Wednesday, May 21, 2014

"Annals of Internal Medicine on Syrian sarin"

My colleague Fred Lawson sent me this: "A rather odd article appears in the 4 May 2014 issue of the Annals of Internal Medicine, As'ad, in which eight Israeli doctors discuss the lessons that they learned from trying to analyze the 2013 nerve gas release outside Damascus on the basis of watching on-line videos.

Here is their methodology:  "We searched for videos uploaded between
21 August and 15 September 2013 on YouTube about the sarin attack in Syria.  We used the keywords 'Syria,' 'sarin,' 'nerve agents,' or 'August 21st' in English or Arabic.  ...We included videos that focused on at least 1 victim for at least 3 seconds, enabling the investigators to get a clinical imporession about predefined clinical signs."

Later, they remark that "one major disadvantage of the study is the inability to validate the authenticity of the videos.  Also, some deliberate selection of the uploaded material was probably done, which could bias our findings.  Therefore, it is preferred that we look on the following findings qualitatively rather than quantitatively."  The primary findings are that most of the victims were children, and almost all were severely injured in other ways.

It turned out to be hard to identify the signs of sarin gas, but with effort such indications were found:  "By watching the videos closely, one can notice the typical clinical signs of OP poisoning.  This well-defined toxidrome consists of hypersecretion and miosis, respiratory distress, convulsions and loss of consciousness.  Although the recognition of this toxidrome might be challenging, it was done successfully here."

The authors conclude that the spread of social media is likely to help us track future "toxicologic mass casualty events," even though such "videos are not medically oriented, and the medical information is incomplete in many cases.  We do not have data about the clinical events before or after the videos end.  Nevertheless, the abundance of material helps to validate the data in a way similar to 'crowd wisdom,' in which quantity and diversity compensate for quality."
Speaking only for myself, As'ad, I'm not sure that I want a crowd of amateurs to diagnose the next health problem that I suffer.  But perhaps modern medicine is heading that way."