Sir: It may be helpful for clinicians to appreciate that the great weight of recent evidence indicates that a spectrum model best explains serotonin syndrome phenomena. Serotonergic side-effects merge imperceptibly into ‘toxic’ effects or serotonin syndrome. Much confusion exists in the literature because in many reports an insufficiently precise distinction is being made between side-effects and toxicity.

At present the evidence is that life-threatening morbidity or mortality, only arises from combinations of monoamine oxidase inhibitors (this definitely does include so-called ‘RIMAs’ (reversible inhibitors of monoamine oxidase A) such as moclobemide) and drugs able to act as serotonin reuptake inhibitors (which includes some narcotic analgesics). The risk remains unclear for catechol-O-methyltransferase inhibitors.

I also wish to draw attention to some valuable prospectively gathered and systematically documented data specifically addressing the issues of what symptoms and signs characterise toxicity from various drugs when take in overdose. These data come from Ian Whyte's group. In a series of over 5000 cases of self-poisoning 10% were with a single, primarily serotonergic, drug. Of these, 16% met the Sternbach criteria for serotonin syndrome.

The only serotonin reuptake inhibitor that was significantly different from the reference drug (sertraline) in its frequency of association with the serotonin syndrome was clomipramine, with which serotonin syndrome was only one-tenth as frequent (odds ratio 0.1 and 95% CI was 0.0-0.9). This may be because clomipramine is a potent 5-HT_2A antagonist.

Our extensive database of references about serotonin syndrome is available to researchers at www.psychotropical.com.