“…For 36 studies, the risk of bias was determined as low (Supplementary Table 4, http://links.lww.com/LGT/A242). High risk of bias was in most studies caused by the unrepresentativeness of the cohort for the full population in a predefined period 11–13,24,27,28,30,32–34,36,37,40,42–44,48,50,51,53,57,59,61,71,73–77,79,83–88,91–93,95 and the low number of participants in the study 8,9,11,12,26,27,34,36,37,39,42–44,49,53,56,58,62,63,68,69,71–73,75,76,78,81,84,95 . The subgroup analysis for studies with low or high risk of bias showed no statistically significant differences between the 3 different outcomes for each CIN grade.…”