Sir: We read with interest the drug information quarterly by Taylor (Psychiatric Bulletin, December 2000, 24, 465-468) in which the author concluded that “low doses of typical antipsychotics offer no advantages over higher doses”. This statement, if true, would be of major clinical importance as it refutes evidence that the apparent advantage of the new ‘atypical’ antipsychotics, in terms of tolerability, is largely owing to excessive doses of the typical drug being used in the clinical trials (Reference Geddes, Freemantle and HarrisonGeddes et al, 2000). This statement is, however, opinion rather than evidence-based.

Taylor makes no attempt to review this area systematically, instead relying on the selective citation of a few articles, of variable quality, that support his view. This practice is exemplified on page 466 of the article where, for reasons best known to himself, Taylor cites six isolated ‘clinical trials’ and neglects to mention the many randomised controlled trials (RCTs) and meta-analyses that cast doubt on the whole thrust of his argument. Taylor measures the tolerability of typical drugs using three indices, hyperprolactinaemia, tardive dyskinesia and placebo levels of extrapyramidal side-effects (EPS). In doing so he biases the article by failing to consider the wide range of potentially distressing side-effects that are associated with antipsychotic treatment. The adoption of hyperprolactinaemia as a proxy for tolerability is particularly confusing as it is accepted that only a proportion of people with it will experience an adverse event. Finally, the receptor binding studies cited in the article in support of Taylor's hypothesis do indeed show that high levels of D2 occupancy are needed for the therapeutic efficacy of typical drugs, but no mention is made of the fact that this may be responsible for early clinical relapse following the withdrawal of ‘atypical’ drugs (Reference Seeman and TallericoSeeman & Tallerico, 1999).

Taylor does, however, remind us that there is a trend for psychiatrists to use higher doses of typical drugs than can be justified on the basis of the available evidence. Whether such a finding supports Taylor's view or simply indicates that such drugs are used improperly is somewhat debateable. In citing one trial that compares an atypical to haloperidol he also reminds us of the high propensity of this drug to cause EPS even at low doses. To suggest that this fact in itself justifies atypicals as a first-line treatment is as ridiculous as it is to assume that EPS are the only consideration.

Data from well-conducted meta-analyses of RCTs do indeed confirm that ‘atypical drugs’ are less likely to cause EPS. However, when atypicals are compared to lower doses of ‘typicals’, efficacy is equal, the burden of total side-effects from both drugs is similar (Reference Kennedy, Song and HunterKennedy et al, 2000) and patients are no more likely to continue taking an atypical drug than an older one (Reference Geddes, Freemantle and HarrisonGeddes et al, 2000). Taylor's concluding remark flies in the face of almost 50 years of clinical trials and clinical experience in psychiatry. Such ‘drug misinformation’, if acted upon, would serve only to increase costs with no discernible benefit to patients. Why Taylor should wish for the advantages of atypicals over older drugs to be grossly overstated in this way is a matter for him to clarify.