Clinical Diagnoses and Autopsy Findings: A Retrospective Analysis of 252 Cases in Greece

Archives of Pathology & Laboratory Medicine, Feb 2005 by Spiliopoulou, Chaido, Papadodima, Stavroula, Kotakidis, Nikolaos, Koutselinis, Antonios

Context.-Despite medical and technologic advances, clinicians may misdiagnose a patient's situation and the cause of death. Autopsy may be valuable in uncovering the most frequent diagnostic pitfalls and helping clinicians to learn and to develop the medical art and science.

Objective.-To compare the clinical diagnoses with postmortem findings and evaluate the frequency of diagnostic errors assessed by autopsies.

Design.-Retrospective analysis of the protocols of 252 consecutive cases of adult patients autopsied in the Department of Forensic Medicine and Toxicology of Athens Medical School during the period 1999-2003. The outcome measures included concordance between diagnosis before death and at autopsy, sex, age, and length of hospitalization of the patient.

Results.-In 73 cases (29%), the autopsy findings confirmed the clinical diagnosis and the cause of death suggested by the clinicians. In 45 cases (19%), the clinical diagnosis and the cause of death suggested by the clinicians were discordant with the autopsy findings. In 105 cases (42%), the autopsy requests did not include any suggestion about the cause of the patient's death. In 7 cases (3%), several diagnoses were suggested by the clinicians, and in 16 cases (6%), the comparison between clinical and postmortem diagnosis was not possible. The most frequently misdiagnosed diseases were coronary disease and pulmonary embolism.

Conclusions.-It is concluded from this study that autopsies may reveal unexpected findings that are of critical importance and that a continued emphasis on autopsy evaluation is necessary to improve the quality of patient care.

(Arch Pathol Lab Med. 2005;129:210-214)

The importance of autopsy in clinical practice has long been the subject of discussion, and it has recently attracted even more interest. Autopsy is a valuable tool for evaluating the accuracy of clinical diagnosis, investigating and discovering unsuspected diseases, and recognizing causes of death. Moreover, it is an essential element of medical auditing and teaching, as well as for the shady of cause, natural history, and course of diseases. However, the relative number of autopsies performed over recent decades has been declining worldwide.1-4 This fact has been attributed to increased costs, fear of malpractice litigation, and advances in medical technology.5,6

Undoubtedly, the recent technologic advancements in medical areas have allowed more sensitive and reliable methods for clinical diagnosis during life; nevertheless, significant discrepancies between clinical diagnosis and autopsy findings for patients who died in the hospital have been reported, and it is possible that autopsy may reveal unexpected data that, had the information been known before death, may have changed patient management.7-11

Autopsies in Greece are performed by forensic pathologists only after a juridical order. Autopsy is mandatory in case of sudden or violent death. Most cases concern patients whose situation or short stay in the hospital did not allow the establishment of a clear cause of death. In a smaller number of cases, the order for an autopsy may arise as a consequence of a legal procedure for medical negligence.

The aims of the present study were to compare the clinical diagnosis with the autopsy findings of cases referred to the Department of Forensic Medicine and Toxicology of Athens Medical School during the period 1999-2003 and to highlight the most common discrepancies. It is worth mentioning that, to our knowledge, it is the first such study conducted in Greece.

MATERIALS AND METHODS

A total of 252 cases of patients who were hospitalized and died in a hospital and whose postmortem examination was referred to our department during the period 1999-2003 were retrospectively analyzed. The cases were consecutively selected according to the registration number of the autopsy records. The criteria for exclusion were as follows: (1) age of patient younger than 15 years old, (2) time of stay in hospital less than 1 hour, and (3) violent death. Data concerning the clinical diagnosis established during life, as well as the cause of death suggested by the clinicians, were collected from the autopsy requests.

Complete autopsy was performed within 24 hours of death, and the procedure included macroscopic examination and histopathologic investigation of the internal organs. Toxicologic analysis was also performed when indicated. Causes of death were classified according to the 9th edition of the International Statistical Classification of Diseases and Related Health Problems,12 which is still in use in Greece.

Data were grouped into 5 categories: (1) cause of death suggested by the clinicians and confirmed by the autopsy findings: concordant cases; (2) cause of death suggested by the clinicians but not confirmed by the autopsy findings: discordant cases; (3) cause of death not suggested by the clinicians; (4) several causes of death suggested by the clinicians; and (5) comparison between clinical and postmortem diagnosis not possible.