Overall, 857 autopsies were performed in the study period. Fifty-nine cases were excluded from the study due to a lack of relevant information in autopsy reports. The demographic and circumstantial characteristics of the 798 cases included are shown in Table 1. Most cases were males (70.7%) and of Italian nationality (77.4%), with a mean age of 55.6 (SD: 17.2) years, with half of the subjects aged between 30–60 years (52%) followed by those above 60 years (41%). As for manner of death, natural death (66.9%) was the most common, followed by accidental deaths (21.7%), suicide (6.9%) and homicide (4.5%). As far as mechanism of death is concerned, cardiac death was the most common (53.5%), followed by traumatic/hemorrhagic shock (17.9%). The remaining causes of death were asphyxia (8.0%), MOF (Multi Organ Failure) (7.9%), cerebral/pulmonary edema (4.6%), cerebral hemorrhage (4.5%), pulmonary embolism (2.8%) and others (0.7%). Among unnatural deaths, traffic accident (27.3%) prevailed, followed by acute intoxication (25.8%) and asphyxia deaths (17.1%). The most common place of death was a private house (67.3%), followed by public space (18.7%), health facility (8.2%), workplace (33.0%) and place of confinement (1.8%). In terms of corpse conservation state, the majority of the individuals were characterized by early changes (79.5%), followed by decomposition (19.3%), carbonization (1.0%), skeletal stage and mummification (one case each, only). Toxicological analyses were requested in 504 cases (63.0%): 330 (65.5%) had a positive toxicological profile, while 174 (34.5%) were negative. Toxic substances were considered the leading cause of death in 68 cases (13.5%).

Negative results were detected in 174 of 504 cases for which toxicological investigations were carried out between 2017–2022. The population was composed of 48 females and 126 males A majority of the individuals were Italians (75.9%). Death occurred in autumn in 48 cases (27.6%), followed by spring and winter (25.9% each), and summer (20.7%). Mean age was 51.0 ± 18.5. The age group between 30 and 60 years (55.8%) was the most widely represented. The most common cause of death in these negative cases was natural death (67.8%), followed by accidental death (19.0%), homicide (6.9%) and suicide (6.3%). The most common cadaveric state was early changes (85.6%). No skeletal and mummification stages have been recorded. The most common place of death was a private house (64.4%), followed by public space (23.0%), workplace (8.1%), health facility (4.0%) and place of confinement (only onecase).

Positive results were detected in 330 of the 504 toxicological investigations. These cases included 234 males (70.9%) and 96 females (29.1%), mostly Italians (72.7%). A larger share of death in private house settings (63.3%) emerged, followed by public space (22.1%), health facility (7.0%), workplace (3.9%) and, finally, place of confinement (3.6%). Occurrence of death was slightly higher in the summer (28.8%) than in winter (26.4%) and spring (25.2%), and far more higher than in the autumn (19.7%). The manner of death in toxicological positive cases was mainly natural death (50.0%), followed by accidental death (35.2%), suicide (10.9%) and homicide (3.9%). As far as cadaveric state, the toxicological analysis were reported in cases of early changes (73.3%), followed by decomposition (24.6%).

Table 2 summarizes the association between risk of positive toxicological findings and relevant demographic and circumstantial characteristics, in both crude and adjusted analyses. No association emerged for sex, while a positive association with increased ORs was reported for the 30–60 age group and, in part, for those > 60 years compared to those aged < 30 years. As regards place of death, we found a positive association with residence in health facilities and places of confinement, and a strong inverse association with workplace. A strong positive association was found for accidental death and suicide in terms of manner of death. In the multivariable model, the correlations in the unadjusted analysis were generally confirmed and strengthened.

Overall, 139 different analytes were detected in toxicologically positive cases. These substances were subdivided into the followed classes: ethanol, carbon monoxide, benzodiazepines, antipsychotics, antidepressants, anticonvulsants, opioid agonists, drugs of abuse, antihypertensive and others (Fig. 2). The most frequently observed classes of substances were benzodiazepine (41.2%), followed by alcohol (35.5%) and drugs of abuse (24.8%). Full details about the detected substances are reported in Supplementary Table S2.

Distribution of positive toxicological findings. Different groups of compounds are depicted using decreasing grey-scale patterns and in the order of the legend

Figure 3 shows the distribution of the identified classes across groups in relation to manner of death: the most widely detected class of substances was benzodiazepines, with a prevalence of 68.6% in suicide cases, followed by accidental deaths (42.9%), homicide (38.5%) and natural deaths (34.7%). Drugs of abuse were most frequently found in accidental deaths (39.3%), while the lowest number was found in suicides (5.7%).

Relative to alcohol class, no substantial differences between manner of deaths groups have been highlighted (Fig. 3).

Distribution of positive toxicological findings divided by manner of death. Different groups of compounds are depicted using decreasing grey-scale patterns and in the order of the legend

The occurrence of two or more classes of substances was also investigated. Supplementary Figure S1 shows the co-occurrence rates of the investigated toxicological substances divided by main types and number of classes. For all classes of substances, two or more classes were detected in at least half of the samples. As regards licit (prescribed) drugs, the percentage of the number of different substances detected at the same time in positive cases (poly-assumption) is reported in Fig. 4. In approximately 40% of cases, the subject assumed 2 or more drugs with a median value of 1 drug (IQR: 0–3). Polypharmacy (i.e. ≥ 5 drugs) occurred in 6.1% of cases (N = 20). Supplementary Figure S2 reports the distribution of licit drugs stratified by sex and age categories. Slightly lower median values were found in males (N = 1, IQR: 0–2) compared to females (N = 1.5, IQR: 1–3), but with 6.8% (N = 16) and 4.2% (N = 4) of polypharmacy, respectively. Analysis divided by age demonstrated that younger subjects (< 30 years) generally assumed less prescribed drugs compared to older individuals (Supplementary Figure S3).

Distribution of number of prescribed (licit) drugs in positive cases (N=330). The numbers in the x-axis represent the simultaneous occurrence of prescribed drugs

Figure 5 presents spider plots with rates of contemporary detection for the investigated classes. Except for antihypertensives and carbon monoxide, multi-detection of classes of substances can be noted. In particular, benzodiazepines were detected in 30% of all cases, with higher rates over 60% in association with antidepressants, antipsychotics, anticonvulsants, and opioids. In addition, benzodiazepines have been detected alongside alcohol in more than 20% of all cases. Similarly, both antipsychotics and antidepressants (though with lower rates of co-detection compared to benzodiazepines) were found in over 30% of subjects who are also positive to benzodiazepines, antidepressants, anticonvulsants and opioids.

Spider plots of xenobiotic positivity divided by main classes of substances. For each xenobiotic, the percentage of co-occurrence of other classes is reported

Toxic substances were neither considered as the first cause nor a fundamental cause of death in 256 cases, while the cause of death was ascertained as fatal acute intoxication in 68 cases thanks to toxicological findings (13.5% of the total number of cases with toxic investigation and 20.6% of the positive cases).

Supplementary Table S2 outlines drugs and illicit substances detected through toxicological analysis. In terms of individual analytes, ethanol was found in 117 cases (23.2% of all cases and 35.5% of all toxicologically positive cases), benzoylecgonine in 60 cases (11.9%—18.2%) and lorazepam in 54 cases (10.7%—16.4%).

Among illicit drugs, the most highly consumed substance was cocaine (n = 54, 10.7%—16.4%; benzoylecgonine: n = 60, 11.9%—18.2%), followed by heroin (6-MAM: n = 31, 6.2%—9.4%; morphine: n = 38, 7.5%—11.5%) and cannabis (11-Nor-9-carboxy-delta-9-tetrahydrocannabinol: n = 22, 4.4%—6.7%, delta-9-tetrahydrocannabinol n = 19, 3.8%—5.8%, and 11-Hydroxy-delta-9-tetrahydrocannabinol n = 16, 3.2%—4.8%). Additionally, cocaethylene was also detected n = 34, 6.7%—10.3%). Among benzodiazepines, lorazepam (n = 54, 10.7%—16.3%) and diazepam (n = 34, 6.7%—10.3%) were the most widespread. Among antipsychotics, the most widely used was quetiapine (n = 24, 4.8%—7.3%), while trazodone and citalopram were the most frequently detected antidepressants, retrieved in 15 (3.0%—4.5%) and 13 cases (2.6%—3.9%), respectively. Among opioid agonists, the most widely used was methadone (n = 24, 4.8%—7.3%). Among anticonvulsants, pregabalin (n = 15, 3%—4.5%) was most widespread, as was amlodipine among antihypertensives (n = 16, 3.2%—4.8%).

Supplementary Tables S3-S11 present crude and adjusted OR of positivity to classes of toxics, also considering subjects’ demographic and circumstantial characteristics. For drugs of abuse, an inverse association was found for female sex, age > 60 years, and non-Italian nationality. Similarly, all places of death showed a negative association, compared to private house. Conversely, a direct association was present for accidental manner of death.

Detection of alcohol occurrence was inversely associated with female sex and increasing age, and directly with non-Italians. Public space showed a stronger positive association compared to private house and other places of death. Association with manner of death was generally positive, although weaker compared to other characteristics.

Benzodiazepine detection was strongly associated with female sex and age older than 30 years, while it was the opposite with non-Italian nationality. As for places of death, work and public spaces showed inverse association, while a positive one could be noted for health facilities and place of confinement, compared to private house. The autumn season tended to be positively associated with benzodiazepine presence as well as both accidental death and suicide. Antidepressants showed a pattern similar to benzodiazepines, except for a weak negative association with health facility as place of death. They showed no association with season, and a weak direct association with manner of death.

For antipsychotics, a negative association was found for non-Italians and public spaces. Conversely, a strong positive association was noted for place of confinement and both accidental death and suicide. Anticonvulsants showed a weak positive association with female sex and with age > 30 years. Similarly, a positive association could be noted for health facility and place of confinement. All other seasons compared to spring showed a negative association, stronger for winter and summer. A positive association was also found with accidental death, suicide and homicide. Antihypertensives were more commonly detected in non-Italians, individuals aged > 60 years and for deaths occurred in work and public spaces. For all other characteristics, no clear associations emerged. Opioid agonists were more commonly detected in individuals aged more than 30 years, as compared to younger subjects. Places of confinement showed the strongest direct association, as did accidental death. Carbon monoxide was more commonly detected in females, over 60 years and in non-Italians. Carbon monoxide was strongly associated with accidental death. A direct weak association was also observed with the winter season.