Is the clinical autopsy no longer relevant?

Abstract

Background: The total number of autopsies has dropped during the last decades to less than a quarter of the number during the 1980-ies. We studied the consequences of a decrease in the autopsy frequency for the accuracy of the death certificates and the cause of death registry.

Aim: The overall aim of the project was to study the background of the low levels of autopsies. One specific aim was to investigate the clinical usefulness of autopsies for determining the cause of death and as a source of information for relatives. (I) Another aim was to describe the changes over time in clinical and forensic post-mortem examinations of males and females at different ages and to study the relevance and accuracy of the cause of death registries from the National Board of Health and Welfare and how this has been affected by the decline in the autopsy rate. (II)

Methods: In study I the process of post-mortem examination was investigated. A total of 53 autopsies was studied using data from medical records, including nurse’s records, autopsy request forms, autopsy reports and death certificates. Study II was a national registry study (1969–2016) of mortality statistics and a local retrospective study (2008–2009) of medical records and death certificates on the cause of death of patients who died at a university hospital.

Results: The autopsy routines were sufficient to fulfill clinical needs in a reasonable time. In six cases, kin requested the examination, while in half of the cases, autopsy was recommended by the responsible physician only. In two-thirds of the cases, the autopsy significantly altered or complemented the clinical picture in a way that was noted in the death certificate. In two-thirds of the cases, it was documented that the next of kin was informed of the result of the autopsy. (I) The total autopsy rate has fallen from 40,000 autopsies to about 10,000 autopsies per year from 1969–2016. Clinical autopsies performed in the healthcare system have dropped from 30,000 to less than 5,000 per year. The frequency of clinical autopsies was age dependent and had low values after the age of 70. We could not confirm sex-dependent differences in the rate of clinical autopsies. We investigated 221 deaths at Karolinska university hospital from 2008–2009. 188 cases were not autopsied and in 26 of these the cause of death was unspecific and irrelevant, while in 33 cases in which an autopsy had been performed, only 2 had an unspecified cause of death. (II)

Conclusion: The data confirm that clinical autopsy is a valuable source of information for the death certificate and for the next of kin. The diagnostic value of autopsy was the same as when autopsy frequencies was above 50%. We conclude that the decrease in number of autopsies requested and performed reduces the value of the cause of death registry and should be increased to improve the understanding of disease and cause of death, as well as to better inform the next of kin.