A comparison of estimators from self-controlled case series, case-crossover design, and sequence symmetry analysis for pharmacoepidemiological studies.

The estimations of the three self-controlled methods depended on various underlying assumptions, and the violation of these assumptions may cause non-negligible bias in the resulting estimates. Pharmacoepidemiologists should select the appropriate self-controlled method based on how well the relevant key assumptions are satisfied with respect to the available data.