The diagnostic value of the diatom test for drowning, II. Validity: analysis of diatoms in bone marrow and drowning medium.

In part two of this series, we investigated the characteristics of diatom frustules recovered from bone marrow and samples of putative drowning medium in case of freshwater drownings. A total of 52 cases of freshwater drowning with diatoms in the femoral bone marrow in which a sample of the putative drowning medium was also collected were available for analysis. The same types of diatoms were found in the bone marrow and putative drowning medium in 47 cases (90%) indicating that the water samples were representative of the site of drowning in at least 90% of freshwater drownings. In the remaining 5 cases (10%), the diatoms in the water samples did not match those in the bone marrow indicating that the site of body and water sample recovery were likely geographically remote from the site of drowning. In cases with matching diatoms in the bone marrow and drowning medium, the diatoms were consistently the smallest and most abundant types represented in the water samples. In addition, there were highly stereotyped physical characteristics of typical "drowning-associated" diatoms with small pennate diatoms representing the most common type of frustule found in the bone marrow. In an additional 34 cases of putative drowning in water that lacked detectable diatoms, 29 cases (85%) lacked diatoms in the bone marrow. Analysis of the diatom content of samples of putative drowning medium by month revealed that winter months had the highest frequency of samples devoid of diatoms. These data indicate that the true positive rate of the diatom test for drowning is at least 90% and that small pennate frustules are most commonly associated with drowning, particular in non winter months.