As a rule, studies of deaths that have occurred outside hospital tend to focus on forensic issues or sudden unexpected death. The author has found no studies of the frequency of medical autopsies of deaths outside hospital, either in Norway or internationally. A comparison with the autopsy material for the entire country from 2005, which primarily refers to deaths in hospital, shows that those who died outside hospital and were autopsied at Akershus University Hospital were significantly younger (12 and 11 years of difference in median age for men and women respectively), the proportion of men was higher (67 % compared to 56 % nationwide), neoplasms were more rare (4 % compared to 25 % nationwide) and that deaths from external causes were more frequent (12 % compared to 3 % nationwide) (11). The proportion of deaths from cardiovascular diseases was identical, which is due to a high proportion of deaths from this cause in the group with no known previous history and observed acute death.
When the current Act relating to hospital autopsies was written in the 1970s, most deaths occurred in hospital, and the autopsy frequency in institutions with a pathology department occasionally reached 80 % (3, 12). Today, most deaths occur in nursing and health institutions (4). These institutions only rarely have a tradition for requesting an autopsy. Only ten of the deaths in our material originated in nursing or health institutions.
The proportion of deaths occurring outside hospital or other health institutions has remained relatively stable, and account for approximately 20 % of the total (4). A large proportion of these deaths are unnatural according to the legislative definition. With the exception of those that obviously fall under this category, such as homicides, suicides and substance abuse, cases of unexpected death, deaths from unknown causes or deaths after recent contact with the health services may also be unnatural (Box 1). Many of these were therefore previously submitted to a forensic examination (13). In the early 1990s, the responsibility for the costs of forensic post mortems was transferred from the Ministry of Justice to the police districts – and the proportion of forensic post mortems fell sharply (8).
The statistics for deaths that are currently investigated forensically permit no detailed analysis on the basis of gender, age or diagnosis, and there are large geographic variations in practices related to requisitioning (13). Nevertheless, we may assume that the need for a better analysis of the group of deaths that occur outside institutions, which no longer are investigated forensically, may be one explanation for why the number of medical autopsies of these cases has increased, in the country as a whole as well as in Akershus University Hospital. Even though the autopsy records reviewed do not systematically register whether the death was reported as unnatural or not, the circumstances and the clinical information indicate that at least half of the autopsies must have related to deaths that were regarded as unnatural prior to this procedure.
The distribution of medical autopsies of deaths outside hospital between the first and second half of the year, shown in Figure 2, is striking. Since many of these cases fall within the definition of unnatural death, it is not unreasonable to speculate that the reason why the number of autopsies increases towards the end of the budget year may be found in restrictive practices with regard to requests for forensic post mortems in the police districts.
A forensic toxicological examination was deemed to be useful in 34 of the 89 cases of death (38 %). In four cases, such examinations revealed that the cause of death was intoxication with pharmaceutical drugs, in three cases probably ingested with suicidal intent. In this context, it is noteworthy that public suicide statistics fluctuate in line with the frequency of post mortems (14, 15). The cost of a forensic toxicology examination in the Norwegian Institute of Public Health exceeds the reimbursement rate for a medical autopsy of a death that has occurred outside hospital (7). The discrepancy between the cost and the reimbursement rate may induce the pathologists to be more restrictive in their requests for forensic toxicological examinations. The number of such examinations for the autopsies performed is therefore most likely the minimum deemed to be necessary.
The Cause of Death Registry has previously estimated that in 11.7 % of all medical autopsies performed in 2009, death had occurred outside a health institution, but it is uncertain whether deaths in ambulances are included in this estimate (8). For the same reason, the figures from the Cause of Death Registry that are used in this study are not comparable to the development observed in Akershus University Hospital. The fact that figures from the registry show an increase of 45 % from 2007 to 2010, while the number of medical autopsies has remained stable, is nevertheless an indication that the number of autopsies for deaths that have occurred outside hospital is increasing nationwide as well.