On the strength of evidence in forensic pathology

Excerpt

One of the key issues in forensic pathology is the strength of evidence required to draw conclusions about the cause of death [1]. This is a thorny and controversial issue for, at least, three reasons. First, the legal system often tries to comprehend the strength of a pathologist’s opinion relative to the legal standards of balance of probabilities and beyond a reasonable doubt. But, it is unusual for a pathologist to frame their opinions based on a legal standard. Most pathologists make determinations based on being satisfied that there is adequate evidence to support their conclusion in a given case. Second, it seems that there is a tacit assumption among some pathologists that the severity of the legal outcome ought to be linked to a higher medical evidential standard for an opinion. For example, the standard for the diagnosis of ischemic heart disease in a natural death is probably lower than the standard applied to an opinion about an asphyxial death in a murder trial. Third, differences of opinion in difficult areas of forensic pathology are most often related to how strongly different pathologist’s hold their opinion on a controversial topic. For example, the line drawn by two different pathologists between the determination of stillbirth or live birth may relate to a qualitative difference on ‘how much lung inflation is enough’ to establish live birth, rather than if lung inflation is a valid criterion for live birth [2]. This gives rise to a natural and acceptable inter-observer variability in forensic pathology and often explains seemingly wide differences of opinion. This concept has been well studied in surgical pathology and has challenged our thinking about what constitutes an error compared to a difference of opinion. Some surgical pathology diagnoses, due to the nature of the disease or the morphology of the entity, have a higher or lower rate of diagnostic concordance than other diagnoses (higher or lower kappa scores). For example, in breast cancer, the highest agreement is observed with tubular and mucinous carcinoma which has the most uniformly and easily recognized histologic pattern [3]. Indeed, in medicine, some diagnoses are less difficult and attain more general agreement than other diagnoses that may be more difficult or have less clear diagnostic criteria. Forensic pathology is no exception to other branches of medicine. …