The most common cause of anal sphincter injury is obstetric trauma sustained during vaginal delivery. One study using anal endosonography found external anal sphincter tears in 35% of 79 primiparous women who had vaginal deliveries, with 13% having some degree of fecal incontinence 6 weeks after delivery.1 Two recent reports found a similar frequency of tears,2,3 although two others found 20%4 or 11.5%.5 Naturally, concern has been expressed that the higher figures suggested do not equate with the incidence in clinical practice. For example, the incidence of overt third-degree tears in the United Kingdom is less than 1%.6,7 Also, if the incidence of sphincter damage was 35%, this would be expected to generate up to 400 referrals with fecal incontinence from a unit delivering 3500 women a year,5 many more than are seen in practice.7 This discrepancy could be partly explained by the underreporting of fecal incontinence or delayed presentation until the effects of menopause summate with those of sphincter damage to produce overt symptoms.8 However, the significance of external sphincter trauma should not be understated, as repair of external sphincter tears is of proven benefit,9 suggesting a direct link between sphincter damage and incontinence.
The majority of workers in this field have used a mechanically rotated 7-MHz crystal (B&K Medical, Sandhoften, Denmark) designed for rectal scanning. A 10-MHz crystal, adapted for use in the anal canal, has now been available for several years. The higher frequency and near-field focusing have greatly improved anatomic resolution of the anal structures. The development of three-dimensional ultrasound10 and comparisons with endocoil magnetic resonance imaging have further clarified the endosonographic anatomy of the sphincter complex.11–14 In previous studies tears involving the external sphincter were not distinguished from other closely related structures, therefore all these lesions have been considered external sphincter tears. Another benefit of three-dimensional imaging is that once a volume of data has been acquired, it can be reviewed in its entirety at will, making this an ideal medium for reviewing paired examinations.
Using this technology and improved anatomic interpretation, we conducted a prospective study to assess the incidence of obstetric trauma to the external sphincter and related structures and to evaluate the functional consequences of disruption of each component.
Materials and Methods
During a 6-month period, primiparous women were recruited from the antenatal clinic of Northwick Park Hospital between 28 weeks' gestation and full term. Women with a history of previous anal surgery, anorectal dysfunction, or complicating medical condition that might have precluded spontaneous vaginal delivery were excluded.
The Harrow Research Ethics Committee approved the study, and written informed consent was obtained. Each woman had anal manometry, three-dimensional anal endosonography, and completed a questionnaire assessment of continence, before and 8 weeks after delivery. Obstetric management was not influenced by participation in the study.
Maximal resting and maximal voluntary squeeze pressures were measured with a hand-held system, using a 1-cm air-filled micro balloon (Stryker 295-1 pressure monitor, Stryker Corporation, Kalamazoo, MI) and station pull-through technique.1,15,16
Three-dimensional anal endosonography was performed using the B&K Sirius 3D system (B&K Medical, Sandhoften, Denmark). This system utilized a 10-MHz frequency rotating transducer to provide axial images that were stored continuously during mechanical withdrawal of the probe from the anal canal. Subjects were examined in the left lateral position. The prone position is technically better17 but was impossible in advanced pregnancy. To maintain constancy, the left lateral position was also used after delivery. The probe was inserted into the distal rectum and the assembly stabilized in a self-locking rig. Once the probe had been positioned, data acquisition was automatic, averaging 15 seconds to acquire a full dataset, with a similar time to render the volume and check satisfactory image acquisition.
Each subject completed a St Mark's continence questionnaire,18 which assesses independently the presence and frequency of incontinence to solid, liquid stool, or flatus, and includes an assessment of fecal urgency. A zero score denotes total continence; the maximum score of 25 denotes complete incontinence to solid stool on a daily basis.
Each predelivery and postdelivery pair of three-dimensional datasets was reviewed in the axial plane independently by two observers (ABW and CIB) who were blinded to the manometric and continence data. Tears were defined by the presence of scar tissue, presenting on ultrasound as a homogeneous area of low reflectivity crossing or deforming normal anatomic boundaries. Tears of the external sphincter were defined as scarring involving the sphincteric ring, of the puboanalis by scarring in the medial aspect of the puborectalis, and of the transverse perineii by scarring lateral to the external sphincter ring and posterior to the bulbospongiosus. (Puboanalis is formed by strands of striated muscle originating from the levator and inserting into the longitudinal layer of the anal canal medial to puborectalis.) Internal sphincter tears were defined by disruption of any part of the smooth muscle ring. Postdelivery scans were classified as intact or as having a defect affecting either the external sphincter or other perineal structures (either puboanalis or perineii muscles).
A consensus decision was reached when there was disagreement between the two observers. The point at which the external anal sphincter formed a complete ring was noted, and considered symmetric if this was between 11 and 1 o'clock.
The scans were then reviewed in the coronal plane to examine the anterior part of the external anal sphincter where it closes to form a complete ring. This area is of most interest in the context of postdelivery sphincter damage. Both observers reviewed the scans assessing whether trauma had been sustained to the external anal sphincter or any other perineal structure. These results were then correlated with the results of simple axial imaging.
The paired parametric manometry data were analyzed using paired t testing, with nonpaired t testing for comparison between delivery methods. The continence data were analyzed using Wilcoxon signed rank scoring for paired data and Mann Whitney U testing for non-paired data. Associations were assessed using Fisher exact testing. The assessment of agreement for reporting the scans was made using a weighted kappa statistic. Scans were classified in increasing severity with a normal scan being the least damage present, damage to either the transverse perineii or the puboanalis being intermediate, and external anal sphincter trauma being the most severe damage. The confidence intervals for the incidence of sphincter trauma were calculated using Clopper-Pearson test for proportions and a 95% confidence interval (CI).
Of the 350 nulliparous pregnant women invited to take part in the study, 71 were recruited and 55 returned for follow-up. The data presented are from the 55 women for whom paired data were available. Their median age was 30 years (range 18–47 years), and they were scanned at a median gestation of 33 weeks (range 23–42 weeks) and 10 weeks after delivery (range 7–22). The median gestation was 40 weeks (range 34–42).
Forty-five had a vaginal delivery, which was unassisted in 31 and instrument assisted in 14 (seven forceps and seven Ventouse vacuum extractions). The overall episiotomy rate was 29%; 57% in the forceps delivery group, 71% with Ventouse extraction, and 23% without instrumental assistance. All episiotomies were mediolateral in orientation. Cesarean delivery was done in ten for failure to progress or fetal distress during established labor.
In the group delivered vaginally, the mean predelivery resting pressure was 91 mmHg (standard deviation [SD] 13 mmHg), which decreased to 85 mmHg (SD 22 mmHg) after delivery; the difference was not significant (P = .058). The decrease in the mean squeeze pressure from 157 mmHg (SD 22 mmHg) to 130 mmHg (SD 33 mmHg) after delivery was significant (P < .001) (Figure 1). The incontinence score, irrespective of delivery mode, increased significantly from a median of 0 (range 0–9) to a median of 2 (range 0–9) (P = .001). These changes were most marked with forceps-assisted delivery. Women who had an emergency cesarean showed no change in manometric pressures or continence score.
When the axial scans were reviewed independently, there was initial disagreement in seven cases, three of which concerned possible external anal sphincter trauma. In all cases a consensus diagnosis was reached, and the weighted Kappa analysis for agreement was 91%, with a Kappa of 0.67 (95% CI 0.45, 0.89) indicating good agreement.
Only one woman had a clinically evident third-degree tear at time of delivery. In the 45 women delivered vaginally, 13 (29%, CI 16%, 44%) showed evidence of trauma on endosonography. The external sphincter was involved in five (11%, CI 4%, 24%) (Figure 2), the puboanalis in nine (20%, CI 10%, 35%) (Figure 3), and the transverse perineii in three (7%, CI 1%, 18%) (Figure 4). In four women, more than one structure was involved: two had external sphincter and puboanalis tears, one an external sphincter and transverse perineii tear, and one a puboanalis with a transverse perineii tear. No woman had an internal anal sphincter tear, and no tear occurred in the group that had a cesarean delivery. Imaging in the coronal plane in prenatal women clearly showed the symmetric anterior closure of the external anal sphincter (Figure 2). Loss of this symmetry after delivery often alerted the physician to the possibility of a sphincter defect. Symmetry of closure had a negative predictive value for an external sphincter tear of 95%. Asymmetric closure was strongly associated with a tear in the external sphincter or puboanalis (P < .02). The finding was confirmed on coronal imaging in all cases where sphincter trauma was reported on axial imaging. Furthermore, no additional sphincter defect was detected on coronal imaging that was not seen on axial scanning.
In all women, squeeze pressure was lower postpartum. The presence of an external anal sphincter defect after delivery was associated with a significantly greater reduction in squeeze pressure from a mean of 163 mmHg (SD 23 mmHg) to 105 mmHg (SD 20 mmHg) compared with those without any tear (155 mmHg [SD 21 mmHg] decreasing to 135 mmHg [SD 28 mmHg]; P = .035). A tear to puboanalis or transverse perineii alone was not associated with any additional decrease in pressure compared with women with a normal scan (mean 160 mmHg [SD 31 mmHg] decreasing to 133 mmHg [SD 24 mmHg]; not significant) (Figure 5).
The sphincter trauma group also had a significant increase in incontinence score compared with women with a normal endosonographic postdelivery examination (median 1 [range 0–2] increasing to 4 [0–9] in the sphincter tear group compared with a median of 1 [0–9] increasing to 1 [0–8] in the normal group; P = .02). Tears of the puboanalis or transverse perineii, either alone (n = 6 and 1, respectively) or in combination (n = 1) were associated with a median incontinence score predelivery of 0 (0–4) increasing to 2 (0–8) after delivery. This change was not statistically different from that in women with a normal postdelivery scan.
The overall episiotomy rate was 29% (23% in women who had a normal vaginal delivery and 64% for the instrument-assisted group). There was no association between episiotomy and the incidence of external sphincter or other tears. The episiotomy rate was 33% in women who had sphincter trauma and 29% in women who had intact sphincters after delivery.
We found an incidence of external sphincter trauma after vaginal delivery of 11%, which is much lower than in other studies except that of Varma et al,5 who reported 6.8% for primiparous deliveries. However, this should be modified to 11.5% to exclude those who had a cesarean delivery.5 The overall incidence of trauma to the perineum and sphincter in this study was 29%, as 18% had tears to the puboanalis or transverse perineii. The previous comparative studies were therefore not strictly inaccurate, as there is about a 30% incidence of tears. With hindsight, however, imprecise anatomic localization of tears at the time meant that these were all considered to involve the external sphincter, whereas probably less than half actually did so.
This study highlights the differentiation of external sphincter trauma from tears to other structures and explains much of the discrepancy between the incidence of incontinence after vaginal delivery and the previously reported incidence of trauma. Damage to the external anal sphincter was associated with a significantly greater reduction in squeeze pressure and increase in incontinence score, compared with intact sphincter on endosonography, whereas damage to the puboanalis or transverse perineii had no detectable functional significance. Although the total incidence of trauma to the perineum was 29%, external sphincter trauma had a functional consequence in only 11% of vaginal deliveries. This is close to the 13% reported to have some degree of postdelivery incontinence.1
The three-dimensional ultrasound system described has many advantages. The datasets were reviewed initially in the axial plane to make this study comparable to routine scanning. The ability to review the entire study in detail by using the three-dimensional volume as a cine loop and to repeatedly investigate suspicious areas proved helpful in making comparisons and arriving at a consensus. Anomalous insertions of the external sphincter have been reported in up to 14%,19 and these anomalies could be investigated in detail on both the predelivery and postdelivery datasets, to avoid confusing an anomaly with a tear.
The ability to image the sphincter in the coronal plane was a useful adjunct for confirming sphincter trauma, but it did not show any trauma not already diagnosed using the conventional axial imaging plane. There are two disadvantages to coronal imaging in endosonography. First, because the image slices are flat while the sphincter is cylindrical, to show the full extent of a tear often requires oblique coronal slicing (Figure 6). Second, sphincter symmetry is important in the diagnosis of trauma, and even slight rotation of the axial image can result in misleading coronal images. It is therefore essential to ensure that the axial images are correctly orientated before imaging in the coronal plane. Lack of symmetry of closure of the external anal sphincter can result from a tear in the puboanalis or external sphincter and, although not a specific finding, suggests the need for close scrutiny to exclude a tear.
Endocoil magnetic resonance imaging has enabled aspects of perineal ultrasound anatomy to be clarified.20 The most common structure torn during vaginal delivery is the puboanalis, which is a slip of striated muscle that arises from the medial border of the puborectalis and inserts into the longitudinal layer of the anal canal.21–24
1. Sultan AH, Kamm MA, Hudson CN, Thomas JM, Bartram CI. Anal-sphincter disruption during vaginal delivery. N Engl J Med 1993;329:1905–11.
2. Rieger N, Schloithe A, Saccone G, Wattchow D. A prospective study of anal sphincter injury due to childbirth. Scand J Gastroenterol 1998;33:950–5.
3. Donnelly V, Fynes M, Campbell D, Johnson H, O'Connell PR, O'Herlihy C. Obstetric events leading to anal sphincter damage. Obstet Gynecol 1998;92:955–61.
4. Zetterstrom J, Mellgren A, Jensen LL, Wong WD, Kim DG, Lowry AC, et al. Effect of delivery on anal sphincter morphology and function. Dis Colon Rectum 1999;42:1253–60.
5. Varma A, Gunn J, Gardiner A, Lindow SW, Duthie GS. Obstetric anal sphincter injury: Prospective evaluation of incidence. Dis Colon Rectum 1999;42:1537–43.
6. Sultan AH, Kamm MA, Hudson CN, Bartram CI. Third degree obstetric anal sphincter tears: Risk factors and outcome of primary repair. BMJ 1994;308:887–91.
7. Walsh CJ, Mooney EF, Upton GJ, Motson RW. Incidence of third-degree perineal tears in labour and outcome after primary repair. Br J Surg 1996;83:218–21.
8. Donnelly VS, Fynes M, O'Connell PR, O'Herlihy C. The influence of oestrogen replacement on faecal incontinence in post menopausal women. Br J Obstet Gynaecol 1997;104:311–5.
9. Felt-Bersma RJ, Cuesta MA, Koorevaar M. Anal sphincter repair improves anorectal function and endosonographic image. A prospective clinical study. Dis Colon Rectum 1996;39:878–85.
10. Gold DM, Bartram CI, Halligan S, Humphries KN, Kamm MA, Kmiot WA. Three-dimensional endoanal sonography in assessing anal canal injury. Br J Surg 1999;86:365–70.
11. deSouza NM, Puni R, Kmiot WA, Bartram CI, Hall AS, Bydder GM. MRI of the anal sphincter. J Comput Assist Tomogr 1995;19:745–51.
12. Hussain SM, Stoker J, Lameris JS. Anal sphincter complex: Endoanal MR imaging of normal anatomy. Radiology 1995;197:671–7.
13. Hussain SM, Stoker J, Zwamborn AW, Den Hollander JC, Kuiper JW, Entius CA, et al. Endoanal MRI of the anal sphincter complex: Correlation with cross-sectional anatomy and histology. J Anat 1996;189:677–82.
14. Rociu E, Stoker J, Zwamborn AW, Lameris JS. Endoanal MR imaging of the anal sphincter in fecal incontinence. Radiographics 1999;19:S171–7.
15. Orrom WJ, Williams JG, Rothenberger DA, Wong WD. Portable anorectal manometry. Br J Surg 1990;77:876–7.
16. Taylor BM, Beart RW Jr, Phillips SF. Longitudinal and radial variations of pressure in the human anal sphincter. Gastroenterology 1984;86:693–7.
17. Frudinger A, Bartram CI, Halligan S, Kamm M. Examination techniques for endosonography of the anal canal. Abdominal Imaging 1998;23:301–3.
18. Vaizey CJ, Carapeti E, Cahill JA, Kamm MA. Prospective comparison of faecal incontinence grading systems. Gut 1999;44:77–80.
19. Sultan AH, Kamm MA, Hudson CN, Nicholls JR, Bartram CI. Endosonography of the anal sphincters: Normal anatomy and comparison with manometry. Clin Radiol 1994;49:368–74.
20. Tan IL, Stoker J, Zwamborn AW, Entius CA, Calame J, Lameris JS. Female pelvic floor: Endovaginal MR imaging of normal anatomy. Radiology 1998;206:777–83.
21. Lawson JO. Pelvic anatomy. II. Anal canal and associated sphincters. Ann R Coll Surg Engl 1974;54:287–300.
22. Gorsch RV. The sigmoid, rectum, and anal canal. Relations, attachments and pelvic spaces. Clin Symptoms 1960;12:35–61.
23. Lawson JO. Pelvic anatomy. I. Pelvic floor muscles. Ann R Coll Surg Engl 1974;54:244–52.
24. Eckardt VF, Jung B, Fischer B, Lierse W. Anal endosonography in healthy subjects and patients with idiopathic fecal incontinence. Dis Colon Rectum 1994;37:235–42.