The predictive value of history and scene investigation for toxicology results in a medical examiner population

Medical examiner offices vary in the extent to which they pursue postmortem toxicology. Our office routinely tests decedents for ethanol and drugs of abuse, and we decided to evaluate the usefulness of our practice. We reviewed 1180 medical examiner cases examined in 2002-2003. History and scene investigation indicated that alcohol or drugs of abuse were likely to be detected in 369 cases, yet toxicology testing revealed an intoxicating substance in 589 cases, a prevalence of 50%. Screening for toxicology testing based on investigative findings had a sensitivity of 0.47, a specificity of 0.84, and a positive predictive value of 0.74. Moreover, even in the 811 cases where initial investigation did not suggest substance abuse, toxicology testing revealed at least 1 substance that was pertinent to the subsequent investigation in one third of the cases (260), and the intoxicating substance was sufficiently important to merit inclusion as a cause of death or contributing factor in nearly half of those cases (113). We conclude that investigation alone is ineffective at predicting the presence of intoxicating substances within decedents.