Anesth Analg 2000;91:344-346
© 2000 International Anesthesia Research Society
CRITICAL CARE AND TRAUMA
Postoperative Death and Malpractice Suits: Is Autopsy Useful?
Philippe Juvin, MD*,
Frédéric Teissière, MD
,
Fabrice Brion, MD,
Jean-Marie Desmonts, MD*, and
Michel Durigon, MD
*Service d’Anesthésie et de Réanimation, Hôpital Bichat, Paris; and
Service de Médecine Légale, Hôpital Raymond Poincaré, Garches, France
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.
 |
Abstract
|
|---|
Implications: This report demonstrates the extremely high yield of autopsies performed in the case of postoperative death with suspicion of malpractice. They frequently identified undetected complications. They could also suggest faulty or negligent practice that would otherwise go unrecognized. This report supports the widespread use of autopsies to investigate perioperative death.
|
Introduction
|
|---|
The 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 (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 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 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
|
|---|
-
Marwick C. Pathologists request autopsy revival. JAMA 1995;273:1889–91.[Abstract/Free Full Text]
-
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.[Abstract]
-
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.[Web of Science][Medline]
-
Battle RM, Pathak D, Humble CG, et al. Factors influencing discrepancies between premortem and postmortem diagnoses. JAMA 1987;258:339–44.[Abstract/Free Full Text]
-
Charatan F. Medical errors kill almost 100,000 Americans a year [letter]. BMJ 1999;319:1519.[Free Full Text]
Accepted for publication April 26, 2000.
This article has been cited by other articles:
|
|
|
|
 
M H C Grade, S Zucoloto, J K Kajiwara, M T P Fernandes, L G F Couto, and S B Garcia
Trends of accuracy of clinical diagnoses of the basic cause of death in a university hospital
J. Clin. Pathol.,
April 1, 2004;
57(4):
369 - 373.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
