Cannabis is the most widely used illegal drug but is rarely considered a causal factor in death. Indeed, its use‐related toxicity is considered insignificant in fatal cases [4]. In support of this theory, it should be considered that the psychotropic effects that occur when taken in high doses are quite unpleasant and, before dangerous reactions occur, tend to self‐limit further intake [23, 24]. For the above reasons, cases of adult individuals dying after cannabis assumption, regardless of the modality, are very rare, if not extremely rare. The only detectable cases in the literature involve adult subjects, often suffering from preexisting disease patterns. Drug‐induced deaths are rarely considered in the analysis of fatal cases involving children, even though the increasing legalization of drugs is leading to an increasing amount of drug intoxication even in young subjects.

Regarding the presence of THC and its metabolites 11‐OH‐THC and THC‐COOH in the blood sampled in the emergency room, the concentrations of 14, 6.9, and 77 ng/mL, respectively, are widely within the ranges reported in the literature as associated with cases of intoxicated subjects after cannabinoid intake [2, 25, 26]. It must be considered that the values obtained from toxicological analysis refer to whole blood and not plasma, as occurs in the relevant literature. Considering the low partition coefficient of erythrocytes [22], the concentrations measured in plasma are twice as high as those obtained in blood, and therefore, the latter should be considered doubled in view of performing an adequate comparison.

Furthermore, regarding the interpretive blood cut‐offs in use in major European countries for the assessment of disabling effects related to cannabinoid use in adults (Belgium–THC: 1 ng/mL; Luxembourg–THC: 1 ng/mL; Germany–THC: 1 ng/mL; United Kingdom–THC: 2 ng/mL) [27], it is evident that compared with these interpretive values, those found in our case are much higher.

Supporting the above is the finding of a symptomatology consistent with that known to be associated with intoxication by THC and the active metabolite 11‐OH‐THC [2]. In fact, from 1 p.m. onward, the child staggered with loss of balance and appeared absent. While being transported to the hospital, the child was inclined to fall asleep and, when admitted to the hospital, showed a tendency to doze off showing bradypnea with apnea.

Regarding the pharmacokinetic profile of cannabis, it is necessary to consider that the lower amount of adipose tissue typically present in children compared with adults justifies a lower distribution of the drug in that tissue resulting in higher concentrations in the blood than in adults, considering the same amount of absorbed substance [28]. In addition, children up to the age of 3 years possess a lower gastric pH than that of the adult, implying a higher bioavailability of THC due to less degradation of the substance at the gastric level [29].

According to a model proposed by Huestis in the living subject, it would be possible to cautiously estimate, based on the blood concentration of THC and THC‐COOH, the elapsed time since the last intake. In this case is estimated that the intake had occurred approximately 1:50 h (1 h 46′) before blood sampling. This result should be considered purely as indicative, since the formula was originally intended for estimating the elapsed time since intake in the case of adult subjects taking THC as a smoke. Moreover, Huestis' predictive model should be interpreted cautiously when applied to cadaveric samples, since THC is subject to post‐mortem redistribution.

Considering that cadaveric blood in forensic cases is not usually immediately sampled and the corpse is not immediately put in a rephrigerated environment, given both the redistribution phenomena and the degradative kinetics of THC in cadaveric blood [30], it can be hypothesized that blood THC concentration in the child moments before death was higher than that observed in the cadaveric sample.

Considering all of the above, it is believed that the toddler was in a state of acute cannabinoid intoxication at the time of the access to the emergency room.

In deaths studied in animal models with acute administration, the dose of THC that kills 50% of animals (LD50) is 40 mg/kg intravenously in the rat and 130 mg/kg in the dog and monkey [10]. Extrapolating from animal to human, it would result in an e.v. lethal dose of between 2.6 and 8.45 g THC for a 65‐kg adult. When administered orally, LD50 in rats is over 500 mg/kg, which is a human equivalent oral dose of about 806.45 mg/kg [31]. Considering a 10‐kg child, the lethal oral dose of THC would be approximately 8 g. In our case, 1 g of hashish (presumably at 28% THC) was measured in the gastric content. It is to be noted that the child ingested the substance a couple of hours before the admission to the hospital, so its enteric absorption had already begun, even if it was not complete at the time of death. It is reasonable to assume that the dose ingested was significantly higher, even if it is difficult to estimate.

Therefore, in the absence of other causes potentially responsible for death, it can be stated that the infant's death is causally attributable to the ingestion of cannabis.

Finally, according to fentanyl and ketamine pharmacokinetics [32, 33], the detection of fentanyl, ketamine, and norketamine at concentrations of 1.8, 726, and 265 ng/mL, respectively, is consistent with the documented administration of these sedatives at the dosage of 20 mcg (fentanyl) and 30 mg (ketamine) for orotracheal intubation maneuvers.

Regarding the autopsy sampling, the THC concentrations detected in the peripheral blood sample (6.9 ng/mL), considering the plasma‐to‐blood ratio as explicated in the previous paragraph, are well above those reported in literature as causative of death in adults without pathological preexisting conditions [3]. Furthermore, the urinary presence of THC [34, 35] and the metabolites 11‐OH‐THC and THC‐COOH at the concentrations of 4.6, 3.8, and 51 ng/mL, respectively, is indicative of a recent use of cannabis, as explained hereafter. Given the relatively high concentrations of THC and metabolites (THC‐COOH and 11‐OH‐THC) found in blood, the relatively low concentrations detected at the urinary level can be attributed to the initial and thus incomplete metabolism of the active ingredient THC and to the early stages of its excretion when death occurred [5], nor is any data available on the metabolic efficiency of the child.

As to keratin matrices, the detection of cannabinoids in head hair (THC, 0.91 ng/mg; CBD, 0.030 ng/mg; and CBN, 0.030 ng/mg) implies several mechanisms of incorporation: by blood capillaries, through sebum or sweat [22] or through hair contact with the sweat of a third active user [36]. Given the lipophilicity of the THC molecule, its ability to bind to hair melanin is reduced; the concentrations detected in hair are therefore lower than they would have been in the case of other substances of abuse [22]. The presence of fentanyl at a concentration of 0.001 ng/mg does not contrast with the documented administration of this sedative drug at a dosage of 20 mcg in the ER at about 2:10 p.m. As regards the detection of ketamine and its metabolite norketamine, in concentrations of 1.38 and 0.28 ng/mg, respectively, it is indicative of systemic exposure to the substance. On this point, it should be noted that the first‐aid report shows the administration of ketamine at a dose of 30 mg, necessary for the orotracheal intubation maneuvers. Given the high concentrations found, it is reasonable to infer that the incorporation of both THC and ketamine in the hair was due to the sweat produced by the subject in the agonic or preagonic phase. However, especially for cannabinoids, it cannot be excluded that there was also an exposure in the 3 months before death. Before considering the other drugs found in the hair of the victim, it should be noted that children's hair is finer and more porous than those of an adult, making them at greater risk of sweat contamination [37].

As regards cocaine, the absence of cocaine and its metabolites in the blood, urine, and bile reveals that the child was not exposed to cocaine recently and was not intoxicated by cocaine just before death. The same consideration holds for methadone, EDDP, morphine, codeine, and 6‐AM absent in the cadaveric fluids/tissues and present only in hair, thus excluding a recent acute heroin and methadone intoxication. However, the presence of metabolites (benzoylecgonine, norcocaine, EDDP, morphine, codeine, and 6‐AM) may suggest that the parent compounds underwent metabolism in the child's body prior to incorporation into hair. It is not possible, however, to determine whether these were taken by the young victim (e.g., by accidental ingestion) or by adults with whom the child had contact, resulting in the transfer from their sweat to the toddler's hair. However, based on the concentrations found on keratin matrices, it can be affirmed that the child had certainly been exposed to heroin, cocaine, and methadone either actively or passively.

Considering THC concentrations in fluids, these are compatible with the development of symptoms of acute central neuronal intoxication. In fact, from early afternoon, the child presented recurrent episodes of bradypnea with subsequent apnea. The acute toxicity of cannabis is due to THC and is mediated by its effects on the neuronal system. In fatal cases caused by acute administration and derived from animal models, apnea always occurs or cardiac arrest if apnea is prevented [10]. Data from the analysis of the liver sample, which resulted in the detection of 11‐OH‐THC and THC‐COOH at concentrations of 9.8 and 81.0 ng/mg, respectively, indicate that cannabis has been ingested, absorbed, and metabolized in the liver.