Park et al (Reference Park, Ross-Degnan and Adams2005) cite the results of clinical trials as evidence supporting their hypothesis that the use of antiparkinsonian drugs in schizophrenia is an indication of extrapyramidal symptoms (EPS). This may be true for clinical trials (most of which include young adults with no comorbidity) but may not hold true for their observational study, in which other factors such as prescribing habits and comorbidity may affect the reason for prescription of antiparkinsonian drugs. As the mean age of their sample was 48.6 years, which falls within the range in which Parkinson's disease often develops, some patients could have been receiving antiparkinsonian drugs for the illness per se. Although this is mentioned as a limitation of the study, it has an adverse impact on the central hypothesis. Since decrements and increments in antiparkinsonian medication followed expectations from changes in antipsychotics (Reference Tran, Hamilton and KuntzTran et al, 1997), the results could well reflect the prescribing pattern of the general practitioners (GPs) rather than be true evidence for the presence of EPS.

One of the main limitations of the study is the lack of data regarding the reason for switching antipsychotics. As it is mandatory to submit data of all major illnesses (presumably including Parkinson's disease), any indication for prescribing or altering medication and any adverse drug reaction to the General Practice Research Database (GPRD; Reference Walley and MantganiWalley & Mantgani, 1997), the data could have been provided and would have helped in the interpretation of the results. Furthermore, during the period studied more than 400 GPs provided data to GPRD but data from only 266 were analysed. It is not clear why the data from some GPs were excluded.

Park et al (Reference Park, Ross-Degnan and Adams2005) classified their study population as those switched from typical to atypical antipsychotics (TA group) and those switched from typical to different typical antipsychotics (TT group). However, when we add up the total figures provided (3% and 99% were receiving atypicals and typicals respectively in 1992, which changed to 47% and 70% in 2000), it appears that some patients were receiving a combination of both classes of antipsychotics. This could have influenced the trend for prescribing antiparkinsonian drugs.