Lal, Mira MD, MRCOG; Mann, Christopher H. MRCOG, MD; Callender, Roger FRCOG; Radley, Simon FRCS, MD
Anal incontinence, especially fecal incontinence, is a devastating symptom with a considerable negative impact on the physical and psychosocial health of the sufferer. The higher prevalence in women is related to childbirth, but the patient may be too embarrassed to volunteer information.1,2 Although there are differences in definitions and populations studied, the reported postpartum incidence after vaginal delivery varies between 1% and 45%.3–6
Vaginal delivery may lead to anal incontinence via direct anal sphincter trauma3,5–8 or pudendal neuropathy.9 Primiparas are at higher risk as are those who have had an instrumental vaginal delivery.10,11 The place of cesarean delivery is unclear4,7 with previous studies limited by small cesarean samples and those of mixed parity. Elective cesarean delivery has been considered protective.4,5–7 Two studies4,7 have reported a possible association between emergency cesarean in established labor and anal incontinence. One of these4 observed that the effect of emergency cesarean on anal incontinence “is still inconclusive,” and the other7 concluded that emergency cesarean may not prevent neurologic injury and subsequent anal incontinence.
The implications of a rising cesarean rate,12 its associated morbidity,10 and cost implications13 cannot be discounted. Recently published data14 indicate that the cesarean rate is showing the start of a downward trend in the United States and Nordic countries, but the implications of the previously high rate will influence future obstetric decision making. Although prevention of anal incontinence is desirable, the role of cesarean delivery14 needs further clarification.
The purpose of this study was to compare the incidence and severity of anal incontinence in primiparas after cesarean (emergency and elective) with those after noninstrumental vaginal delivery.
MATERIALS AND METHODS
This is part of a larger, complex study looking also at postcesarean urinary, sexual, and perineal problems along with primiparous psychosocial health, which cannot be assimilated into a single article. The study received ethical approval from the Dudley Ethics Committee, and all participants took part voluntarily after giving written informed consent.
Wordsley Hospital is part of the Dudley Group of Hospitals, National Health Service Trust. It has a wide catchment area with a mix of urban and rural areas. Five percent of the population is ethnic minority, and the rest is white.15
Wordsley Hospital deals with 4000 deliveries annually. Labor is managed according to standard protocols. Midwives conduct normal deliveries, which represent 70% of the total number. Obstetricians are available for complicated deliveries. Information on labor and delivery is recorded contemporaneously by notekeeping, including labor ward partograms. During the study period, the primigravid population constituted 32% of the deliveries. Their total cesarean rate was 13% with an elective cesarean rate of 5%. The instrumental delivery rate was 8%. Thirty-three percent had episiotomies, 15% second-degree tears, and 0.6% third-degree tears.
A literature search of all English-language publications–using MEDLINE (since 1966), Cochrane, CINAHL, and EMBASE databases–was carried out until October 2001. The key words used were “cesarean,” “caesarean,” “caesarian,” “cesarian” (on their own and combined with “section” or “birth”) and “anal incontinence,” “fecal incontinence,” “faecal incontinence,” and “flatus incontinence.” The relevant publications were critically appraised and then cross-referenced and a similar appraisal repeated.
Based on the evidence available at the time, a power calculation would have been a crude estimate at best. MacArthur et al4 missed out flatus incontinence, but extrapolating from their data, an incidence of 6.9% for postcesarean anal incontinence was calculated. This would mean seeing a total of 383 cesarean mothers (emergency and elective) and 383 controls for an 80% chance of a 5% significance test detecting a 92.4% difference. It was decided to limit our study (flatus incontinence included) to a definite period when the hospital's primiparous elective cesarean deliveries would be likely to approach an acceptable number comparable to the above. The recruitment period was also constrained to ensure the group of mothers under study did not become pregnant for a second time.
Consecutive primiparas who had delivered live singletons between April 1997 and September 1998 were selected from the hospital's computerized database, which registered demographic details, date, and mode of delivery. Letters were mailed to general practitioners informing them of the study and to find out whether there was any objection to any of their patients taking part.
Nine months after delivery, primiparas received letters inviting them and informing them about the ensuing phone discussion. Their willingness to take part was confirmed. If interested, further discussion involved checking whether there was any cause for their exclusion and explaining what participation in the study would involve. An appointment was made for a home interview and the option given to leave the study at any stage if they changed their minds. Women who preferred to be seen at Wordsley Hospital (31%) were appointed to a biweekly clinic run for the study. Postal questionnaires were not used as they can lead to misinterpretation and underreporting4 when investigating sensitive issues such as anal incontinence. The same female obstetrician (ML), who was not known to the participants before the study and was not part of their care-providing team, contacted and interviewed the women.
The total number of letters of invitation sent was 566 (346 targeting cesarean mothers and 220 vaginally delivered). Of these mothers, 126 (22.3%) had moved from their registered address and could not be contacted by letter or phone. Of the 440 mothers who could be contacted, another 105 (18.6%) were ineligible for various reasons, such as being wrongly coded as primiparous, pregnant again, having had an operation, severe psychiatric or medical problem, neurologic problem, urinary tract anomaly, or diversion, or an ill infant, and the study was concluded. A total of 335 (213 cesarean and 122 vaginally delivered) mothers were eligible to participate. Of the eligible mothers, 49 (14.6%) declined to take part giving a response rate of 85%.
The restrictions imposed by the cesarean rate, the inclusion and exclusion criteria enforced by letter and phone, time constraints, and the need to recruit before a second pregnancy resulted in the study being underpowered in relation to the power estimate. A total of 286 mothers were interviewed 10 ±.2 months after delivery, but two were excluded from the analysis as their notes could not be completed. Of the 284 participants included for the purposes of this study, 184 consecutive mothers who delivered by cesarean (104 emergency and 80 elective) were selected as the index group. Another 100 consecutive women with cephalic presentations, who underwent noninstrumental vaginal delivery, served as controls matched for age, parity, catchment area, race, and period of delivery. Compared with cesarean mothers, most vaginally delivered mothers had returned to full-time work resulting in time constraints, which restricted interview appointments and slowed recruitment. Exclusion criteria related not only directly to anal incontinence (ie, bowel diversion, anorectal surgery) but also to other pelvic floor or psychiatric symptoms, which could affect the assessment of anal incontinence (see previous paragraph). A comprehensive structured questionnaire about delivery, bowel function, and relevant medical history was completed. Although the questioning aimed to investigate the postdelivery period, pertinent questions about pregnancy and prepregnancy functioning were also included.
The questionnaire was based on the validated “Birmingham Bowel and Bladder Questionnaire.” Hiller et al16 have validated this questionnaire by measuring its content, construct and criterion validity, internal consistency, reliability, and responsiveness in a sample of 630 women. Content validity was supported by reports and missing data patterns. Factor analysis indicated a good internal structure, and comparison with clinical and physiological tests showed criterion validity. Internal consistency was shown with key domain question analysis and Cronbach.αs. The κ values showed good testretest reliability, and key question correlation over time proved responsiveness.
Questions relating to defecatory symptoms included those about frequency and consistency of stools, laxative use, as well as those more specific to anal sphincter function (ie, urgency, flatus incontinence, urge and passive incontinence, soiling, and pad use). The questions specific for anal sphincter function reflected varying degrees of external or internal anal sphincter dysfunction presenting as anal incontinence. Anal incontinence was defined as involuntary loss of solid or liquid stool or gas (ie, fecal or flatus incontinence). Urgency was defined as inability to defer defecation for 5 minutes. Urge incontinence was defined as inability to defer defecation after a desire to defecate with the subject being aware, and passive incontinence as leakage of feces where there was no desire to defecate and the subject was unaware when it was occurring.
Emergency cesarean was defined as operative delivery carried out as an emergency, before or when in established labor. Early labor was defined as cervical dilatation less than 8 cm and late labor more than 8 cm, according to Fynes et al.7 Elective cesarean delivery was carried out as a planned procedure, or immediately after the onset of labor in anyone due for a planned section.17
All episiotomies were mediolateral. Perineal tears were recorded on the birth record by the attending doctor (registrar/consultant) or experienced midwife and classified according to standard definitions as follows:
1. First-degree tear: laceration extending through the vaginal mucosa and perineal skin only.
2. Second-degree tear: laceration extending to the perineal body and not the anal sphincter.
3. Third-degree tear: laceration extending into the anal sphincter without involvement of the anorectal mucosa.
4. Fourth-degree tear: laceration extending into the perineal muscle involving the anorectal mucosa.
“New” as in “new symptoms” denotes symptoms of anal incontinence occurring for the first time after delivery. After the interviews, obstetric details were confirmed by ML from the handwritten hospital obstetric records, including partograms. The contemporaneously written notes of the attending obstetrician, midwife, pediatrician, and anesthesiologist and discharge summaries were systematically searched for relevant obstetric data, such as mode of delivery, gestational age, cervical dilatation, infant birth weight, etc, and transferred to a form for consistency. In a few instances where the infant was transferred to the special care infant unit at birth, details were confirmed from its computerized database. All data were rechecked and transferred to a dedicated database (Microsoft Excel; Microsoft, Seattle, WA). Anorectal physiologic tests and anal endosonography were not done.
Participants' characteristics were defined using counts and percentages for categoric variables and mean ± standard deviation or median (range) for continuous variables. Categoric variables were assessed for homogeneity using χ2 test or Fisher exact test as appropriate. Continuous variables in independent groups were compared using analysis of variance. The Kruskal-Wallis test was used for non-normally distributed data. A P value of .05 or less was considered to be of statistical significance. The magnitude of the effect was shown by using relative risks (RR) and confidence intervals (CI). Statistical analysis was carried out using SPSS 10.0.7 for Windows (SPSS Inc., Chicago, IL).
Our sample was representative of the study population at the time with 13 (5%) of the participants being Afro-Caribbean or other ethnic minorities and 271 (95%) being white. The demographic and obstetric details recorded are presented in Table 1.
When compared using analysis of variance, the cesarean group (n = .184) was similar to the controls (n = .100) with respect to maternal age (P = .109), infant weight (P = .331), and head circumference (P = .186). Our sample was homogenous in that they were all English-speaking primiparas with live singletons delivered at the same hospital where obstetric care was similar. No statistically significant difference in maternal age, infant weight, or head circumference was seen. However, gestational age was not normally distributed, as the mean duration of gestation for the elective cesarean mothers was less than that of the other participants. When gestational age was compared between the groups using the Kruskal-Wallis test, a statistically significant difference was shown (P < .001).
The emergency cesarean group included 21 women before the onset of labor and 83 women after the onset of labor (63 in early labor and 20 in late labor). In addition to fetal distress, indications for emergency cesarean included failure to progress, where labor was judged to be arrested in the first or second stage, and “preeclampsia,” where there was increasing fetomaternal compromise with conservative management but vaginal delivery was not feasible. None of the elective cesarean participants in this study had cervical dilatation, and they were not considered to be in labor. Among the indications for elective cesarean delivery, only three were for maternal request from healthy pregnant women. The term “miscellaneous” included those with premature rupture of membranes or a large infant. Whether participants were seen at home or at the clinic, there was no significant difference among the participants or data from the two locations.
The Afro-Caribbean/ethnic participants reported no defecatory symptoms. Two previously healthy cesarean mothers (both elective) developed occasional urge incontinence during pregnancy, and two others (one emergency cesarean, one vaginally delivered) developed severe flatus incontinence during pregnancy, reverting back to prepregnancy continence after delivery.
Two in the elective cesarean group were diabetic (one a gestational diabetic and the other an insulin-dependent diabetic), but neither had anal incontinence. There were no diabetics among those delivered by emergency cesarean or vaginally. Fourteen (10 emergency, four elective) cesarean mothers and six mothers delivered vaginally had irritable bowel syndrome known to be associated with anal incontinence,18 but of these, only one emergency cesarean mother developed postpartum anal incontinence. Six (three emergency, three elective) cesarean mothers and eight controls were still breast-feeding. Among these women, menstruation had returned to normal in all, except one with amenorrhea (emergency cesarean), one who had just started menstruating (elective cesarean), and one with amenorrhea (control). None of the mothers who were breast-feeding reported anal incontinence despite reported association.19 No mother with new symptoms of anal incontinence had a history of chronic constipation or laxative use.
Anal incontinence was first present in nine (5%) of the overall cesarean group and eight (8%) of the controls (χ2 = 1.113, df = 1, P = .427, RR 0.611, 95% CI 0.25, 1.53). Of the cesarean participants, new symptoms were noted in six (6%) of those who had an emergency cesarean (n = .104) and three (4%) of those who had an elective cesarean (n = 80).
The patterns of anal incontinence related to the mode of delivery are shown in Figure 1. At least two symptoms were reported by two (3%) of the elective cesarean, one (1%) of the emergency cesarean, and two (2%) of the controls. Anal incontinence, severe enough to require pad use, was reported by two (3%) cesarean mothers after elective cesarean delivery and one (1%) of the controls (Fisher exact test, P = .716).
Of the 21 mothers who had emergency cesarean delivery when not in labor, two (10%) who had antepartum hemorrhage with fetomaternal compromise subsequently developed anal incontinence. Three of 63 emergency cesarean mothers (5%) in early labor and one of 20 mothers (5%) in late labor developed new anal incontinence. This was not a statistically significant difference (Fisher exact test value = 3.496, P = .184, RR 1.03, 95% CI 0.43, 2.47).
Five (5%) of 101 (21 emergency, 80 elective) cesarean mothers who were not in labor had new anal incontinence and when compared with vaginal delivery, the RR was 1.62 and the 95% CI was 0.81, 3.23. Four (5%) of the 83 emergency cesarean mothers who were in labor developed new anal incontinence and when compared with vaginal delivery, the RR was 1.66 and the 95% CI was 0.83, 3.32, which was similar to cesarean mothers not in labor.
New symptoms were reported by one (3%) with an intact perineum and five (23%) with a second-degree tear. This is a statistically significant difference (Table 2, Fisher exact test value = 9.697, P = .014). The incidence of anal incontinence after elective cesarean delivery reported by three (4%) mothers was comparable to that reported after vaginal delivery with an intact perineum.
This study includes a large sample of primiparous cesarean mothers from a single obstetric unit who were matched with a noninstrumental vaginally delivered group. We found that postpartum anal incontinence is not uncommon. The incidence after vaginal delivery is comparable to that reported in several previous studies.4,19–22 It is, however, at variance with all except two recent studies22,23 regarding the presumed protective effect of cesarean delivery on the pelvic floor and subsequent anal incontinence. We report that elective cesarean or prelabor emergency cesarean is not always protective, and symptoms can be severe.
MacArthur et al in their study4 from Birmingham reported on 906 patients who were interviewed by different observers 10 ± 2 months postdelivery. They reported an incidence of anal incontinence of 4% overall but did not include flatus incontinence. Their cesarean sample (n = .113) was of mixed parity. The small elective cesarean cohort was asymptomatic, but six mothers who had an emergency cesarean in labor developed anal incontinence. Their study, like ours, is prone to recall bias, but this is less likely to occur after the first childbirth, as noted in a previous study24 where there has been a good recall of events surrounding delivery.
A recent multicenter study23 on fecal incontinence using a postal questionnaire had a 71% response rate. A prevalence of postpartum anal incontinence of 9.6% for all modes of delivery was reported, though questions on fecal urgency were not asked. The authors observed that cesarean delivery offered some protection. A multivariate regression analysis was carried out on primiparas, with both types of cesarean combined into one group because of concerns about sample size. In our study, multivariate regression analysis was not reported, as there were so few newly identified cases of anal incontinence that any conclusions from the multivariate analysis would be too tentative.
In two other British prospective studies5,21 in primiparas, Sultan et al found an incidence of anal incontinence of 13% and 4%, respectively. The latter figure (nearer to our report) was the incidence in vaginally delivered controls without instrumentation. The slightly higher figure in the former study could be attributed to the inclusion of instrumental vaginal deliveries and third- and fourth-degree tears, which the latter study had excluded, as we have. Symptomatology was our only investigative tool, whereas Sultan et al5 also used anorectal physiologic tests and anal endosonography. Their elective (n = 7) and emergency cesarean (n = 16) mothers were asymptomatic, but the pudendal nerve terminal motor latency (left) was prolonged, and the perineal plane on straining was lower in mothers who had emergency cesarean in labor. Any relevance of these observations to symptomatology could not be ascertained. Others have made similar observations, and the clinical importance of these tests remains contentious.25
Fynes et al7 from Dublin, when investigating the relationship between the timing of cesarean and anal sphincter injury in primiparas, found a reduction in the increment of squeeze pressure and a prolonged pudendal nerve terminal motor latency after cesarean in late labor. (more than 8 cm), but not early labor (less than 8 cm). Again, the implication of these findings to clinical management remains unknown, as their small cesarean sample (eight elective, 26 emergency) was without symptoms. Using the same criteria for cervical dilatation, we did not find any difference in the incidence of anal incontinence, whether in early or late labor.
Our observation that significantly more mothers developed anal incontinence after second-degree tears has not been reported before. This needs to be interpreted with caution because of small numbers. Third-degree tears can be misdiagnosed as second-degree tears by the relatively less experienced,26 but our rate of 0.6% during the study period is comparable to that in other studies.27,28 Recognition can be improved by training.29 Alternatively, second-degree tears may have a hitherto unrecognized influence on the development of postpartum anal incontinence, possibly related to trauma to the perineal body. Anal ultrasound may have helped in identifying anal sphincter trauma in these women.
Crawford et al20 from the United States reported an incidence of postpartum anal incontinence of 6% in their vaginally delivered controls without third- and fourth-degree tears (n = 35), which was similar to ours. Unlike our study, however, they did not subdivide their controls according to the degree of perineal trauma, and so the association of each category to symptoms is unclear. Abramowitz et al in a prospective French study30 found a significant association between anal incontinence and perineal tears. Again, the perineal tears were not categorized into first- and second-degree tears, so the association of each type to symptomatology is not evident. In contrast to these and our study, Sultan et al5 did not find such an association.
MacLennan et al22 observed that pelvic floor morbidity, including anal incontinence, can follow cesarean birth and that its prevalence is comparable to that after noninstrumental vaginal delivery. Our finding is similar though more specific in reporting that the risk after elective cesarean is comparable to that after noninstrumental vaginal delivery with an intact perineum. Theirs was a general population survey, in which one randomly selected male or female respondent (n = 1546) from each family was interviewed at home regarding pelvic floor symptoms in any family member. Obstetric details were not confirmed. They reported a prevalence of fecal incontinence of 3.5% and flatus incontinence of 10.9% in women with a mean age of 44.8 years. Their observation that pregnancy (greater than 20 weeks) was also implicated agrees with our finding.
Based on our literature search, this study seems to be the first of its type to report anal incontinence in primiparas after elective cesarean severe enough to necessitate pad use. When compared with vaginal delivery, the RR of cesarean when not in labor was similar to that of cesarean when in labor. This would call to question the belief that anal incontinence is caused solely by the effect of labor on the pelvic floor and can always be avoided by the performance of a cesarean before the onset of labor. Whether our findings can be extrapolated to different populations is unclear.
There remains considerable debate as to the role of cesarean in those women who have previously sustained an anal sphincter injury.10,31 Our study supports those who argue against routine elective cesarean delivery31 after repair of anal sphincter injury. This requires further evaluation, particularly as new sphincter injuries in subsequent deliveries are uncommon.32 Other intrinsic factors, such as the properties of collagen/connective tissue, hereditary susceptibility, or impaired rectal sensation9 could also contribute to postpartum anal incontinence in the absence of labor.
Future prospective long-term studies should investigate the questions raised by this study, with a shift in emphasis from labor to other factors such as pregnancy and intrinsic factors. Until the place of cesarean delivery is further clarified, we should continue to assess individually all pregnant or laboring women. Decisions about the mode of delivery should be made according to the clinical situation rather than the presumed protective effect of cesarean delivery on the pelvic floor. Routine enquiry should be made pre- and postnatally about anal incontinence. Women should be encouraged to carry out pelvic floor exercises, during pregnancy and postpartum, irrespective of the mode of delivery.
1. Johanson JF, Lafferty J. Epidemiology of fecal incontinence: The silent affliction. Am J Gastroenterol 1996;91:33–6.
2. Leigh RJ, Turnberg LA. Faecal incontinence: The unvoiced symptom. Lancet 1982;1:1349–51.
3. Varma A, Gunn J, Gardiner A, Lindow SW, Duthie GS. Obstetric anal sphincter injury prospective evaluation of incidence. Dis Colon Rectum 1999;42:1537–42.
4. MacArthur C, Bick DE, Keighly MRB. Faecal incontinence after childbirth. Br J Obstet Gynaecol 1997;104:46–50.
5. Sultan AH, Kamm MA, Hudson CN, Thomas JM, Bartram CI. Anal sphincter disruption during vaginal delivery. N Engl J Med 1993;329:1905–11.
6. Fornell EKU, Berg G, Hallböök O, Matthiesen LS, Sjödahl R. Clinical consequences of anal sphincter rupture during vaginal delivery. J Am Coll Surg 1996;183:553–8.
7. Fynes M, Donnelly VS, O'Connell R, O'Herlihy C. Cesarean delivery and anal sphincter injury. Obstet Gynecol 1998;92:496–500.
8. Snooks SJ, Henry MM, Swash M. Faecal incontinence due to external anal sphincter division in childbirth is associated with damage to the innervation of the pelvic floor musculature: A double pathology. Br J Obstet Gynaecol 1985;92:824–8.
9. Snooks SJ, Setchell M, Swash M, Henry MM. Injury to the innervation of the pelvic floor sphincter musculature in childbirth. Lancet 1984;1:546–50.
10. Sultan AH, Stanton SL. Preserving the pelvic floor and perineum during childbirth–elective caesarean section? Br J Obstet Gynaecol 1996;103:731–4.
11. Haadem K, Ohrlander S, Lingman G. Long-term ailments due to anal sphincter rupture, caused by delivery–a hidden problem. Eur J Obstet Gynecol Reprod Biol 1988;27:27–32.
12. Notzon FC, Chattingius S, Bergsjø P, Cole S, Taffel S, Irgens L, et al. Cesarean section delivery in the 1980s: International comparison by indication. Am J Obstet Gynecol 1994;170:495–503.
13. Henderson J, McCandlish R, Kumiega L, Petrou S. Systematic review of economic aspects of alternative modes of delivery. Br J Obstet Gynaecol 2001;108:149–57.
14. Thomas J, Paranjothy S. The National Sentinel Caesarean Section Audit report. London: RCOG Press, 2001:1–3.
15. The annual report of the Director of Public Health 1993: Department of Public Health Medicine, Dudley Health Authority.
16. Hiller L, Radley S, Mann CH, Radley SC, Begum G, Pretlove ST, et al. Development and validation of a questionnaire for the assessment of bowel and urinary tract symptoms in women. Br J Obstet Gynaecol 2002;109:413–23.
17. Government Statistical Service. NHS maternity statistics, England: 1989–90 to 1994–95. London: Department of Health, 1997:28.
18. Donnelly VS, O'Herlihy C, Campbell DM, O'Connell PR. Postpartum faecal incontinence is more common in women with irritable bowel syndrome. Dis Colon Rectum 1998;41:586–9.
19. Zetterström JP, López A, Anzén B, Dolk A, Norman M, Mellgren A. Anal incontinence after vaginal delivery: A prospective study in primiparous women. Br J Obstet Gynaecol 1999;106:324–30.
20. Crawford LA, Quint EH, Pearl ML, DeLancey JOL. Incontinence following rupture of the anal sphincter during delivery. Obstet Gynecol 1993;82:527–31.
21. Sultan AH, Kamm MA, Bartram CI, Hudson CN. Anal sphincter trauma during instrumental delivery. Int J Gynecol Obstet 1993;43:263–70.
22. MacLennan AH, Taylor AW, Wilson DH, Wilson D. The prevalence of pelvic floor disorders and their relationship to gender, age, parity and mode of delivery. Br J Obstet Gynaecol 2000;107:1460–70.
23. MacArthur C, Glazener CMA, Wilson PD, Herbison GP, Gee H, Lang GD, et al. Obstetric practice and faecal incontinence three months after delivery. Br J Obstet Gynaecol 2001;108:678–83.
24. Simkin P. Just another day in a woman's life? Part II: Nature and consistency of women's long-term memories of their first birth experience. Birth 1992;19:64–81.
25. Wattchow DA, Rieger N. Faecal incontinence after vaginal delivery. Lancet 1999;354:965.
26. Sultan AH, Kamm MA, Hudson CN. Obstetric perineal trauma: An audit of training. Br J Obstet Gynaecol 1995; 15:19–23.
27. Sørensen M, Tetzschner T, Rasmussen OØ, Bjarnesen J, Christiansen J. Sphincter rupture in childbirth. Br J Surg 1993;80:392–4.
28. Sultan AH, Kamm MA, Hudson CN, Bartram CI. Third degree obstetric anal sphincter tears: Risk factors and outcome of primary repair. Br Med J 1994;308:887–91.
29. Keighley MRB, Radley S, Johanson R. Consensus on prevention and management of post-obstetric bowel incontinence and third degree tear. Clin Risk 2000;6:231–7.
30. Abramowitz L, Sobhani I, Ganansia R, Vuagnat A, Benifla JL, Darai E, et al. Are sphincter defects the cause of anal incontinence after vaginal delivery? Dis Colon Rectum 2000;43:590–8.
31. Given FT, Browning G. Repair of old perineal lacerations. Am J Obstet Gynecol 1988;159:779–84.
32. Fynes M, Donnelly V, Behan M, O'Connell PR, O'Herlihy C. Effect of second vaginal delivery on anorectal physiology and faecal incontinence: A prospective study. Lancet 1999;354:983–6.