The prevalence of anal incontinence in this study was low, and none of the participants had high St. Mark’s scores. Nine women had previously undiagnosed obstetric injury to the anal sphincters. Anal incontinence was more common among the participants with undiagnosed injuries than those with intact anal sphincters, and the women with injuries also had reduced length of the perineal body and height of the anal sphincter complex.
Our findings confirm that not all OASIS are detected in the delivery ward.
Andrews et al found that 14% of women who deliver vaginally have unrecognized OASIS(11). Groom et al performed a study at a hospital in England, and found that the prevalence of OASIS increased from 2,5% before the study to 15% in a group of women who were re-examined by a specially trained research fellow(12). Our results are a bit lower than the findings of both of these studies from England. The discrepancy could be explained by different delivery techniques and programs for the training of health personell in detecting OASIS at the study hospitals, and in the difference in recruitment of study participants. Especially since our study included both primiparous women and multiparous women, while the other two studies included only primiparous women, which are known to have higher rates of OASIS.
The majority of the women found to have unrecognized injuries in our study had a St. Mark’s score of less than seven, and complete rupture of the external anal sphincter were only present in two women.
Most of the injuries were partial, and the torn part of the anal sphincters were located in the upper half of the muscles. This could complicate an accurate diagnosis in the delivery ward. Our findings imply that it will not be possible to detect all OASIS by the means of rectal exploration and visual inspection of the pelvic floor immediately postpartum. This is consistent with the findings of Frudinger et al (16), who concluded that sphincter injuries are difficult to detect by clinical examination.
However, at examination by a mean time of 14 months postpartum, we found that the length of the perineal body and the height of the anal sphincter complex upon rectal exploration were significantly shorter in the women with defects in the anal sphincters as compared to those with intact anal sphincters. This is consistent with the findings of Ozyurt et al (17). Even though these findings are based on comparision of mean values between women with and without OASIS verified on ultrasonography, obstetricians, gynecologists, general practitioners and physioterapists should be aware that such anatomical findings in parous women could indicate anal sphincter injury and women who experience symptoms should be examined with ultrasonography. Women who develops anal incontinence due to OASIS rarely address this problem when talking to doctors (5, 7, 18), and both gynecologists and other doctors should ask directly about fecal urgency, fecal leakage and involuntary leakage of flatus.
Involuntary leakage of gas was common among the participants found to have intact anal sphincters, but fecal urgency and fecal leakage were more common among the participants found to have unrecognized injuries to the anal sphincters. The participants with unrecognized injuries also had a significantly higher St. Mark’s score. Previous studies have shown that parous women who have sustained OASIS are at greater risk of developing anal incontinence than parous women who have not sustained OASIS. Several case-control studies have been conducted to compare the frequency and severity of anal incontinence among women who have sustained OASIS and women who have not.
A study by Cornelisse et al in the Netherlands found that 39% of women with OASIS suffered from anal incontinence four years postpartum, as compared to 20% of women who delivered vaginally without OASIS (19). Cornelisse et al found that most of the women suffered from leakage of flatus, 31% of the women with OASIS and 18% of the women without OASIS experienced this. Soiling were present in 12,1% in the OASIS group and 4,1% in the control group, and leakage of solid stools affected only 1,4% in the OASIS group and 1% of the controls (19).
Similarly, Pollack et al showed in a Swedish study that 42% of women with OASIS presented with leakage of flatus and 11% with fecal leakage by five years postpartum. They also had a control group of women who had delivered vaginally without sustaining OASIS, and found that 27% of controls had leakage of flatus and 5% had fecal leakage (4). In another Swedish case-control study by Wagenius et al, 33% of women with OASIS experienced leakage of flatus and 21% fecal leakage by four years postpartum, versus 15% leakage of flatus and 6% fecal leakage among women who delivered vaginally without sustaining OASIS (3).
In an American study by Evers et al women with OASIS were compared to women who had delivered vaginally without sustaining OASIS and women who delivered by cesarean section. They found a prevalence of 31% leakage of flatus, 21% fecal leakage among the women with OASIS by five to ten years postpartum. Among the controls leakage of flatus were present in 23% in the vaginal delivery group and 15% in the cesarean section group, and fecal leakage in 8% in both groups (1).
The same study by Evers et al found an odds ratio of 2.32 for anal incontinence among women with OASIS as compared to women who delivered by cesarean section. They also showed that the prevalence of anal incontinence and general quality of life were similar between women who delivered vaginally without sustaining OASIS and women who delivered by cesarean section (1).
All of these studies show a significant and strong association between postpartum anal incontinence and sustained OASIS at the index delivery. This is coherent with our findings, that women with OASIS are more likely to develop anal incontinence than women without OASIS.
The study has limitations. The model of recruitment, by which invitations were mailed to women without relationship to the hospital beyond having delivered there, may have contributed to non-response bias. Women who were contacted who did not have symptoms of anal incontinence may not find any personal gain in participating. Non-response bias is plausible, meaning that the true incidence of unrecognized obstetric anal sphincter injuries is probably lower than the 9.3% among our participants.
The main strength of the study is the combination of self-reported symptoms and clinical examination including endoanal sonography, performed at least four months postpartum (shortest time between delivery and clinical examination in this study was four months, whereas the mean time passed from delivery to examination was 14 months). Four months was set as a cut-off value for inclusion to assure the pelvic floor had had time for natural healing and are no longer swollen due to recent vaginal delivery. Findings of unrecognized OASIS were evaluated together with the symptoms presented by each affected woman. This made it possible to evaluate the degree to which women found to have unrecognized OASIS have symptoms of anal incontinence and are in need of treatment, or not. The endoanal sonography ensures a reliable diagnosis of anal sphincter injuries.