The diagnostic use of the diatom test for drowning has been under investigation for more than a century. Despite continuing research, its true usefulness remains controversial and under debate. Data regarding the extent to which diatoms can penetrate the lungs and other organs of drowning victims are conflicting
similar discrepancies exist as to the presence of diatoms in the organs of living individuals
and as to the occurrence of postmortem (PM) contamination. To shed light on current understanding, we conducted a systematic review based on the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) to investigate how the presence of diatoms PM may be interpreted during medico-legal investigations of drowning. Following sequential screening of records based on our predetermined eligibility criteria, we assessed scientific evidence and risk of bias by use of the SPICOT framework. A total of 17 studies reporting diatom concentrations in victims of drowning, in non-drowned controls, and in non-drowned immersed controls were eligible for this review. Our findings suggest that diatom testing may be of use in medico-legal investigations, although its evidentiary value remains uncertain because both quantitative and qualitative results from the literature are insufficiently comparable. Variations in study design, methodology and reporting approach also fail to provide a comprehensive understanding of the significance of false-positive and false-negative results. Further research is warranted on antemortem and PM contamination, and on standardized autopsy and laboratory procedures, as well as on automated and certified diatom-counting and -identification systems. Moreover, since diatom taxonomy lies outside the specialty of forensic medicine, we underscore that collaboration with expert diatomologists is necessary for analysis and interpretation. Until these issues are adequately addressed, the evidentiary value of diatom testing for the diagnosis of drowning will continue to remain elusive and contentious.