Data extraction
Data were extracted by the three authors and synthesised into summary tables presenting information about study characteristics, methods and outcomes of interest. For the part covering perceptions of teratogenic risk, the authors aimed to examine whether the results of the included studies indicated a true (proper) estimation of the teratogenic risk or not. If not, the perceived estimations of the teratogenic risk were examined to see if they were higher than the true risk (over-estimated) or lower than the true risk (under-estimated). Therefore, the numeric value of the perceived teratogenic risk extracted from the results section of each paper was compared to the true value of the teratogenic risk that was extracted either from the methods or the results section. The authors followed the method shown in Figure 1 to assign the perceived teratogenic risk of every medication included in the review into one of three categories: properly-estimated, over-estimated, or under-estimated.