This review answered a clear research question. The authors searched for studies in a number of databases, although the search dates were not reported and it was unclear exactly what journals were included in the search. Publication bias was assessed, but the reliability of this assessment is unclear given the small number of studies included in the review; however, some attempts were made to retrieve unpublished material through the searching of conference abstracts. Lack of detail about the review methodology, including details of the validity assessment and whether data were extracted in duplicate, make it difficult to assess and confirm the reliability of the authors' findings.
The review findings were based on data from a small number of case-control studies, which are likely to be affected by recall bias. The analysis was also limited by a lack of detail about the original studies, e.g. the drugs involved and which confounding factors had been included in the adjusted analyses. The authors acknowledged that the level of drug exposure was often not reported, thus the results were restricted to exposure versus no exposure; this makes it difficult to assess dose-related effects.
Statistical heterogeneity was considered in the analysis, but it is unclear how clinically similar the studies were with respect to the pooling of data. Finally, the generalisability of the findings to other populations may be limited since the majority of the participants were older women living in the USA.
Overall, the reliability of the review's findings is unclear given the poor reporting of review methods and the potential methodological weaknesses of the study data.