Materials and Methods
From January to July 1998, we informed nulliparous women (those with no previous deliveries after 22 weeks' gestation) about the study during antenatal clinic visits, and eligible women were included immediately after delivery. Inclusion criteria were vaginal delivery and absence of anal sphincter tears (third- or fourth-degree perineal tears). Women were excluded if anal sphincter ruptures were diagnosed clinically (third- or fourth-degree perineal tears). Partial ruptures of external anal sphincters were considered third-degree perineal tears. Women delivered by cesarean were not eligible because the probability of anal sphincter rupture and fecal incontinence was low in that group.7 Participants were asked at recruitment whether they had fecal incontinence before or during pregnancy. All deliveries were by certified midwives or senior residents under the supervision of board-certified obstetricians. Evaluations of perineums after delivery were done by senior residents under supervision of board-certified obstetricians.
Members of the study team (DLF, SB, or CS) were called for anal endosonography immediately after vaginal deliveries, before repair of any perineal tears. Endosonography was done with a Bruel & Kjaer Medical (Copenhagen, Denmark) system 3535 with a 10-MHz, 360° Endosonic Probe Type 1850. Anal sphincter rupture was defined as a gap in the hyperechogenic ring of internal or external anal sphincter.8–11 Assessment of intraobserver variability soon after delivery was not possible, and formal assessment of interobserver variability in the delivery room would not have been easily accepted by the women. Therefore, two sonographers evaluated videotaped recordings of 42 consecutive participants. The overall proportion of agreement was 0.83, and the Kappa statistic was 0.58, which was satisfactory.12 Every attempt was made to conceal the details of deliveries from the sonographers. Forceps or vacuums were removed from the delivery room and sonographers were instructed not to evaluate perineums clinically. Obstetricians and participants were not informed of anal endosonography findings, thus the procedure did not modify the care of participants or their outcomes.
Three months after delivery, we mailed questionnaires on fecal continence to participants. If we received no answer after a month, we sent a reminder. After another month, we tried to contact the women by telephone to encourage them to return the forms. The questionnaire evaluated incontinence to flatus, liquid, or solid stools. A positive response to any of the questions was considered fecal incontinence. Items in the questionnaire had been asked of women treated for postobstetric fecal incontinence and were reliable, valid, and sensitive.13,14 In a preliminary study of 100 women, we compared answers to the questionnaire with an interview conducted by an experienced proctologist. There was good agreement between the questionnaire and the interviewer for detecting fecal incontinence (K = .73).
We assessed the effects of fecal incontinence on quality of life with the short-form incontinence effect questionnaire, which evaluated the effects of incontinence on seven items, including household chores, physical recreation, entertainment activities, travel more than 30 minutes from home, social activities, emotions (nervousness, depression), and frustration.15 Each item was scored by women on a 10-cm-long visual analog scale, anchored with the words “not at all” and “very much.” The scores of all items can be pooled to calculate a mean effects score. Although the questionnaire was designed primarily for urinary incontinence, it had good face validity to measure the effects of fecal incontinence.
Differences between groups were assessed by unpaired t test for continuous variables and Fisher exact test for categoric variables. Variables statistically associated with symptoms in univariate analysis were entered in a multivariate logistic regression model to determine the combination of independent variables that best predicted symptoms of fecal incontinence and calculate the odds ratios (ORs) with their 95% confidence intervals (CIs). Birth weight was entered in the model as two categories, with cutoff at the 75th percentile of the variable distribution (birth weight above 3500 g), chosen because the risk of fecal incontinence increased in the upper quartile, (9, 1%; 8, 1%; 11, 4%; 30, 8%; respectively, for quartiles 1 to 4). Analyses were done with the Statistical Package for Social Science (SPSS Inc., Chicago, IL). A sample of 25 women with anal sphincter ruptures was needed for precision of ± 20% in the estimate of positive predictive value of anal endosonography. Hypothesizing that an anal sphincter tear would be diagnosed in 20% of participants, we needed to recruit 125 women. One hundred fifty women were recruited to account for eventual losses to follow-up. The study protocol was approved by the institutional ethics committee, and all participants gave written informed consent.
Between January and July 1998, 1420 women delivered in our department, of whom 486 were eligible. One hundred fifty women were included in the study. Four women reported incontinence to flatus before pregnancy (2.7%, 95% CI 1%, 7%), and 11 reported incontinence to flatus during pregnancy (7.3%, 95% CI 4%, 13%). None reported daily fecal incontinence. One hundred fifteen women (77%) had epidural analgesia, 46 (31%) had instrument deliveries, of which 35 were with forceps. Eighty women (53%) had episiotomies, of which 32 (40%) were median and 48 (60%) were mediolateral. The latter episiotomies were done mainly in cases of instrument deliveries. Maternal age, birth weight, proportion of epidural analgesia, instrument deliveries, and episiotomies were similar between studied and eligible women.
Tears of anal sphincters were diagnosed by anal endosonography immediately after delivery in 42 women (28.0%, 95% CI 21%, 36%). The external anal sphincter alone was involved in 30 women (20.0%), the internal anal sphincter alone in two women (1.3%), and both in ten women (6.7%). Figure 1 shows a sonograph of an intact anal sphincter and Figure 2 a ruptured anal sphincter. The procedure was well tolerated, even by women who had no epidural analgesia.
Completed questionnaires on fecal incontinence were returned by 144 women (96%), with a median delay of 142 days (range 103–334 days) after delivery, not different between continent and incontinent women. Six women were lost to follow-up. Fecal incontinence was reported by 22 women (15%, 95% CI 10%, 22%). Twenty women were incontinent to flatus, ten reporting daily episodes of incontinence. Sixteen women reported incontinence to flatus only (16 of 22, 73%, 95% CI 50% 89%). Six women were incontinent to liquid or solid stools, four of them reporting daily episodes of incontinence. One woman already had secondary anal sphincter repair. Among 22 women who reported fecal incontinence, 12 (55%, 95% CI 32%, 76%) placed a mark beyond 1 cm on the visual analog scale for at least one item of the short-form incontinence effects questionnaire, compared with 15 (12%, 95% CI 7%, 19%) of 122 women who did not report fecal incontinence (OR 8.6, 95% CI 3.2, 23.3). The associations between delivery variables and subsequent fecal incontinence are reported in Table 1.
Fecal incontinence was reported by four of 35 women who had forceps deliveries, five of 19 women who had vacuum deliveries, and two of two women who had forceps deliveries after failed attempts at vacuum delivery. The differences were not statistically significant.
The relationship between results of anal endosonography and development of subsequent fecal incontinence is reported in Table 2. Clinically undetected anal sphincter tears diagnosed by anal endosonography were associated with fecal incontinence 3 months after delivery (OR 7.9; 95% CI 3.0, 19.5). When only women who reported effects of fecal incontinence on quality of life were taken into account, the association of symptoms with endosonographic diagnoses of anal sphincter defects was similar (OR 8.1; 95% CI 2.1, 27.1). Sensitivity and positive predictive value of anal endosonography immediately after childbirth for predicting fecal incontinence were 68% (95% CI 49%, 88%) and 37% (95% CI 22%, 51%), respectively. When incontinence with an effect on quality of life was the outcome, the sensitivity and positive predictive value of anal endosonography were 73% (95% CI 46%, 99%) and 20% (95% CI 7%, 32%), respectively.
Factors associated with fecal incontinence by univariate analysis (ie, birth weight and fecal incontinence during pregnancy) were entered with tears of anal sphincters in a logistic regression model (Table 3). Anal sphincter tears diagnosed by endosonography, birth weight above 75th percentile (3500 g), and fecal incontinence during pregnancy were statistically significant and independently associated with fecal incontinence.
This study confirmed the high prevalence of fecal incontinence 3 months after vaginal delivery, even in women with no clinically diagnosed third- or fourth-degree perineal tears. Although women who delivered by cesarean and who had clinically evident anal sphincter tears were excluded, the reported rate of fecal incontinence was similar to that reported by others.1,2 Most affected women reported incontinence to flatus only. Half the incontinent women in our cohort reported effects of fecal incontinence on their quality of life, emphasizing the importance of treatment of perineal tears after delivery.
Anal sphincter tears diagnosed by anal endosonography 6 weeks after delivery were associated with fecal incontinence.7 Our study showed that diagnosis of clinically undetected anal sphincter tears by anal endosonography immediately after delivery predicted fecal incontinence. Delivery histories were masked to sonographers and test results were masked to the obstetricians and participants, so an incontinence treatment or reporting bias was unlikely.
Few obstetric variables are good clinical predictors of fecal incontinence. In the present study, only birth weight above the 75th percentile and fecal incontinence during pregnancy were associated with fecal incontinence 3 months after delivery. The association between results of endosonography and subsequent fecal incontinence remained significant after adjusting for effects of birth weight and fecal incontinence during pregnancy.
Other studies suggested that instrument delivery was a significant risk factor for fecal incontinence.1,7 The present study differed from previous ones because we excluded clinically diagnosed anal sphincter tears. After instrument deliveries, tears of anal sphincters might be better diagnosed clinically because they can be assessed more carefully or more extensively for perineal damage, allowing better exposure of the anal sphincter. The proportion of instrument deliveries (37%) might seem high in this study, but the denominator of that proportion was women who delivered for the first time, and only vaginal deliveries were included.
Anal endosonography after vaginal delivery was well tolerated, even when women had no epidural analgesia. We anticipated that the quality of endosonographic examinations might be less than optimal immediately after delivery because of tissue edema, lacerations, and bleeding. However, the strength of association of anal sphincter tears diagnosed by endosonography with subsequent fecal incontinence was in the same range as that described by authors who did endosonography 6 weeks after delivery, in optimal conditions.7 We found assessment of the internal anal sphincter to be particularly difficult. In only two women (1.3%) were isolated tears of internal anal sphincter diagnosed, so restricting diagnoses to external anal sphincter defects would allow recognition of almost all women at risk for fecal incontinence.
We found a relatively high proportion of women who reported occasional incontinence to flatus during pregnancy. Those women were at higher risk of subsequent fecal incontinence, independently of results of anal endosonography. We believe that a physiopathologic mechanism other than anal sphincter tear is responsible for those symptoms, which could also explain why sensitivity and positive predictive value of anal endosonography for predicting fecal incontinence were not higher. Our relatively small sample limited the precision of our estimates of predictive characteristics of anal endosonography immediately after childbirth. The burden of fecal incontinence on women's health and well-being is not small. There is need for further research to evaluate interventions to prevent anal sphincter tears and decrease risks of fecal incontinence.
1. MacArthur C, Bick DE, Keighley MRB. Faecal incontinence after childbirth. Br J Obstet Gynaecol 1997;104:46–50.
2. Fornell UEK, 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.
3. Thacker SB, Banta HD. Benefits and risks of episiotomy: An interpretative review of the English language literature, 1860–1980. Obstet Gynecol Surv 1983;38:322–38.
4. Crawford LA, Quint EH, Pearl ML, DeLancey JOL. Incontinence following rupture of the anal sphincter during delivery. Obstet Gynecol 1993;82:527–31.
5. Sorensen M, Tetzschner T, Rasmussen OO, Bjarnesen J, Christiansen J. Sphincter rupture in childbirth. Br J Surg 1993;80:392–4.
6. 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.
7. Sultan AH, Kamm MA, Hudson CN, Thomas JM, Bartram CI. Anal-sphincter disruption during vaginal delivery. N Engl J Med 1993;329:1905–11.
8. Sultan AH, Nicholls RJ, Kamm MA, Hudson CN, Beynon J, Bartram CI. Anal endosonography and correlation with in vitro and in vivo anatomy. Br J Surg 1993;79:104–6.
9. Deen KL, Kumar D, Williams JG, Olliff J, Keighley MRB. Anal sphincter defects. Correlation between endoanal ultrasound and surgery. Ann Surg 1993;2:201–5.
10. Sultan AH, Kamm MA, Talbot IC, Nicholls RJ, Bartram CI. Anal endosonography for identifying external sphincter defects confirmed histologically. Br J Surg 1994;81:463–5.
11. Sentovich SM, Blatchford GJ, Rivela LJ, Lin K, Thorson AG, Christensen MA. Diagnosing anal sphincter injury with transanal ultrasound and manometry. Dis Colon Rectum 1997;40:1430–4.
12. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159–74.
13. Sangalli MR, Marti MC. Results of sphincter repair in postobstetric fecal incontinence. J Am Coll Surg 1994;179:583–6.
14. Osterberg A, Graf W, Karlbom U, Påhlman L. Evaluation of a questionnaire in the assessment of patients with faecal incontinence and constipation. Scand J Gastroenterol 1996;31:575–80.
15. Uebersax JS, Wyman JF, Shumaker SA, McClish DK, Fantl JA. Short forms to assess life quality and symptom distress for urinary incontinence in women: The Incontinence Impact Questionnaire and the Urogenital Distress Inventory. Continence Program for Women Research Group. Neurourol Urodyn 1995;14:131–9.