Authors' response to: Prevalence of urinary incontinence (UI) 20 years after childbirth in a national cohort study in singleton primiparae after vaginal or caesarean delivery


Authors' Reply


We thank John et al.[1] for their interest in our paper[2] published recently in BJOG. The authors have chosen to comment on the title of our paper, which they have unfortunately incorrectly stated in the heading of their Letter to the Editor. The title reads ‘The prevalence of urinary incontinence 20 years after childbirth: a national cohort study in singleton primiparae after vaginal or caesarean delivery'. The title follows the recommendation from BJOG's Instructions for Authors that the title should include the methodology at the end of the title after a colon.

John et al. comment on the relationship between urinary incontinence (UI) and maternal age and current body mass index. As all women were followed for approximately 20 years (the mean follow-up was 21.5 years [SD 1.5] in the vaginal delivery group and 21.8 years [SD 1.1] in the caesarean section group) use of current age or age at delivery will give the same result for prevalence data in the analysis. Further in the logistic regression analysis, a higher maternal age was associated with an increased risk of UI (odds ratio 1.03; 95% confidence interval 1.02–1.04), which corresponds to an annual risk increase of 3%.

Concerning the question raised by John et al. regarding the method of obtaining current height and weight: As stated in the Methods section of the paper ‘The 31-item questionnaire included questions about current height and weight'. Self-administered postal questionnaires ‘offer advantages’ because they ‘may yield more accurate data on sensitive or embarrassing topics because they are more anonymous' compared with telephone interviewer-administered instruments.[3] Any individual difference between actual weight and reported weight was assumed to be evenly distributed among the women of these large cohorts and will therefore not alter the calculated difference in UI prevalence between the caesarean section and vaginal delivery groups.

John et al. have noted that we assessed the severity of UI with the Sandvik[4] score but did not present data on this assessment. It is correct that the Sandvik score was mentioned in the Methods section. The information generated about severity of UI is being presented in a separate paper submitted for publication describing the prevalence, severity and bothersomeness associated with the three main subtypes of UI (stress, urge and mixed) 20 years after vaginal delivery or caesarean section.

They also comment on tables 4–6 where we have performed analyses regarding risk factors for UI. We have noted that John et al. would have preferred a further analysis where confounding variables were compared with the independent variable to understand the net effect of the independent variable.

We agree with John et al. that including birthweight and head circumference in tables 5 and 6 did not offer any further information about nonlinear or threshold effects. Still, we think that this (negative) information is of interest, not least clinically. The different categorisations of infant birthweight were made in an attempt to find possible threshold effects for these parameters (which has been done in many earlier studies in this field) and to obtain sufficient power in the analysis, especially for the subgroup analysis of the two different modes of delivery.

John et al. misinterpret our use of the term NNT (number-needed-to-treat). The term NNT is purely a statistical measure calculated from the difference of prevalence between the control group (caesarean section) and the exposed group (vaginal delivery). It is widely used in the literature to compare results between studies in epidemiology. Our paper contains no recommendation about mode of delivery and it was further noted that ‘operative delivery by caesarean section also involves a degree of risk for morbidity and mortality over and above that of vaginal delivery'.