In this study, we investigated a decade of fatal PTEs at our institute with the advantages and limitations of forensic work.

Generally, forensic autopsy-based studies that have investigated fatal PTEs from the last decade have shown fairly heterogeneous cohorts with mean ages of decedents ranging from 40 to 67.8 years and a varying male: female ratio (0.5–1.5:1) [15,16,17,18,19,20,21]. Although the results of our study add to the heterogeneity of the overall forensic case work, our cohort constitutes the more centrist parts of the intervals. Further corroborating the heterogeneity, the prevalence of a fatal PTE ranges from 0.6 to 4.6% in autopsy cohorts [15,16,17,18, 20, 21]. This heterogeneity may reflect the basic limitations of forensic autopsy-based studies, with an inherent selection bias of cases determined by the police and/or authorities.

From a Danish perspective, annual reports from 2010 to 2019 on CODs from Statistics Denmark revealed that a total of 1,844 out of a total of 527,492 deaths (0.3%) were due to PTEs [22]. A recent Swedish study revealed that the decreasing prevalence of clinical autopsies overall reduced the value of the COD registry [23]. In Finland, where the forensic autopsy rate has remained stable at approximately 30% of all deaths for many years and has only recently experienced a decline, fatal PTEs constituted 2.3% of all deaths [24]. This finding may indicate that the true prevalence of fatal PTEs is underestimated in the Danish COD registry.

Based on several forensic autopsy-based studies that have examined fatal PTEs, the mean age of decedents appears to follow a bimodal distribution. For example, the present study, along with other studies, identified mean ages in the prime of the 50th decade [16, 19, 21], and others reported a mean age in the middle or ultimate part of the 60th decade [15, 18, 25]. Although general guidelines for referral for a forensic autopsy are similar across many countries, i.e., when death is sudden and/or unsuspected, certain details and national practices vary. Denmark has an inversely proportional relationship with advancing age and the probability of referral for a forensic autopsy [26], which may, to some extent, explain the relatively low mean age of our cohort. Although investigating national differences in national practices for referrals for a forensic autopsy is beyond the scope of the present study, we highlight the results from a study by Ylijoki-Sørensen et al., which revealed that broad referrals for a forensic autopsy in Finland had a marked impact on the low number of unknown CODs [27].

With respect to misclassifications of the COD in the clinical setting, an autopsy is valuable, especially with respect to PTEs. In the systematic review by Winters et al., PTE along with myocardial infarction, pneumonia, and Aspergillus infection constituted more than one-third of all Class I misdiagnoses according to the Goldman classification scheme, with PTE eclipsing the other three causes in the total number of studies identifying Class I and Class II misdiagnoses [9]. Winters et al. additionally proposed a mathematical model with a hypothetical value of 100% of decedents from a hospital intensive care unit undergoing an autopsy and reported an estimated incidence of 6.3% Goldman Class I misdiagnosis [9]. Furthermore, Ylijoki-Sørensen et al. reported that a systematic investigation with a medicolegal autopsy in all cases with an unclear COD only 2–3/1000 deaths would be coded as having an ill-defined or unknown COD compared with the present number of 41/1000 deaths [27].

Although a previous Danish study on acute PTE reported that cancer was the most prevalent risk factor [5], only approximately 1/6 of the deceased individuals in our cohort had a cancer diagnosis. This finding most likely reflects that deceased individuals with a known, potentially lethal disease—i.e., cancer—are less likely to be referred for medicolegal inquest and a potential subsequent forensic autopsy.

The impact of a patient’s BMI combined with the previously mentioned nonspecific clinical symptoms should raise the suspicion of a PTE. In accordance with other studies [15, 17, 19, 21, 25], the vast majority of our decedents were defined as overweight or obese. The prevalence of obesity in Denmark has increased during recent decades, and in 2021, almost one in every five adults (18.4%) could be classified as obese [28]. Although results from a forensic, retrospective study design rarely can be directly extrapolated to the general population, our rather consistent finding of fatal PTE in persons with a BMI > 25 and even more often with a BMI > 30 underscores the morbidity and mortality of obesity.

Many of the decedents in our study died in an out-of-hospital setting, which emphasizes the sudden and potentially fatal outcome of a PTE. Furthermore, only a few experienced symptoms prior to death, which is in line with previous studies [18, 21] and highlights the insidiousness of the condition.

Several genetic variations are associated with an increased risk of developing a DVT and, potentially, a subsequent PTE. A comprehensive autopsy-based study by Meißner et al. identified several genomic alterations related to an increased risk of DVT at different parts of the coagulation and fibrinolysis cascade. Specifically, genetic modulations related to FVL, Factor XIII, Beta fibrinogen, TFPI, and HIVEP1 were emphasized as points of interest in additional studies [12].

At our institute, we only referred cases for investigating FVL mutations in selected cases and only ¼ of our included cases were analyzed for this mutation. Fineschi et al. conducted a systematic, autopsy-based study investigating the prevalence of genetic mutations in cases of fatal PTE and reported that only 1/43 decedents had an FVL mutation [20]. We did not systematically investigate genetic mutations in our cohort, but the low prevalence of FVL mutation (2/58) may corroborate the conclusions of Fineschi et al. that only young decedents without major risk factors should be screened for genetic-based thrombophilia [20].

Although FVL mutation significantly increases the risk of a PTE [29], the mutation is associated with an even greater risk of developing a DVT. This discrepancy between the two entities in FVL mutations is dubbed the Factor V Leiden paradox. This paradox underscores the necessity of distinguishing the two conditions separately in thrombophilia studies and the importance of looking beyond FVL mutations in cases of fatal PTEs [12].

Because an autopsy is still regarded as the gold standard diagnostic tool for PTEs, an autopsy-based study into the characteristics of fatal PTEs constitutes a prime basis for the correct inclusion of cases. However, as with all studies based on forensic autopsies, the deceased individuals that are referred for a forensic autopsy are subject to inherent selection bias, which results in a small sample size. Because referral for a forensic autopsy and the age of the decedent are inversely related in Denmark [26], the current study focused on fatal PTEs, for which age is a known risk factor. Therefore, an unknown number of cases of fatal PTEs have not been referred for a forensic autopsy and ultimately were not included in our material. Finally, this study was retrospective; therefore, we relied on information extracted from police records, which precluded direct influence on the type of information that was available.