Evaluation of postmortem toxicological results

Sammanfattning: Postmortem diagnosis of fatal intoxications is complicated. Autopsy findings indicative of an intoxication related death are scarce and mostly unspecific. Instead, the diagnosis is more dependent on the circumstances surrounding death, and on the toxicological results. In order to correctly evaluate the toxicological results, the forensic investigator needs to know the concentrations of various substances that can be considered “normal” and which concentrations that may be “lethal”. While a wealth of scientific data exist concerning concentrations and effects in experimental animals and in living human subjects, these data cannot simply be translated into postmortem concentrations. Due to various changes occurring already in the early phase after death, the blood postmortem concentration of a substance postmortem often does not mirror the concentration antemortem. In order to correctly evaluate postmortem toxicological results there is a need for postmortem reference values, allowing for the separation between the lethal and non-lethal concentrations. Toxicological data based on the analysis of whole blood from the femoral vein collected at autopsies conducted between 1992 and 2010 were used to establish multiple categories of reference concentrations; Group A (intoxications with a single substance only), Group B (mixed intoxications) and Group C (controls). A well-defined automated selection process followed by a standardized multi-reviewer process and a case-by-case evaluation were applied to generate reference values for each group. Paper I-IV present postmortem reference concentrations of 48 substances. In most cases there are separate reference concentrations for different types of intoxications (group A and group B) as well as controls (group C). In general, there is a substantial overlap in concentrations between the two intoxication groups, whereas the controls often show much lower levels. In addition to the establishment of postmortem reference concentrations the different studies included in this thesis provide some additional information. Paper I presents a fatal toxicity index of sedative and hypnotic drugs; Paper II presents the occurrence of antipsychotics among different manners of death; and in Paper III an attempt is made to assess differences in concentrations between acute, acute-on-chronic and chronic intoxications. Further, Paper IV presents the impact of sample size on the precision, power and risk for type 1 error of postmortem reference concentrations. Using the method presented in paper I-IV, >5 cases in each group are needed to reduce the risk of type 1 error (<5%) and >10 cases in each group is needed to have a high power (>0.95). It is suggested that a target number of between 20-30 cases in each group provides a reasonable stability of the reference concentration.