This article corrects:

  1. Does the mode of delivery predispose women to anal incontinence in the first year postpartum? A comparative systematic review Volume 115, Issue 4, 421–434, Article first published online: 6 February 2008

It has been noted since the publication of ‘Does the mode of delivery predispose women to anal incontinence in the first year postpartum’ (BJOG 2008, March 115(4):421–34) that data from Dr C MacArthur’s paper, ‘Obstetric practice and faecal incontinence 3 months after delivery’ (BJOG 2001, July 108(7):678–683) had been misinterpreted and entered into the systematic review as referring to incontinence of faeces and flatus when the paper actually refers to faecal incontinence alone.

When this error is corrected, the broad sweep of the paper, that mode of delivery does affect symptoms of anal incontinence in the first year postpartum is unchanged. However, the type of symptoms experienced by women is changed. These changes are demonstrated graphically in the re-analysed forest plot (Figure 1, formerly Figure 3 in original paper).

Figure 1.

 The effect of mode of delivery on symptoms of anal incontinence (the constituent studies are detailed in tables 4 and 5).

When caesarean section is used as the control group for symptoms of faecal incontinence alone, statistical significance is now reached in all groups, with forceps, ventouse and spontaneous vaginal delivery all having increased risk of symptoms when compared to caesarean section.

When vaginal delivery is used as the control group for faecal incontinence alone, forceps and ‘any instrumental’ have significantly more symptoms than spontaneous vaginal delivery. Forceps do not cause significantly more symptoms when compared to ventouse for faecal incontinence alone.

When examining any symptoms of anal incontinence (i.e. solid and or liquid and or flatus), there is now only one study examining the difference between caesarean section and vaginal delivery which is too small to reach statistical significance and draw any valid conclusions from.

Interestingly when any symptoms of anal incontinence are studied using spontaneous vaginal delivery as the control group, women delivered by ventouse or ‘any instrumental’ had statistically significantly more symptoms than the control group but the group delivered by forceps were not statistically more likely to have symptoms.

When Fitzpatrick’s study is added to look at the difference between forceps and ventouse, forceps again appear to cause more symptoms.

The authors apologise for these errors.