Interpretation
The rate of OASIS in primiparous women in the United Kingdom is 6.1% (19), which was similar to that detected by the accoucher. However, when women were re-examined by the trained research fellow, the incidence of OASIs increased from 6% to 9%. Previous studies have also shown that OASIs are missed by doctors and midwives (6, 13, 14). Andrews et al performed a prospective study of 254 primiparous women immediately after delivery, and found that the prevalence of OASIs increased from 11% to 24.5% when women were re-examined by a trained clinical research fellow (13). Of these, 87% of OASIs were missed by midwives compared to 24% by doctors. Similarly, Groom et al independently re-assessed 121 women’s perineum following a vaginal delivery and found that 40% of OASIs were being missed by accouchers (6). Both of these studies however are more than 15 year old and subsequently there has been an increase both in awareness and training in the diagnosis and management of OASIs (16). Hands on training courses in the management and diagnosis of OASIs are now mandatory for trainees in Obstetrics and Gynaecology the UK and have been shown to improve both the knowledge of perineal anatomy and techniques of identifying (15-17). Most missed OASIs however appear to be by midwives. In the UK there is no standardised assessment for the midwives in the diagnosis and repair of perineal tears. For obstetricians the Royal College of Obstetricians and Gynaecologists have developed objective structured assessment of surgical and technical skills for perineal tears. These are mandatory for obstetric trainees, and consideration should be given to implementing these for midwives as well.
This is the first study where TPUS was performed immediately after delivery and prior to suturing. 90.5% of anal sphincter defects seen clinically were confirmed on ultrasound and 99% of the clinically intact anal sphincters appeared to have an intact anal sphincter on 3D TPUS. We found that TPUS had a low positive predictive value to diagnose a sphincter defect and a high negative predictive value to detect an intact anal sphincter. This is in keeping with the findings of a cross-sectional study of 250 women who sustained OASIs that were scanned at 6-12 weeks postpartum by TPUS, where they found that 3D TPUS had a low positive predictive value (0.37-0.63) and a high negative predictive value (0.85-0.95) when compared with EAUS. They found that 3D TPUS could identify an intact anal sphincter but only had a fair ability to diagnose an EAS and IAS defect (9). Although Taithongchai et al’s (9) study is the most adequately powered study to date to compare the TPUS, TVUS with EAUS, it must be acknowledged that their training is of high standard and may not be generalisable.
TPUS had also been used to evaluate the anal sphincter complex during the early postpartum period. García-Mejido et al found that of the 146 women where TPUS was performed within 48 hours postpartum, 12 OASIs were detected by clinical examination and all such injuries were confirmed by TPUS (20). These results, however, should be interpreted with caution due to the small number of women who sustained OASIs and the study was not adequately powered.
In our study there were two anal sphincter defects (1%) seen on ultrasound and classified as a second degree tear clinically. These may have been missed by the clinical research fellow. It may also represent a genuine “occult” sphincter injury. Genuine ‘occult’ anal sphincter injuries are defects that are only seen on ultrasound and not detected clinically (21). These injuries were previously believed to be “occult”, but subsequently have been shown to represent clinically missed OASIs (13). Our findings were consistent with another observational study where 1% of OASI were seen on ultrasound but were not clinically detectable (13).
We found two OASIs that were diagnosed clinically but not seen on TPUS. The plausible explanation of the discrepancy between the clinical and ultrasound findings might be because of the technical difficulties performing an ultrasound examination immediately postpartum. It can be challenging due to oedema, haematoma and suture material (22) which may impair the image quality of the EAS. The IAS, however, seems to be easier to interpret due to its hypoechoic nature. In our study eight (3.5%) ultrasound images were excluded due to quality of the images. This finding is consistent with another observational study by García-Mejido et al (23) who performed TPUS on women immediately postpartum and found that three anal sphincter defects (2%) were not diagnosed on clinical examination but were detected in the subsequent ultrasound scan. They excluded five (3%) images due to poor quality for analysis caused by the intense perineal oedema.
All OASIs in our study were classified according to the RCOG classification. Using this classification if less than 50% of the EAS is torn it is classified as a 3a tear, and as a 3b tear when there is more than 50% involvement. We found that there is poor correlation between the size of an EAS defect seen clinically, when compared with ultrasound findings using TPUS similar to that seen on EAUS (13).