To examine the incidence of obstetric anal sphincter injury in women who had a successful vaginal birth after a previous caesarean delivery (VBAC).
To examine the incidence of obstetric anal sphincter injury in women who had a successful vaginal birth after a previous caesarean delivery (VBAC).
Retrospective analysis of prospectively gathered data.
A tertiary referral university institution.
All secundiparous women with a previous caesarean delivery who had a VBAC from 2001 to 2011.
Details of maternal demographics, intrapartum characteristics and outcomes were examined in cases of VBAC with accompanying anal sphincter injury.
Rates of obstetric anal sphincter injury and associated risk factors.
During the study period there were 3071 trials of labour in secundiparous women with a previous caesarean delivery; 65% (1981/3071) of these had a successful VBAC. Women having a VBAC were at greater risk of anal sphincter injury than nulliparous women having a vaginal delivery over the same period (5% [98/1981] versus 3.5% [1216/34 496], P = 0.001, odds ratio 1.4, 95% CI 1.15–1.75). The rate of instrumental delivery in woman having a VBAC was 39% (771/1981). On multiple logistic regression analysis an increased rate of instrumental delivery was a strong predictor of sphincter injury (P = 0.03, odds ratio 1.15, 95% CI 1.01–1.3). When the first labours of women with sphincter injury in the VBAC group were examined, 70% (60/86) had been in labour before undergoing their caesarean delivery.
The incidence of anal sphincter injury in women undergoing VBAC is 5% and birthweight is the strongest predictor of this. The rate of instrumental delivery in this group was also increased.
Injury to the anal sphincter mechanism (third-degree tear) occurs after extension of a perineal tear or episiotomy. Anal sphincter injury following labour is the most common cause of anal incontinence in women, and can severely diminish quality of life. The incidence of clinical third- and fourth-degree lacerations varies widely; it is reported to be between 0.5 and 3.5% in Europe and between 5.9 and 8.9% in the USA,[2-5] possibly because of a more liberal use of midline episiotomy. It has been reported that a quarter of women who sustain anal sphincter injury at vaginal birth will have symptoms of compromised continence 6 months postnatally.[6, 7]
Risk factors for anal sphincter injury include nulliparity, birthweight, occipito-posterior position and operative vaginal delivery.[8, 9] Although these risk factors are part of normal labour and delivery practice, certain modifiable risk factors exist. These include episiotomy and the choice of instrument for operative vaginal delivery. Midline episiotomy, compared with mediolateral, is associated with significantly higher rates of anal sphincter injury. This predisposition is further compounded by instrumental delivery, with rates of anal sphincter injury of up to 50% reported when forceps delivery and midline episiotomy coexist.
Vaginal birth after caesarean (VBAC) was suggested to be associated with an increased risk of anal sphincter injury in deliveries after a previous caesarean section in a cohort containing some women with previous vaginal deliveries. Another study suggested that VBAC carries the same risk of sphincter injury as primiparous labour, but this was in a cohort with extremely high rates of sphincter injury. Large population-based studies have also suggested an increased rate of sphincter injury in women with a previous caesarean delivery and no previous vaginal birth; however, there has so far been no clear investigation of the characteristics of labour in such women.
VBAC may be associated with decreased maternal morbidity compared with routine repeat caesarean delivery, including complications of future pregnancies such as abnormal placentation.[14, 15] The rate of caesarean delivery has increased during the last 50 years,[16, 17] and as a result women with a previous caesarean delivery form an increasing proportion of the obstetric population.
Rates of maternal morbidity, such as uterine rupture, associated with trial of labour after caesarean delivery (TOLAC) have been widely reported but markers of maternal morbidity associated with VBAC in this group have been less thoroughly investigated. Therefore we examined the rates of and risk factors for anal sphincter injury in a discrete population, namely secundiparous women with a previous caesarean delivery.
This is a retrospective analysis of prospectively gathered data including all secundiparous women with a previous caesarean delivery who attempted VBAC over an 11-year period from January 2001 to December 2011. The National Maternity Hospital, Dublin is a tertiary referral university institution in which approximately 9000 women deliver annually. These data were compiled as part of continuous audit of labour and delivery and hence specific ethical approval was not deemed necessary.
Women who present to the hospital in their second pregnancy having had a previous caesarean delivery are counselled about TOLAC on their first visit to the clinic. Women are generally encouraged to attempt VBAC after one previous caesarean delivery, but clear requests for elective repeat caesarean sections are accepted.
Women who wish to attempt VBAC receive standard antenatal care and are seen on a weekly basis from 36 weeks of gestation. Spontaneous labour is awaited and women are allowed to progress beyond their due date. Induction is performed if spontaneous labour does not occur before 42 weeks of gestation.
Intrapartum management of women with a previous caesarean section is by one-to-one midwifery care and continuous electronic fetal monitoring is used to monitor fetal wellbeing. Patients are assessed for cervical dilatation in labour at regular 2-hourly intervals. The standard delivery management protocol in our institution requires that every woman who has a vaginal delivery has an examination of the anal sphincter, carried out by the attending midwife/obstetrician, immediately after their delivery. All episiotomies carried out at our institution are mediolateral. Episiotomies are not performed as routine but rather at the discretion of the birth attendant when indicated. All suspected cases of anal sphincter injury are examined and diagnosed by an obstetrician who has received specialist teaching in the diagnosis and repair of anal sphincter injury. Repairs are carried out in an operating theatre under regional or general anaesthesia. All women receive laxatives and broad-spectrum antibiotics for a period of 5–7 days immediately after repair, and are also reviewed by a physiotherapist and given instructions on pelvic floor exercises. Follow up of all women with sphincter damage 3 months after delivery is at a specially designated perineal clinic where a continence score and quality of life assessment are performed in addition to anal manometry and endoanal ultrasound.
Statistical analysis was performed using the chi-square test and Fisher's exact test. Comparison of means was accomplished with a Student's t test. Two groups were compared: those with and without anal sphincter injury in our VBAC cohort and those in the VBAC population compared with nulliparous women delivering in the same time period. Multiple logistic regression analysis was used to produce two multivariate models. The first model had parity as the dependent categorical variable and coefficients adjusted for anal sphincter injury, maternal age, birthweight, episiotomy, gestation and instrumental delivery. The second multivariate model was built to examine the specific factors associated with anal sphincter injury in the VBAC population alone, with sphincter injury as the dependent categorical variable adjusted for the same outcome parameters as mentioned above. Results are presented as adjusted odds ratios (exp B) with associated 95% confidence intervals and P values. The SPSS software package (version 20.0; SPSS, Chicago, IL, USA) was used and a two-tailed probability value of P < 0.05 was considered significant.
During the period from 1 January 2001 to 31 December 2011 there were 95 035 deliveries at the National Maternity Hospital, Dublin of which 4704 were deliveries of women in their second pregnancy with a previous caesarean section. One-third (1633/4704) of these women underwent an elective prelabour caesarean delivery, leaving 3071 who attempted to have a VBAC. Of these women who underwent TOLAC, 65% (1981/3071) had a successful VBAC and 35% (1090/3071) required a second caesarean delivery. The rate of anal sphincter injury in women who had a successful VBAC was 5% (98/1981).
In women who had a successful VBAC, the rate of instrumental delivery was 39% (771/1981), whereas 50% (49/98) of those who had sphincter damage required an instrumental delivery. Of 49 women who had an operative vaginal delivery and a subsequent anal sphincter injury 28 (58%) had a forceps delivery, with the remaining 21 having a vacuum delivery. Of the women in both groups who had an instrumental delivery, those with sphincter damage were more likely to have had a forceps than a vacuum delivery (58% [28/49] versus 33% [241/722], P = 0.001). A comparison of baseline labour characteristics between those who suffered sphincter injury while having a VBAC and those with an intact sphincter after having a VBAC can be seen in Table 1.
|VBAC with anal sphincter injury (n = 98)||VBAC with intact anal sphincter (n = 1883)||P-value|
|Instrumental delivery (%)||50 (49/98)||38 (722/1883)||0.02|
|Maternal age (years)||31.2 ± 4.2||32.1 ± 4.6||0.06|
|Birthweight (g)||3754 ± 528||3581 ± 458||< 0.0001|
|Gestation (weeks)||40.1 ± 1.1||39.8 ± 1.6||0.12|
|Episiotomy (%)||56 (55/98)||65 (1233/1883)||0.06|
|Epidural (%)||66 (65/98)||67 (1270/1883)||0.8|
During the 11-year study, 32% (31/98) of infants born to mothers with anal sphincter injury were macrosomic (birthweight > 4 kg). Women who had a VBAC and an anal sphincter injury were more likely to have a macrosomic infant (n = 31) than women who had a VBAC without sphincter injury (n = 346) (32% [31/98] versus 18% [346/1883], P = 0.002).
When length of labour was examined, the mean duration of labour in those with anal sphincter injury was 318 ± 225 minutes. There was no significant difference in the duration of labour between women who had an anal sphincter injury and those who did not (318 ± 225 minutes versus 293 ± 189 minutes, P = 0.2). One-third of women who had anal sphincter damage (33/98) laboured quickly, delivering in less than 3 hours, and of these 12 women delivered within 1 hour of admission, but the rate of rapid labour (labour duration < 3 hours) was similar among women who did and did not have an anal sphincter injury (33% [33/98] versus 33% [630/1883], P = 1.0). Similarly, there was no difference in the rate of prolonged labour (duration > 12 hours) between the group with anal sphincter injury and those without (4% [4/98] versus 2% [37/1883], P = 0.15). Among those parturients who had a successful VBAC there was no difference in the incidence of oxytocin use in those women who had sphincter damage and those who did not (23% [23/98] versus 19% [367/1883], P = 0.36). A multiple logistic regression model comparing mothers with anal sphincter injury after VBAC with those with an intact sphincter after VBAC can be seen in Table 2.
|VBAC with anal sphincter injury (n = 98)||VBAC with intact anal sphincter (n = 1883)||B||OR||95% CI||P-value|
|Instrumental delivery (%)||50 (49/98)||38 (722/1883)||0.497||1.64||1.08–2.5||0.02|
|Maternal age (years)||31.2 ± 4.2||32.1 ± 4.6||0.053||1.05||1.0–1.1||0.02|
|Birthweight (g)||3754 ± 528||3581 ± 458||0.001||0.99||0.99–1.0||0.001|
|Gestation (weeks)||40.1 ± 1.1||39.8 ± 1.6||0.027||0.97||0.89–1.1||0.5|
|Episiotomy (%)||56 (55/98)||65 (1233/1883)||0.68||0.5||0.32–0.8||0.004|
|Epidural (%)||66 (65/98)||67 (1270/1883)||0.25||0.78||0.49–1.23||0.28|
Women having a VBAC were at higher risk of having an anal sphincter injury than nulliparous women having a vaginal delivery over the same period (5% [98/1981] versus 3.6% [1189/32 514], P = 0.004, odds ratio 1.4, 95% confidence interval 1.15–1.75). The rate of instrumental delivery in woman attempting a VBAC was significantly higher than the rate seen in nulliparous women delivering vaginally at the institution over the study period (39% [771/1981] versus 30% [9733/32 514], P < 0.0001). The rate of episiotomy in women having a VBAC was 65%, compared with an institutional rate of episiotomy in nulliparous women delivering vaginally of 57% over the study period. A comparison of baseline labour characteristics between secundiparous women with a previous caesarean section and nulliparous women can be seen in Table 3. A multiple logistic regression model demonstrating comparison of secundiparous women with a previous caesarean with nulliparous women can be seen in Table 4.
|VBAC (n = 1981)||Nulliparous (n = 32 514)||P-value|
|Anal sphincter injury (%)||5 (98/1981)||3.6 (1189/32 514)||0.004|
|Instrumental delivery (%)||39 (771/1981)||30 (9733/32 514)||< 0.0001|
|Maternal age (years)||32.1 ± 4.6||29.5 ± 5.2||< 0.0001|
|Birthweight (g)||3589 ± 464||3522 ± 453||< 0.0001|
|Gestation (weeks)||39.82 ± 1.6||39.85 ± 1.2||0.3|
|Episiotomy (%)||65 (1288/1981)||57 (18 652/32 514)||< 0.0001|
|Epidural (%)||67 (1335/1981)||67 (22 278/32 514)||0.2|
|VBAC (n = 1981)||Nulliparous (n = 32 514)||B||OR||95% CI||P-value|
|Anal sphincter injury (%)||5 (98/1981)||3.6 (1189/32 514)||0.351||1.4||1.25–1.7||0.005|
|Instrumental delivery (%)||39 (771/1981)||30 (9733/32 514)||0.136||1.15||1.01–1.3||0.03|
|Maternal age (years)||32.1 ± 4.6||29.5 ± 5.2||0.785||2.2||2.12–2.26||0.001|
|Birthweight (g)||3589 ± 464||3522 ± 453||0.001||1.00||1.0–1.0||0.0001|
|Gestation (weeks)||39.82 ± 1.6||39.85 ± 1.2||0.001||0.999||0.96–1.04||0.95|
|Episiotomy (%)||65 (1288/1981)||57 (18 652/32 514)||0.108||0.89||0.79–1.01||0.08|
|Epidural (%)||67 (1335/1981)||67 (22 278/32 514)||0.143||1.15||1.02–1.3||0.02|
When details of the preceding first labours were examined, data were available on 88% (86/98) of women with anal sphincter injury. Seventy per cent (60/86) of women were in labour when they were underwent caesarean section in their first pregnancy. Of those in labour, 45% (27/60) had had a caesarean delivery at a dilatation of ≥ 5 cm, whereas 16% (10/60) underwent caesarean delivery despite having reached full dilatation in their first labour.
These results indicate that vaginal delivery in a second pregnancy in a large cohort of women with one previous caesarean delivery is associated with an incidence of anal sphincter injury of 5%. This increased rate of instrumental delivery may play a role in the accompanying rate of anal sphincter injury. However, of those 771 women who had an operative vaginal delivery, 722 (94%) had an intact anal sphincter after delivery. Similarly, the rate of fetal macrosomia was found to be substantial (32%) – this has also been shown to be a risk factor for anal sphincter injury. Birthweight, instrumental delivery and not having an episiotomy were all found to be strong predictors of sphincter injury when multiple logistic regression was carried out to compare those who suffered sphincter injury while having a VBAC with those who did not.
When rates of sphincter injury in secundiparous women with a previous caesarean delivery were compared with those of nulliparous women delivering over the same time period it was found that women having a VBAC were more likely to suffer sphincter injury. On multivariate analysis instrumental delivery, maternal age, birthweight and epidural anaesthesia were all found to be important contributing factors.
When the performance in first labour of women with anal sphincter injury was examined, we found that a large number of these women had progressed to an advanced stage before being delivered by caesarean section. These findings raise the possibility that women who have progressed to an advanced stage in their first labour have a second labour more akin to that of a multiparous woman with more propulsive contractions as a result of previous exposure to the labour process. This coupled with a ‘nulliparous’ perineum may be partly responsible for the increased rate of anal sphincter injury seen in this group of women.
We consider our findings to be robust because of the clearly defined nature of our cohort of parturients, namely it consisted of a large group of secundiparous women with a previous caesarean delivery. This cohort did not contain any women with previous vaginal deliveries, so essentially the pelvic floor of all women involved in the study was equivalent to that of nulliparous parturients. A previous study had suggested an increased rate of sphincter injury in cohorts like ours, but they examined women in their first delivery after a caesarean section and included some women who had had previous vaginal deliveries.
Our study is not without limitations. First, we do not have data on the first labours of those women who did not have an anal sphincter injury while having a VBAC. This is a topic that could form the basis of a future study. We have also not examined the reasons why some women in their second pregnancy elected to have a caesarean delivery rather than attempting VBAC. Information about women opting for an elective caesarean for reasons of fetal macrosomia or a previous failed operative vaginal delivery would have been useful, because these risk factors would have had an effect on sphincter injury rates had these women been allowed to go into labour a second time.
Rates of maternal morbidity, such as uterine rupture, associated with TOLAC have been widely reported in the literature. However, these data suggest an increased rate of anal sphincter injury, and this is a new aspect of maternal morbidity not typically counselled about when discussing VBAC after a previous caesarean delivery in women presenting in their second pregnancy.
In conclusion, secundiparous women with a previous caesarean delivery appear to be at increased risk of anal sphincter injury. Contributory factors may include an increased rate of operative vaginal delivery, increased birthweight and potentially a previous exposure to the labour process. An increased rate of anal sphincter injury, albeit at the 5% level, might now form part of the counselling discussion involving women deciding to attempt VBAC after a previous caesarean section.
None of the authors report any conflict of interest.
MPH contributed to study design, data collection and wrote the manuscript, MF contributed to study design and wrote the manuscript, MC contributed to data collection and revised the manuscript, MM contributed to data collection and revised the manuscript, COH contributed to study design and wrote the manuscript.
These data were compiled as part of a continuous audit of labour and delivery. Medical records for individual patients were not accessed during data collection and it was deemed by the chairman of our institutional ethics committee that specific ethical approval was not necessary.
No funding was received for this study.
The authors would like to acknowledge the contribution of Fionnuala Byrne, Information Officer at the National Maternity Hospital, Dublin.