2016 年 42 巻 2 号 p. 87-97
Atypical antipsychotics are less likely to cause extrapyramidal symptoms (EPS) compared to typical antipsychotics. However, recent studies have suggested that both atypicals and typicals produce a similar risk for EPS. We performed data mining using the Japanese national insurance claims database (NDB) constructed by the Ministry of Health, Labour and Welfare in Japan. Sequence symmetry analysis was performed to identify the risk of EPS after antipsychotic use. In this study, antiparkinsonian drugs with anticholinergic action were used as a marker of EPS. Antipsychotics in combination with antiparkinsonians was examined by prescription sequence symmetry analysis (PSSA). Likewise, event sequence symmetry analysis (ESSA) was undertaken to evaluate the association between antipsychotics and EPS diagnosis. Adjusted sequence ratios (ASRs) with 95% confidence intervals (CI) were calculated. In PSSA, significant associations with antiparkinsonians were found for the whole class of antipsychotics and atypicals with ASRs of 4.82 (95%CI 4.74-4.91) and 2.97 (2.93-3.01), respectively. A significant association between typicals and antiparkinsonians was not found. In ESSA, significant associations with EPS were found for the whole class of antipsychotics and atypicals with ASRs of 1.67 (1.65-1.69) and 1.49 (1.47-1.50), respectively. A significant association between typicals and EPS was not found. The number of patients who were prescribed atypicals first was 1.5 times larger than the number of patients who had typicals initially prescribed. Analysis of NDB demonstrated that antipsychotics increase the risk of EPS. Significant associations between EPS and atypicals (but not typicals) were found. This finding may be attributed to the prescribing sequence of antipsychotics.