Conundrums in neurology: diagnosing serotonin syndrome - a meta-analysis of cases

Background: Serotonin syndrome is a toxic state, caused by serotonin (5HT) excess in the central nervous system. Serotonin syndrome's main feature is neuro-muscular hyperexcitability, which in many cases is mild but in some cases can become life-threatening. The diagnosis of serotonin syndrome remains challenging since it can only be made on clinical grounds. Three diagnostic criteria systems, Sternbach, Radomski and Hunter classifications, are available. Here we test the validity of four assumptions that have become widely accepted: (1) The Hunter classification performs clinically better than the Sternbach and Radomski criteria; (2) in contrast to neuroleptic malignant syndrome, the onset of serotonin syndrome is usually rapid; (3) hyperthermia is a hallmark of severe serotonin syndrome; and (4) serotonin syndrome can readily be distinguished from neuroleptic malignant syndrome on clinical grounds and on the basis of medication history. Methods: Systematic review and meta-analysis of all cases of serotonin syndrome and toxicity published between 2004 and 2014, using PubMed and Web of Science. Results: Two of the four assumptions (1 and 2) are based on only one published study each and have not been independently validated. There is little agreement between current criteria systems for the diagnosis of serotonin syndrome. Although frequently thought to be the gold standard for the diagnosis of the serotonin syndrome, the Hunter criteria did not perform better than the Sternbach and Radomski criteria. Not all cases seem to be of rapid onset and only relatively few cases may present with hyperthermia. The 0 differential diagnosis between serotonin syndrome and neuroleptic malignant syndrome is not always clear-cut. Conclusions: Our findings challenge four commonly made assumptions about serotonin syndrome. We propose our meta-analysis of cases (MAC) method as a new way to systematically pool and interpret anecdotal but important clinical information concerning uncommon or emergent phenomena that cannot be captured in any other way but through case reports.