The concerns discussed in The Psychiatrist ^{Reference Rajpal, Owens and McCauley1} about the causes, complications, management and avoidance of antipsychotic polypharmacy reflect an earlier literature on anti-epileptic polypharmacy. ^{Reference Shorvon and Reynolds2-Reference Reynolds4}

Anti-epileptic polypharmacy is much older than antipsychotic polypharmacy and can be traced at least to the introduction of bromides in the middle of the 19th century. By 1970, patients at epilepsy centres in the UK and Europe were taking a mean of 3.2 drugs per person, of which 84% were anti-epileptics. In the 1960s and 1970s the availability of blood level monitoring of phenobarbitone, phenytoin, carbamazepine and valproate revealed that most of this polypharmacy was unnecessary, unhelpful and in many patients harmful. As well as previously unsuspected subtle toxic effects on cognition, mood and behaviour, polypharmacy also increased seizure frequency in some patients. In newly diagnosed patients, long-term seizure control was achieved in 70-80% with carefully monitored monotherapy. The addition of a second drug to those patients uncontrolled by optimum monotherapy resulted in modest improvement in seizure control in only a third, but led to unacceptable toxicity or increased seizures in a quarter. Monotherapy became the gold standard, licensing authorities demanded clinical trials of monotherapy and the incidence of polypharmacy fell rapidly.

Between 1989 and 2004, ten new anti-epileptic drugs were licensed in the UK, only one of which, lamotrigine, has widely available blood level monitoring. The lessons of the 1970s and 1980s have been largely overlooked and polypharmacy has increased again. Even now there has been only one controlled trial in patients with epilepsy unresponsive to optimum monotherapy comparing the policy of adding a second drug with the policy of switching to alternative monotherapy. The probability of remaining seizure-free over the next year was 16% for patients given a second drug and 14% for those switched to alternative monotherapy. ^{Reference Beghi, Gatti, Tonini, Ben-Menachem, Chadwick and Nikanorova5}

For both psychosis and epilepsy, factors which perpetuate polypharmacy are: a lack of understanding of the therapeutic limits of monotherapy or polypharmacy with either class of drug; unawareness of subtle and sometimes serious chronic toxicity; the need to do something in a distressing and chronic situation; misguided claims about the ‘rationality’ of polypharmacy, based on theoretical views of different mechanisms of drug action or even unproven concepts of synergy.

The problems of monitoring medication and measuring clinical outcome in psychosis are different from epilepsy, but many of the causes of unnecessary and potentially harmful polypharmacy as well as the principles of avoiding such polypharmacy, are likely to be similar.