Influence of the subpubic arch angle on anal sphincter trauma and anal incontinence following childbirth

Authors

  • Andrea Frudinger,

    1. Intestinal Imaging Centre, Northwick Park and St Mark's Hospitals, Harrow, Middlesex, UK
    2. Physiology Unit, Northwick Park and St Mark's Hospitals, Harrow, Middlesex, UK
    3. Department of Obstetrics and Gynaecology, University Graz, Graz, Austria
    Search for more papers by this author
  • Steve Halligan,

    1. Intestinal Imaging Centre, Northwick Park and St Mark's Hospitals, Harrow, Middlesex, UK
    Search for more papers by this author
  • John A.D. Spencer,

    Corresponding author
    1. Department of Obstetrics and Gynaecology, Northwick Park and St Mark's Hospitals, Harrow, Middlesex, UK
      * Mr J. A. D. Spencer, Department of Obstetrics and Gynaecology, Northwick Park Hospital, Watford Road, Harrow, Middlesex, HA1 3AJ, UK.
    Search for more papers by this author
  • Clive I. Bartram,

    1. Intestinal Imaging Centre, Northwick Park and St Mark's Hospitals, Harrow, Middlesex, UK
    Search for more papers by this author
  • Michael A. Kamm,

    1. Physiology Unit, Northwick Park and St Mark's Hospitals, Harrow, Middlesex, UK
    Search for more papers by this author
  • Raimund Winter

    1. Department of Obstetrics and Gynaecology, University Graz, Graz, Austria
    Search for more papers by this author

* Mr J. A. D. Spencer, Department of Obstetrics and Gynaecology, Northwick Park Hospital, Watford Road, Harrow, Middlesex, HA1 3AJ, UK.

Abstract

Objective To assess the relationship between the subpubic arch angle, anal sphincter and perineal trauma, and anal incontinence after childbirth in nulliparous women.

Design Prospective cohort study.

Setting University teaching hospital.

Sample One hundred thirty-four low risk nulliparous women.

Methods Measurements of the bony pelvic outlet were made to calculate the subpubic arch angle. Endosonographic assessment of the perineum and anal sphincter was made before and after delivery. A specific bowel function questionnaire was completed before and after delivery.

Main outcome measures Perineal and anal sphincter trauma, and deteriorating anal continence after delivery.

Results Thirty-two women with a subpubic arch angle of less than 90° had significantly prolonged first and second stages of labour when compared with 102 women whose subpubic arch was wider. Following delivery, anal continence deteriorated in more women with a narrow subpubic angle (69%vs 21%, P < 0.001) but this was unrelated to the incidence of anal sphincter and perineal trauma.

Conclusions A narrow subpubic arch is strongly associated with prolonged labour and postpartum anal incontinence in nulliparous women. However, perineal and anal sphincter trauma, assessed by ultrasound, does not account for the higher rate of postpartum anal incontinence in women with a narrow subpubic arch angle.

Introduction

It is established obstetric teaching that a narrow pelvic outlet predisposes to a difficult vaginal delivery. A variety of simple clinical pelvimetry techniques can be used before delivery to assess adequacy of the pelvic outlet, notably the intertuberous diameter1. A rapid assessment of outlet adequacy may also be made by estimating the angle of the subpubic arch2. This is the angle subtended between the inferior border of the pubic symphysis and the ischial tuberosities on each side. The smaller this angle, the closer together are the ischial tuberosities and the narrower the pelvic outlet. It is believed this angle should be 90° or more if problems during delivery are to be avoided2. Furthermore, a narrow subpubic arch is more likely to displace the fetal head posteriorly, towards the soft tissues of the perineum and anal sphincter2.

Symptoms of anal incontinence are well recognised following vaginal delivery3 and are strongly associated with anal sphincter disruption4. Sphincter disruption is believed to occur in up to one-third of nulliparous women following vaginal delivery. Such trauma is best assessed using anal endosonography because clinical detection is unreliable4,5. There is an increased risk of sphincter disruption associated with forceps delivery but many cases occur following spontaneous vaginal delivery. We hypothesised that a narrow subpubic arch will predispose to perineal trauma and anal sphincter disruption. In addition, we wanted to see whether postpartum symptoms of anal incontinence occurred more frequently in women with a narrow subpubic arch, which we presumed would be the consequence of a greater posterior displacement of the baby's head at the time of delivery. In order to test this hypothesis, we prospectively characterised pre- and post-delivery bowel habit and performed anal endosonography on consecutive nulliparous women whose subpubic arch angle was measured pre-delivery.

Methods

The study was performed in one centre following approval from the local Clinical Research Ethics Committee. Consecutive nulliparous women were recruited during the third trimester from the antenatal clinic of a large university teaching hospital. All women were recruited by one investigator (AF), who explained the purpose of the study and obtained informed consent. Inclusion criteria were nulliparity, low risk singleton pregnancy, no apparent complication, no previous pelvic floor or anal surgery and ability to give informed consent. Approximately 5% of women invited to participate declined.

Pre-delivery bowel function was determined using a validated questionnaire6,7, which was completed by women under direct supervision. Measurements of the pelvic outlet were made in the supine position at the time of routine antenatal examination. The ischial tuberosities were palpated and the distance between their medial aspects measured using standard obstetric callipers (measured in 1-cm increments). Using the same instrument, the distance between the inferior border of the symphysis pubis in the midline and one ischial tuberosity was also measured. The subpubic arch angle was not calculated until analysis of the results after completion of all data collection. Thus, the angle was not known prior to delivery or at the time of postnatal follow up.

Anal endosonography was performed using a B and K Medical ADI 2001 Panther ultrasound scanner fitted with a type 1850 anal endoprobe and type 6004 10 MHz transducer (B and K Medical, Gentofte, Denmark). The transducer was covered with a sonolucent plastic cone that allowed withdrawal into the anal canal. Acoustic coupling was achieved by filling the cone with degassed water and then covering the probe with a condom lubricated with ultrasound gel on both surfaces. The cone had an external diameter of 17 mm. The transducer had a focal range of 5–45 mm, an axial resolution of less than 0.05 mm and a lateral resolution of 0.5–1 mm. With the woman in the left lateral position, the probe was gently inserted into the anus to the level of the anorectal junction. During slow withdrawal, images were taken at predefined anal canal levels (proximal, mid and distal) as per standard endosonographic practice8. The proximal anal canal level was identified by visualisation of the puborectalis sling. The mid canal level was identified when the external anal sphincter ring was seen to be complete anteriorly in combination with maximum internal anal sphincter thickness. The distal canal was defined as the level immediately caudal to the termination of the internal anal sphincter. Sphincter integrity was noted and the presence of any external and/or internal sphincter defects was recorded. Sphincter defects were defined by discontinuity of the external and/or internal sphincter ring4,5. If present, isolated perineal scarring was also noted. This was defined by any discontinuity or asymmetry of the anterior perineum and transverse perineal muscles, representing perineal trauma that did not directly involve the external anal sphincter9. Images were stored using a personal computer fitted with a picture acquisition system (Medimage; Vepro, MHS Medizintechnik, Vienna, Austria).

Following delivery, the same volunteers were asked to return to a postpartum pelvic floor clinic. The actual time of follow up ranged between three and eight months because some women could not keep the initial appointment. Data regarding gestation at delivery, the durations of the first and second stages of labour, need for episiotomy, mode of delivery and the baby's weight and head circumference were obtained from the clinic notes. All women were again interviewed by the same investigator and an identical bowel habit questionnaire was completed in order to identify any postpartum change in bowel habit. Women again underwent anal endosonography, exactly as before delivery, and the presence of any sphincter tears or perineal scarring was noted.

The subpubic arch angle was calculated after completion of the study from the pelvic outlet measurements using standard trigonometry. The space between the inferior pubic rami can be represented by two right angle triangles formed by a midline passing back from the midpoint of the pubic symphysis to the midpoint of a transverse line joining the ischial tuberosities. The subpubic arch angle is divided into two by this midline. The length of the inferior pubic ramus (IPR) represents the hypotenuse of each triangle. Half the intertuberosity (IT) distance is the length of the side opposite to half of the subpubic angle. The subpubic arch angle is therefore 2 × [sin−1 (1/2)(IT/IPR)]. Women with a subpubic arch angle of 90° or more constituted the ‘wide’ group for comparison with women whose subpubic arch angle was less than 90° (the ‘narrow’ group).

Discrete frequencies for categorical data were compared using Fisher's exact test. Continuous variables were compared using the t test, or Mann–Whitney test statistic if Shapiro–Wilk testing confirmed non-normality. Multivariate analysis using logistic regression was used to examine the effect of the subpubic arch angle on anal incontinence while accounting for the confounding effects of other variables, notably sonographic trauma. As only three women had vacuum extraction, these were combined for analysis with women having forceps extraction. Although incontinence was measured using a continuous scale, the large numbers of zero values necessitated collapse into two groups, continent or incontinent. Statistical significance was assigned to any probability level of less than 0.05.

Results

One hundred and thirty-four nulliparous women with full pre- and post-delivery data were included. The median age was 31 years (range 22–39). The mean subpubic arch angle was 104° (median 106°, range 60°–128°). The distribution of the population data was skewed towards larger angles, as shown in Fig. 1. There were 102 women in the wide group and 32 women in the narrow group. There was no endosonographic evidence of perineal scarring and/or sphincter tears in any subject pre-delivery.

Figure 1.

Distribution of subpubic arch angles.

Comparisons between the two groups are shown in Table 1. First and second stage durations were significantly longer in the narrow group but there were no significant differences in mode of delivery or episiotomy. Mean birthweight and mean head circumference of the babies were similar. A significantly greater proportion of women in the narrow group reported a deterioration in anal continence following delivery, but this was not accompanied by a significant difference in anal sphincter trauma seen by ultrasound.

Table 1.  Comparison between wide and narrow subpubic arch groups. Data are mean (95% CI), except second stage [median (interquartile range)], or n (%).
VariableWide group (n= 102)Narrow group (n= 32)P
Birthweight of baby (kg)3.47 (3.38–3.56)3.59 (3.44–3.75)0.19
Head circumference of baby (cm)35.03 (34.73–35.34)35.39 (34.83–35.95)0.26
First stage (hours)5.00 (2.78)6.57 (3.63)0.02
Second stage (hours)0.46 (0.18–1.16)1.34 (0.46–2.86)<0.001
Mode of delivery
Vaginal77 (75)22 (69)0.44
Caesarean12 (12)3 (9) 
Forceps13 (13)7 (22) 
Episiotomy
No51 (50)12 (38)0.23
Yes51 (50)20 (62) 
Ultrasound evidence of injury
None79 (77)20 (62)0.14
Perineal scar8 (8)6 (19) 
Sphincter tear15 (15)6 (19) 
Postnatal deterioration in anal continence
No81 (79)10 (31)<0.001
Yes21 (21)22 (69) 

Eighteen of the 102 women (18%) with a wide subpubic arch reported some degree of anal incontinence pre-delivery (median incontinence score 1, range 1–4) and continence worsened in six of these after delivery. A further 15 women in the wide group developed new symptoms of anal incontinence following delivery. Therefore, anal continence deteriorated following delivery in 21 women (21%) with a wide subpubic arch (median incontinence score 3, range 1–13). Seven of the 32 women (22%) with a narrow subpubic arch angle reported anal incontinence before delivery (median incontinence score 1, range 1–6) and this became worse after delivery in six. A further 16 developed symptoms of incontinence following delivery and so anal continence deteriorated in 22 women (69%) in the narrow group (median incontinence score 6, range 1–16). Thus, postpartum anal continence deteriorated significantly in women with a narrow subpubic arch (69% compared with 21%, Table 1. OR 8.49, 95% confidence interval (CI) 3.49–20.63, Table 2). Figure 2 shows that the subpubic angle distribution was towards the lower values in the group of women whose continence deteriorated, whereas continent women had an angle distribution heavily skewed towards the larger values.

Table 2.  Association between risk factors and anal incontinence.
VariableCategoryOdds ratio (95% CI)
Subpubic arch angleNormal1
Narrow8.49 (3.49–20.63)
Birthweight of baby (kg)2.16 (0.97–4.84)
Head circumference of baby (cm)1.19 (0.94–1.50)
First stage (hours)1.03 (0.91–1.16)
Second stage (hours)1.32 (0.96–1.83)
InjuryNone1
Perineal scar3.94 (1.25–12.48)
Sphincter tear2.69 (1.02–7.09)
Mode of deliveryVaginal1
Caesarean0.32 (0.07–1.51)
Forceps1.71 (0.65–4.55)
EpisiotomyNo1
Yes1.18 (0.57–2.45)
Figure 2.

Distribution of subpubic arch angles according to postnatal continence. The boxes show median with 25th and 75th centiles, and the whiskers represent the range. Note that in the group with no postnatal incontinence, the 25th centile and median values, and the 75th centile and upper limit values, respectively, coincide due to the skew towards larger values.

There was sonographic evidence of trauma in 12 (38%) women with a narrow subpubic arch (six isolated perineal scars and six anal sphincter tears) compared with 23 (23%) of the women in the wide group (8 with isolated perineal scars and 15 anal sphincter tears, Table 1). Sonographic evidence of anal sphincter injury was significantly associated with anal incontinence. Women with perineal scars and anal sphincter tears reported more incontinence than women with no sonographic sign of injury (Table 2). Multivariate analysis of variables likely to have an effect on anal incontinence (P < 0.2) showed that a highly significant relationship between the subpubic arch angle and anal incontinence remained (OR 8.69, 95% CI 3.15–23.94, Table 3). The size of this effect was similar to the individual analysis which suggests that the additional variables, including sonographic evidence of trauma, had little influence on anal incontinence due to a narrow subpubic arch angle.

Table 3.  Adjusted odds ratios for anal incontinence.
VariableCategoryOdds ratio (95% CI)
Subpubic arch angleNormal1
Narrow8.69 (3.15–23.94)
Birthweight of baby (kg)1.31 (0.34–4.99)
Head circumference of baby (cm)1.11 (0.74–1.67)
Second stage (hours)0.89 (0.57–1.39)
Ultrasonic evidence of injuryNone1
Perineal scar2.69 (0.70–10.33)
Sphincter tear2.26 (0.69–7.36)
Mode of deliveryVaginal1
Caesarean0.31 (0.06–1.62)
Forceps1.36 (0.35–5.29)

Women with sonographic evidence of anal sphincter trauma or perineal scars were combined to study the relationship with other variables (Table 4). Thirty-eight percent of women with a narrow subpubic arch had sonographic evidence of trauma compared with 23% of those with a wide arch (Table 4). When examined individually, there was no significant effect of the subpubic arch angle on the incidence of sonographic trauma (OR 2.07, 95% CI 0.88–4.84, Table 4). In contrast, birthweight and episiotomy were significantly associated with sonographic trauma. An increase in birthweight of the baby of 1 kg more than doubled the odds of injury, while those who require an episiotomy had odds of injury four times greater than those who did not. There was also slight evidence of an effect of forceps delivery method on the presence of sonographic trauma (OR 2.42, 95% CI 0.90–6.52, Table 4).When all variables showing evidence of an effect upon sonographic trauma (P < 0.2) were entered into a multivariate model, there was no significant relationship between the subpubic arch angle and the presence or absence of sonographic trauma (OR 1.66, 95% CI 0.66–4.17, Table 5). Thus, after adjustment, the relationship between sonographic trauma and subpubic arch angle was even less strong than that suggested by individual analysis.

Table 4.  Association between risk factors and sonographic evidence of anal sphincter and perineal trauma.
VariableCategoryOdds ratio (95% CI)
Subpubic arch angleNormal1
Narrow2.07 (0.88–4.84)
Birthweight of baby (kg)2.60 (1.10–6.14)
Head circumference of baby (cm)0.98 (0.76–1.26)
First stage (hours)0.97 (0.85–1.10)
Second stage (hours)1.20 (0.85–1.67)
Mode of deliveryVaginal1
Caesarean0.21 (0.03–1.69)
Forceps2.42 (0.90–6.52)
EpisiotomyNo1
Yes4.22 (1.74–10.20)
Table 5.  Adjusted odds ratios for sonographic evidence of anal sphincter and perineal trauma.
VariableCategoryOdds ratio (95% CI)
Subpubic arch angleNormal1
Narrow1.66 (0.66–4.17)
Birthweight of baby (kg)2.25 (0.93–5.45)
Mode of deliveryVaginal1
Caesarean0.38 (0.04–3.43)
Forceps1.37 (0.47–4.06)
EpisiotomyNo1
Yes2.94 (1.08–8.03)

Discussion

Significantly more women with a narrow subpubic arch angle reported a deterioration in anal continence after delivery (69%vs 21%, OR 8.5, 95% CI 3.5–20.6), confirming our hypothesis. As we expected anal sphincter disruption to account for this increase in anal incontinence, we used anal endosonography in all women because most sphincter tears following vaginal delivery are clinically occult. Surprisingly, multivariate analysis showed that the marked difference in postpartum anal incontinence between the two groups could not be attributed to anal sphincter and perineal trauma alone.

A retrospective study of 8603 vaginal deliveries found overt third degree tears in only 50 women (0.6%)10, yet when the same authors used endosonography in a consecutive group of 202 deliveries, there was evidence of third degree tears in 35% of primiparous and 44% of multiparous women4, a startling incidence subsequently confirmed by other researchers5. More recent studies using higher frequency transducers confirm that approximately one-third of women have evidence of anoperineal trauma, but suggest this directly involves the external sphincter in only 11%9. Our data show a previously unreported relationship between pelvic anatomic shape and subsequent incontinence, and contribute to an understanding as to why anal incontinence is not directly related to sphincter trauma.

Only 7 of 23 women with a wide arch and sonographic trauma complained of post-delivery anal incontinence in contrast to 11 of 12 women with a narrow subpubic arch and sonographic trauma. Thus, 70% of women with sonographic trauma in the wide subpubic arch group remained asymptomatic. Although it is well recognised that anal sphincter disruption may not cause immediate post-delivery symptoms, instead only becoming apparent after the cumulative effects of further deliveries, ageing and menopause, it is difficult to explain why sonographic trauma was only associated with symptoms in women with a narrow subpubic arch. We had hypothesised that differences in continence could be explained by a different pattern of anal sphincter and perineal trauma. We expected women with a narrow subpubic arch to experience more external sphincter disruption, whereas those with a wide arch might instead have isolated perineal tears, which are less likely to give rise to symptoms9. However, we found no evidence to support this; the distribution of isolated perineal scars and external sphincter disruption was similar between the two groups (8 isolated perineal scars and 15 tears in the wide arch group compared with six of each in the narrow arch group).

We found no significant differences between the two groups with respect to episiotomy rate, forceps or vacuum extraction or Caesarean section. However, the first and second stages of labour were both significantly longer in women with a narrow subpubic arch angle. Posterior displacement of the fetal head as it passes through the outlet probably results in a need to overcome more soft tissue resistance. Our findings showed that this did not result in a significantly more structural anal sphincter or perineal damage. However, stretching of the pelvic floor nervous innervation, particularly in relation to control of anal function, may have been adversely influenced to a greater extent in this group. It has long been recognised that pudendal nerve latency is prolonged following childbirth11 and this was believed to be the prime cause of postpartum anal incontinence until anal endosonography revealed the true incidence of anal sphincter trauma4. However, our results suggest that structural trauma alone does not account for postpartum anal incontinence in women with a narrow subpubic arch. A recent study of 184 primiparous women found that nearly half had abnormal anorectal physiology postpartum12. The authors also found a prolonged second stage increased the risk of postpartum incontinence and found prolongation related to sonographic anal sphincter damage12. A relationship with the subpubic arch angle was not sought. Pelvic floor innervation will need to be compared in women with wide and narrow subpubic arch angles in order to determine if this is the mechanism responsible for deteriorating continence.

Our finding that 19% of women experienced some degree of anal incontinence, as defined by our questionnaire, before delivery was surprising. All were nulliparous and there was no history of anal surgery. This may be partly explained by the highly sensitive and specific nature of the questionnaire but also stresses the importance of vigorous pre-delivery characterisation of bowel habit. We were able to confidently identify women whose anal continence deteriorated following delivery. Pre-delivery incontinence scores were mild, with a median of one, and there was no significant difference between the two groups studied.

In summary, we have found that a narrow subpubic arch is strongly associated with prolongation of both the first and second stages of labour, and postpartum anal incontinence. However, endosonography indicated that a deterioration in anal continence after delivery was not due to a higher incidence of anal sphincter or perineal trauma.

Acknowledgements

We acknowledge the assistance of Paul Bassett for statistical analysis of the data.

Ancillary