This recent study by Dr. Hilary J. Hamnett and Dr. Itiel E. Dror looked into the possible effects of contextual case information in forensic toxicology testing. Two experiments were conducted in this study. The first focused on the interpretation of immunoassay screening data and the error rates of the participants’ conclusions. The participants were asked to screen for opioids and some were given case information while others were not. The second experiment asked participants to choose which toxicology tests to perform on five postmortem cases. All participants were given a brief history of the case and some were given demographic information like age and ethnicity as well.
In the first experiment, for the control group where no contextual information was given, the error rate was between 6-12.5% for four of the five cases and 21% for the other. When case context was provided, the error rate increased to 15-19% in the four cases. The greatest increase in error rate for this experiment went from 6% with no context to 19% with context that implied the presence of opioids. This case sample was considered a close case because the immunoassay test result was near the cut-off for additional testing, so it was thought that a close case may be more susceptible to bias from contextual information.
The second experiment showed that certain age groups were tested more often for particular drug types. Older people were tested for medicinal drugs at a much higher rate while younger people were more likely to be tested for drugs of abuse.
The study found that decision making, even in an ostensibly objective field of forensics, was affected by the presence of case information. Specifically, where the results were a close call, case information caused the error rates to increase based on the bias. In terms of case strategy, the participants’ personal biases and experiences led them to choose certain tests over others based on the given case information.
Limitations of this study include that the study participants were forensic toxicology students, which means their accuracy rates may differ from practitioners employed in a forensic lab. Also, even though the study was blind, it is possible that study participants may have discerned the purpose of the experiment and changed their answers accordingly.
They study’s authors proposed based on the results of this study that analysts making this type of decision in forensic toxicology laboratories not be given access to irrelevant contextual information. They also propose that forensic toxicology laboratories use a consistent framework for choosing tests, and that deviations be documented and justified in case files.