were performed a few days after the death. In the hypothesis of a late claim, months or years after death, the yield of the autopsies should be reevaluated because of the putrefaction phenomena. However, the conclusions concerning the weak value of the primary diagnosis remain valid. In any case, our report was representative of present and actual early postmortem claims after postoperative death in a population of 2.8 million inhabitants. This report demonstrates the extremely high yield of early postmortem autopsies performed in the case of postoperative death with suspicion of malpractice. They frequently identified undetected complications, including surgical complications and disease processes. They could also suggest faulty or negligent practice that would otherwise go unrecognized. As reported in the present study, the autopsy can help the defense of the patient’s family when a complication, particularly a surgical one, was missed (5), or the defense of unjustly accused physicians when the autopsy findings refute doubts about the appropriateness of their practice. Autopsy can also help conflict solving: when the family suspects that malpractice has occurred, they are unlikely to believe the explanations provided by the physician to justify his management of the case, and in this situation, an autopsy can restore trust and resolve conflicts by providing data viewed as “objective” by the family. In this situation, performing an autopsy can avoid litigation. Finally, the autopsy can help identify inappropriate and hazardous practices with the goal of improving the safety of future patients. For all these reasons, anesthesiologists should request autopsy revival (1). References 1. Marwick C. Pathologists request autopsy revival. JAMA 1995; 273:1889 –91. 2. Goldman L, Sayson R, Robbins S, et al. The value of the autopsy in three medical eras. New Engl J Med 1983;308:1000 –5. 3. Blosser SA, Zimmerman HE, Stauffer JL. Do autopsies of critically ill patients reveal important findings that were clinically undetected? Crit Care Med 1998;26:1332– 6. 4. Battle RM, Pathak D, Humble CG, et al. Factors influencing discrepancies between premortem and postmortem diagnoses. JAMA 1987;258:339 – 44. 5. Charatan F. Medical errors kill almost 100,000 Americans a year [letter]. BMJ 1999;319:1519.