To the Editor: We read the systematic review of tardive dyskinesia associated with second-generation antipsychotics by Christoph U. Correll, M.D. and colleagues (1) with interest in view of its importance in clinical management. We evaluated this article based on the guidelines prescribed in the Cochrane Collaboration Handbook (2) and the Jadad et al. score (3).

There are certain methodological shortcomings in this systematic review. First, only the MEDLINE database was searched for studies in this area. A thorough search of other databases, including studies published in nonindexed journals, could have yielded more studies. Second, the studies were not assessed for their methodological qualities, and no definitive criteria were used that are usually recommended as an integral part of a systematic review (2, 3). Only three of the 11 studies included were randomized, double-blind, controlled trials; the rest were open-label studies. Even among these three, one was a study involving demented patients. Third, the studies included had used different modes of drug administration (injectable/oral), comprised patients of different age extremes, and included patients who had illnesses such as dementia and Parkinson’s disease—all of which have a bearing on the development of tardive dyskinesia.

The authors have acknowledged the limitations of the database with regard to the lack of studies with rigorous methodology on tardive dyskinesia. Hence, we feel that given the heterogeneity of the sample, it was inappropriate for the authors to have calculated the rates of tardive dyskinesia by combining results from the studies. Such calculation might result in a faulty assessment of the risk of tardive dyskinesia with second-generation antipsychotics.

These limitations make it difficult to interpret the results of the review. It is worth considering whether the authors could have written a narrative review instead of a systematic review of the available database.