Hagen S1, Sellers C1, Elders A1, Macarthur C2, Hemming C3, Toozs-Hobson P4, McDonald A5, Herbison P6, Wilson D7, Glazener C5

Research Type


Abstract Category

Pelvic Organ Prolapse

Best Clinical Abstract
Abstract 19
Live Urogynaecology, Female & Functional Urology - Childbirth and its Consequences
Scientific Podium Session 3
Friday 15th October 2021
07:30 - 07:40
Live Room 1
Prospective Study Questionnaire Prolapse Symptoms Incontinence Bowel Evacuation Dysfunction
1. Glasgow Caledonian University, 2. University of Birmingham, 3. Aberdeen Royal Infirmary, 4. Birmingham Women's NHS Foundation Trust, 5. University of Aberdeen, 6. University of Otago, 7. Dunedin School of Medicine

Suzanne Hagen



Hypothesis / aims of study
Pelvic floor dysfunction (PFD) is a common problem both during pregnancy and after childbirth. The prevalence of postnatal PFD (urinary incontinence, faecal incontinence, pelvic organ prolapse) varies in the literature according to definition, method of data collection and follow-up duration [1]. Few studies have followed a large cohort to monitor the prevalence over time. In our longitudinal study, a population of 7,879 primiparous and multiparous women were originally recruited three months after giving birth in a maternity unit in three countries (England, Scotland and New Zealand). These deliveries are referred to as index births. We have continued to follow up our birth cohort; 4,214 women responded at 6 years after index birth, and 3,763 at 12 years.

We present here the analysis of data collected in the UK from this multicentre study on the prevalence of PFD 26 years after index birth according to delivery mode history.
Study design, materials and methods
Data were collected from women by postal questionnaire. Approval was obtained in order to access health service records to ensure that recipients were still living and resident in the UK, and that questionnaires were sent to current addresses. Participants were excluded if they had indicated at a previous follow-up that they did not want to be contacted again. 
For continuity, the core questions asked were the same as at previous follow-up time-points. Women were asked to think about their symptoms over the last 4 weeks. The questions to ascertain the prevalence of urinary incontinence (UI) were “At present do you ever lose any urine when you don’t mean to?” (yes/no) combined with “How much urine do you usually leak (whether you wear protection or not)?” (“none” to “a large amount”). The question to ascertain the prevalence of faecal incontinence (FI) was “Do you ever lose control of bowel motions (stool / faeces) from your back passage in between visits to the toilet?” (“never” to “all of the time”). To ascertain the prevalence of symptomatic prolapse women were asked about “a feeling of something coming down from or in your vagina?” (“never” to “all of the time”), as part of the Pelvic Organ Prolapse Symptom Score (POP-SS) (7 items, range 0 – 28, higher scores indicate more severe symptoms). 
Respondent characteristics were tabulated, observed prevalence rates (with 95% confidence intervals) were reported for UI, FI and something coming down and means and standard deviations were reported for POP-SS. Outcome data were presented by delivery mode history: spontaneous vaginal delivery only (Only SVD); caesarean section delivery only (Only CS); one or more deliveries involving forceps (Any forceps); one or more deliveries using vacuum extraction, but no forceps (Any vacuum (no forceps)); a mix of spontaneous vaginal or caesarean section deliveries only (SVD + CS). Associations with delivery mode history were examined using chi-squared tests for the prevalence of UI, FI and something coming down, and one-way analysis of variance for POP-SS.  Statistical significance was at the 5% level.  Analysis was conducted using Stata v14.
A total of 5,285 study invitations were sent out and 1,845 completed questionnaires returned (response rate 35%). Median length of follow-up was 25.9 years (range 24.8, 26.9). Baseline respondent characteristics are provided in Table 1. The mean age of respondents was 56 years (SD 5), with range 41 to 73 years. Age at first birth ranged from 15 to 44 years, with a mean of 28 years.  

Self-reported symptoms are summarised in Table 2. Overall 60.7% of women reported at least some level of UI in the previous 4 weeks, 21.6% reported FI and 16.8% reported a feeling of something of coming down. The mean POP-SS score was 3.2 (SD 4.0). 

A significant difference (p=0.01) was found between the different delivery history mode categories for UI, with the prevalence ranging from 45.5% for the “Only CS” category up to 67.9% for “Any vacuum (no forceps)”. Rates of FI ranged between 17.0% for the “Only CS” category and 25.7% for “Any forceps”, however no significant difference was found (p=0.12). The lowest prevalence of something coming down was reported for the “Only CS” category (6.1%) while the highest was reported for “Any vacuum (no forceps)” (20.7%), and a significant difference (p=0.01) was found between the delivery categories. POP-SS ranged from a mean of 2.4 (SD 3.3) for “Only CS” up to 3.5 (SD 4.2) for “Any forceps”, but no significant difference was found (p=0.12). 

The percentage of women reporting UI, FI or something of coming down was lowest for those who had delivered only by caesarean section compared to the other delivery mode history categories. These women also had the lowest mean POP-SS score. UI (67.9%) and something coming down (20.7%) were most prevalent in the women in the vacuum extraction (but no forceps) category, whilst FI (25.7%) was most common in women who had had a forceps delivery (one or more).
Interpretation of results
The unadjusted statistical tests showed a significant difference according to delivery mode history for the UI and something coming down measures, but not for FI or POP-SS. 

The overall prevalence rates for UI and FI reported here are higher for all delivery mode history categories than those reported in the full 12-year follow-up (60.7% vs 52.7% and 21.6% vs 12.9% respectively). For UI the greatest difference was for “Any forceps” (61.2% vs 51.4% at 12 years) and “Any vacuum (no forceps)” (67.9% vs 56% at 12 years). These categories also displayed the largest differences in prevalence of FI: “Any forceps”, 25.7% vs 16.7% at 12 years and “Any vacuum (no forceps)”, 24.3% vs 10.9% at 12 years. There was little change in the prevalence of something coming down (16.8% vs 17.0% at 12 years), and only a slight difference in mean POP-SS (3.2 SD 4.0 vs 2.7 SD 3.5 at 12 years) with no obvious differences between the delivery mode history categories.
Concluding message
This unique longitudinal follow-up of women after childbirth has shown higher prevalence of pelvic floor dysfunction at 26 years following index birth than at 12 years. Differences were observed according to delivery mode history. While there is an indication that a delivery history of exclusive caesarean sections may be associated with a reduced risk of experiencing symptoms of pelvic floor dysfunction, the prevalence of UI remains high with almost half of women reporting this symptom. The findings will be further informed by a full longitudinal statistical modelling analysis of these data, adjusting for demographic characteristics, inclusion of data from the New Zealand cohort members and an analysis of responder and non-responder characteristics.
Figure 1 Table 1. Respondent characteristics as reported at 26 year follow up
Figure 2 Table 2. Self-reported symptoms at 26 years by delivery mode history
  1. Abrams, P, Cardozo, L, Wagg, A, Wein, A. Incontinence 6th Edition (2017). ICI-ICS. International Continence Society, Bristol UK, ISBN: 978-0956960733.
Funding Chief Scientist Office, Scottish Government Clinical Trial No Subjects Human Ethics Committee South East Scotland Research Ethics Committee 1 Helsinki Yes Informed Consent Yes
17/04/2024 17:11:43