Context: Some authors have proposed that post-mortem drug concentrations in bile are useful in estimating concentrations in blood. Both The International Association of Forensic Toxicologists (TIAFT) and the US Federal Aviation Administration recommend that samples of bile should be obtained in some circumstances. Furthermore, standard toxicological texts compare blood and bile concentrations, implying that concentrations in bile are of forensic value. Aim: To review the evidence on simultaneous measurements of blood and bile drug concentrations reported in the medical literature. Methods: We made a systematic search of EMBASE 1980-2016 using the search terms ("bile/" OR "exp drug bile level/concentration/") AND "drug blood level/concentration/", PubMed 1975-2017 for ("bile[tw]" OR "biliary[tw]") AND ("concentration[tw]" OR "concentrations[tw]" OR "level[tw]" OR "levels[tw]") AND "post-mortem[tw]" and also MEDLINE 1990-2016 for information on drugs whose biliary concentrations were mentioned in standard textbooks. The search was limited to human studies without language restrictions. We also examined recent reviews, indexes of relevant journals and citations in Web of Science and Google Scholar. We calculated the bile: blood concentration ratio. The searches together yielded 1031 titles with abstracts. We scanned titles and abstracts for relevance and retrieved 230, of which 161 were considered further. We excluded 49 papers because: the paper reported only one case (30 references); the data referred only to a metabolite (1); the work was published before 1980 (3); the information concerned only samples taken during life (10); or the paper referred to a toxin or unusual recreational drug (5). The remaining 112 papers provided data for analysis, with at least two observations for each of 58 drugs. Bile:blood concentration ratios: Median bile: blood concentration ratios varied from 0.18 (range 0.058-0.32) for dextromoramide to 520 (range 0.62-43,000) for buprenorphine. Median bile concentrations exceeded blood concentrations by one order of magnitude for several drugs, including dihydrocodeine, quetiapine and sildenafil; and by two orders of magnitude of for buprenorphine, colchicine and 3,4-methylenedioxy-methamphetamine (MDMA), among others. The minimum and maximum values for the ratio differed by a factor of three or more in three-quarters of the cases where data were available and by a factor of 10 or more for over half of the analytes. Limitations: The data were difficult to find. Medline does not explicitly index the term "drug bile concentration". It may well be that other reports exist, although they would not alter our major conclusion. Many of the papers that contributed data failed to specify the source of the blood samples or the post-mortem interval, so that no judgment was possible regarding post-mortem redistribution in whole blood or bile. Conclusions: For most drugs, there are wide ranges of bile:blood concentration ratios, which means that bile and blood concentrations are generally poorly correlated. Bile concentration measurements cannot readily be used to establish post-mortem blood concentrations; nor can they be extrapolated to ante-mortem concentrations. However, because drug concentrations in bile often exceed those in blood, bile may allow qualitative identification of drugs present, even when the blood concentration is below the limit of detection.
