Grading of obstetric anal sphincter injury: endoanal or transperineal ultrasound?

Introduction

Obstetric anal sphincter injury (OASI) occurs in up to 5% of vaginal deliveries and is associated with significant maternal morbidity.^1,2 OASI is the most common cause of anal incontinence in women of reproductive age.³ Anal incontinence can have a devastating impact on a woman’s mental and physical health, as well as their social wellbeing, significantly impairing quality of life and potentially leading to social exclusion and loss of self-confidence.⁴

Accurate diagnosis and appropriate management of sphincter injuries is crucial to reduce anal incontinence.⁵ 3D endoanal ultrasound (EAUS) is considered the gold standard for imaging and assessing pathology of the anal sphincter.^6,7 EAUS has been validated using histology as the reference standard and findings have been shown to correspond with external and internal anal sphincter defects.^7–9 However, EAUS requires specialised equipment and trained operators, and is not widely available in obstetric units. It is also relatively invasive and may cause patient discomfort.¹⁰

3D/4D transperineal ultrasound (TPUS) is an alternative imaging modality that can assess the anal sphincter complex.¹¹ TPUS is considered less invasive, less expensive, and more widely available in obstetric units compared to EAUS.^10,12 As TPUS probes are placed externally on the perineum, they allow visualisation of the anal sphincter in an undisturbed state, unlike EAUS probes which may distort the anatomy when inserted into the anal canal.^13,14 However, the diagnostic accuracy of TPUS compared to EAUS for detecting anal sphincter defects after OASI is not well established.^10,15

The primary aim of this study was to evaluate the agreement between EAUS and TPUS in detecting anal sphincter defects in women after primary repair of OASI. Secondary aims were to compare the diagnostic accuracy of these modalities and to assess the correlation between ultrasound findings and the severity of anal incontinence symptoms.

Methods

This was a prospective cohort study conducted at a single tertiary perineal tear clinic from 2019–2021 in Sydney, Australia on primiparous women following primary surgical repair of OASI.

All patients underwent clinical examination for grading of OASI according to the Sultan classification¹⁶ shortly after childbirth. Grading was conducted jointly between the accoucheur and a second clinician observing during a routine genito-anal examination as per hospital guidelines. OASI repairs were performed by credentialed registrars, fellows, and consultants. After discharge from hospital, women referred to the Perineal Tear clinic underwent comprehensive evaluation at 6 weeks postpartum. Anal incontinence was recorded using the validated St. Mark’s incontinence score (SMIS), with 0 being no bowel incontinence and 24 being total incontinence.^17,18 The Pelvic Organ Prolapse Quantification (POP-Q) score was documented on clinical examination. All women saw a pelvic floor physiotherapist for review and management. At 6 months postpartum women underwent anal manometry, pudendal nerve terminal motor latency, EAUS, transvaginal ultrasound and for the purposes of the study hypothesis, TPUS. Women gave consent for the extra assessment, and ethics approval was obtained (HREC/10STG/81). EAUS was performed by a single experienced sonographer (Author CP) and TPUS was performed by an experienced urogynaecology trainee previously trained in the use of this modality. Both sonographers were blinded to each other’s findings and to the clinical findings. Each sonographer made their own assessment of their findings. Prior to ultrasound analysis, a test-retest series was carried out to evaluate inter-observer agreement for anal sphincter defects on TPUS, which demonstrated reliability (Cohen’s kappa 0.92).

Patients were asked to empty their bladder before a pelvic examination was performed in the supine position with hips and knees flexed. A clinical examination included a POP-Q score and assessment of the pelvic floor muscle (PFM) on contraction for assessment of PFM strength, using the modified Oxford score. This was followed by TPUS using GE 730 Expert system with RAB 4-8Mz transducer. For anal sphincter assessment, the transducer was placed on the perineum in the coronal plane at a 45-degree angle to rectum. Anal sphincter volumes were obtained on maximum PFM contraction, with the entire length of the anal canal contained in the volume, and with high harmonic settings and one focal zone placed at minimum depth. The resulting images were evaluated using tomographic ultrasound imaging, which produces eight slices evenly divided along the entire length of the anal canal from the puborectalis muscle (slice 1) to the level of the subcutaneous external anal sphincter (EAS) (slice 8); six slices (slices 2–7) are used to demonstrate the entire EAS and internal anal sphincter (IAS).¹¹ A 3a tear was diagnosed if less than 4 of the 6 slices showed abnormality of the EAS only, 3b if the EAS was abnormal in greater than or equal to 4 of the 6 slices, and 3c/4th degree tears if both the EAS and IAS were abnormal in greater than or equal to 4 of the 6 slices.¹⁸ A ‘residual defect’ was defined as a disruption ≥30 degrees visualised in at least 4 out of the 6 slices from slices 2–7.^19,20

EAUS was performed using Bk 3000, 3D 20R3 (9052) anorectal transducer with a 7–10-MHz, 360° rotating endoprobe (BK Medical, Herlev, Denmark). The women were in the left lateral position with flexed knees and hips. Serial images were obtained at the level of the puborectalis muscle (upper anal canal) and the subcutaneous level of the anal canal (mid anal canal and lower anal canal).⁷

The IAS was identified as a continuous hypoechoic ring, whereas the EAS was defined as a mixed echogenic ring. Any remaining sphincter defect was defined as disruption to IAS or EAS on ultrasound.^7,21 The presence or absence of a tear in the IAS or EAS was recorded. The Starck score was used to quantify the residual sphincter damage after primary repair. The Starck score is based upon the length of defect, depth of defect and degree (angle) of defect, with a total score of 0 being no defect and 16 being complete sphincter defect.^21,22

Statistical analysis utilised weighted kappa statistics to evaluate the agreement between clinical grading and ultrasound-based grading according to EAUS and TPUS. The association between anal incontinence symptom severity and ultrasound findings was assessed using parametric and non-parametric tests as appropriate.

Results

A total of 52 primiparous women were referred for evaluation after primary repair of OASI during the study period. Twenty-five (48%) were excluded due to missing data, predominantly from not attending their scheduled clinic visit (21/25). The remaining 27 patients with complete datasets were analysed at a median follow-up of 6 months (range 5–7 months) after OASI. The mean patient age was 31 years (range 26–41 years).

The grade of perineal injury in the 27 patients is described in Table 1. Clinical grading refers to the Sultan score determined post birth by the accoucheur. Overall, the distribution of degree of tear was relatively similar across the three assessment methods. Grade 3a was the most common classification for all three methods, followed by grade 3b. Grades 3c and 4 were less frequently observed. TPUS and EAUS identified one case as grade 2.

One (3.7%) patient graded clinically as a 3a tear was found to have an intact anal sphincter on both TPUS and EAUS imaging. Full agreement between clinical and TPUS grading was noted in 17 (63.0%) women, with a weighted κ of 0.594 (P = 0.002). Full agreement with EAUS grading occurred in 19 (70.4%) cases, weighted κ = 0.713 (P < 0.001).

Complete concordance between TPUS and EAUS grading was seen in 16 (59.3%) cases, with a weighted κ of 0.593 (P = 0.002). In 9 (33.3%) women there was disagreement by one category between EAUS and TPUS grading, but no disagreement by two or more categories.

Overall, potential over-diagnosis relative to clinical exam was noted in 4 (14.8%) cases based on TPUS and in 2 (7.4%) cases based on EAUS. Potential under-diagnosis occurred in 6 (22.2%) cases by TPUS and 6 (22.2%) cases by EAUS.

The association between SMIS and TPUS grading did not reach statistical significance (P = 0.011) at the 6-month median follow-up. Similarly, no significant correlation was found between incontinence symptoms and EAUS grading (P = 0.011) or Starck score (P = 0.96). There was also no significant association between residual defect on TPUS and anal incontinence symptoms (P = 0.88).

Discussion

This study compared the diagnostic accuracy of TPUS to EAUS for evaluating OASI after primary repair. Our findings demonstrate moderate agreement between TPUS and EAUS for grading OASI severity, suggesting TPUS may be a viable alternative to EAUS in certain clinical contexts.

The moderate agreement between clinical examination and both ultrasound modalities (weighted κ = 0.594 for TPUS, 0.713 for EAUS) is consistent with previous research showing discrepancies between clinical and imaging-based assessments of OASI.^23,24 This highlights the potential value of ultrasound imaging in providing additional information beyond clinical examination alone.

Importantly, we found moderate agreement between TPUS and EAUS grading (weighted κ = 0.593), indicating TPUS may offer comparable diagnostic accuracy to the current gold standard. This aligns with findings from Taithongchai et al., who reported fair ability of TPUS to diagnose anal sphincter defects (AUC 0.70–0.74) compared to EAUS.²⁵ However, our study showed higher agreement than previous comparisons, potentially due to advancements in ultrasound technology and operator experience.

The potential for over- and under-diagnosis with TPUS compared to clinical examination (14.8% and 22.2% respectively) is an important consideration. Okeahialam et al. found that using a 30° defect angle cut-off on TPUS could lead to incorrect diagnosis of significant EAS defects in 9–36% of women and IAS defects in 4–15% of women.¹³ This underscores the need for caution in interpreting TPUS findings and suggests further research is needed to optimise diagnostic criteria.

One limitation of TPUS noted in our study was its inability to distinguish between 3c and 4th degree tears, which EAUS could differentiate. While the clinical relevance of this distinction is debatable, it represents a potential drawback of TPUS of which clinicians should be aware.

Interestingly, we found no significant association between ultrasound-based grading and anal incontinence symptoms at 6-month follow-up. This lack of correlation between imaging findings and functional outcomes has been reported in other studies^20,22 and may reflect the complex, multifactorial nature of post-OASI symptoms.

The advantages of TPUS, including its non-invasive nature, wider availability, and lower cost compared to EAUS, make it an attractive option for routine follow-up after OASI repair. Wong et al. demonstrated that TPUS could be successfully performed immediately postpartum, identifying 90.5% of clinically diagnosed OASIs.¹⁵ This suggests TPUS may have a role in both immediate postpartum assessment and longer-term follow-up.

However, the limitations of TPUS must be acknowledged. The potential for misdiagnosis, particularly when using current diagnostic criteria, means that TPUS results should be interpreted cautiously. In cases where precise delineation of sphincter anatomy is crucial for clinical decision-making, EAUS may still be preferable.

Our study has several limitations, including a relatively small sample size and single-centre design. It is hoped, however, that this limited data from a different centre may build on the already limited knowledge on this subject. The 6-month follow-up period may not have been sufficient to capture all long-term sequelae of OASI. Additionally, the exclusion of a significant number of patients due to missing data may have introduced selection bias.

In conclusion, our findings suggest that TPUS demonstrates promise as an alternative to EAUS for evaluating anal sphincter injuries after OASI repair. Its moderate agreement with EAUS, combined with practical advantages, makes it a potentially valuable tool in the follow-up of these patients. However, further research is needed to refine diagnostic criteria and establish optimal cut-off values for TPUS. Large, multi-centre studies with longer follow-up periods are warranted to fully elucidate the role of TPUS in post-OASI assessment and its relationship to long-term functional outcomes. Until then, clinicians should use TPUS as a complementary tool alongside clinical examination and, where available, EAUS in the comprehensive evaluation of women following OASI.