Colpoperineorrhaphy versus pelvic floor muscle therapy in women with improperly healed perineal tears – a randomized controlled trial

Bergman I1, Westergren Söderberg M1, Ek M1

Research Type


Abstract Category

Pelvic Organ Prolapse

Abstract 16
Urogynaecology 1 - Pelvic Organ Prolapse
Scientific Podium Short Oral Session 2
Wednesday 4th September 2019
10:07 - 10:15
Hall H2
Clinical Trial Pelvic Floor Surgery Physiotherapy Female
1.Department of Clinical Science and Education, Södersjukhuset

Ida Bergman



Hypothesis / aims of study
To evaluate subjective and objective outcomes following perineorrhaphy with distal posterior colporrhaphy compared to pelvic floor muscle therapy (PFMT) in women with improperly healed 2nd degree perineal tears.
Study design, materials and methods
A single-center randomized controlled trial with two parallel arms (allocation ratio 1:1). Women with symptomatic, improperly healed perineal lacerations, minimum 6 months after vaginal childbirth, with a detachment of musculus bulbocavernosus and/or musculus transversus perinei superficialis, were randomized into surgical treatment with perineorrhaphy and distal posterior colporrhaphy or conservative treatment with tutored PFMT. A video demonstrating the surgical procedure has been recorded in order to describe the technique. The surgical procedures were performed by five different urogynecologists and the tutored PFMT sessions were performed by two different physiotherapists at Södersjukhuset in Stockholm, Sweden. The follow-up evaluation was conducted by an independent physician (not the surgeons). 

The primary outcome was subjective improvement assessed by the Patient Global Impression of Improvement scale (PGI-1) at 6 months. Secondary outcomes included Pelvic Floor Distress Inventory (PFDI-20) and Prolapse Incontinence Sexual Questionnaire (PISQ-12). Objective outcome measures included the pelvic organ prolapse quantification system (POP-Q) and evaluation of the perineal body thickness using both ultrasound and bidigital palpation. The perineal body was measured in a sagittal plane with a two-dimensional abdominal probe. Surgical characteristics and adverse events during follow-up were registered. 

On the basis of the assumption that 60% of women in the perineorrhaphy group and 20% in the PFMT group would report subjective improvement according to the definition in the PGI-1 scale, we calculated that 28 patients were needed in each group to achieve a power of 80% (alfa-level 0.05). Anticipating a 20% loss to follow-up, we planned to include 70 women.

Outcomes are analyzed on an intention-to-treat bases. Categorical endpoints were analyzed with the use of Fisher's exact test and the Mann-Whitney U-test was used to compare continuous data between groups. Changes in questionnaire scores within groups were analyzed with the Wilcoxon signed rank test. Continuous endpoints are presented as means ± standard deviations and binary outcomes as frequencies.
A total of 70 women were randomized, 35 into surgery and 35 into PFMT. There was one crossover in each group. Three patients in the perineorrhaphy group declined participation post randomization (1.4 %). There were no significant differences in baseline characteristics or questionnaire scores at the time of recruitment. The mean age of the study participants was 36±7 years, mean BMI 23±4 and mean parity 2±1. The mean duration postpartum from the last delivery was 2.9±0.8 years.

In an intention-to-treat analysis, where missing data was imputed conservatively (worst case outcome), 71.4% (25/35) in the perineorrhaphy group and 11.4 % (4/35) in the PFMT group were defined as cured at 6 months according to the definition in the PGI-1 scale, p <0.001. 

The mean PFDI-20 score improved significantly only in the perineorrhaphy group were the total score dropped from 83±45 to 45±47 (p<0.001) compared to the PFMT group were the drop was from 94±50 to 82±53 (NS). The mean PISQ-12 score improved significantly in both groups; mean scores in perineorrhaphy group improved from 31±4 to 37±5 points (p<0.001) and in the PFMT group from 32±7 to 35±6 points (p = 0.02). The improvement in PFDI-20 domains and PISQ-12 scores was significantly higher in the perineorrhaphy group compared to the PFMT group (p <0.001 and p=0.01, respectively). 

The thickness of the perineal body on ultrasound at recruitment was 7.9±2.4 mm in the perineorrhaphy group and 8.8±2.2 mm in the PFMT group (non-significant difference). The measurements at 6 months were 15.2±2.7 mm (perineorrhaphy group) and 9.0±2.4 mm (PFMT group). The improvement in perineal body thickness was significant only in the perineorrhaphy group.

The mean operation time in the surgery group was 35±12 minutes and the mean blood loss was 23±15 ml. De novo dyspareunia was not reported by any patient in the perineorrhaphy group at 6 months. A total of 10 patients reported resolution of painful intercourse after surgery. Three patients had to undergo re-operation due to bleeding from the surgical site postoperatively. One of the patients had a blood loss requiring blood transfusion. Three patients were diagnosed with a postoperative, surgical site infection and were successfully treated with per oral antibiotics.
Interpretation of results
Pelvic floor muscle therapy does not seem to improve pelvic floor dysfunction symptoms in women with an improperly healed 2nd degree perineal tear but results in slightly improved sexual functions. Perineorrhaphy with distal compartment repair resulted in a higher rate of successful treatment with improvement of pelvic floor dysfunction symptoms as well as sexual functions.
Concluding message
For women with symptomatic, defectively healed 2nd degree perineal body injuries following childbirth, surgical treatment, as compared with physiotherapy, results in higher short-term rates of subjective and objective improvement but comes with a higher rate of adverse events.
<span class="text-strong">Funding</span> Tore Nilsson Foundation <span class="text-strong">Clinical Trial</span> Yes <span class="text-strong">Registration Number</span> Identifier: NCT02545218 <span class="text-strong">RCT</span> Yes <span class="text-strong">Subjects</span> Human <span class="text-strong">Ethics Committee</span> Research Ethics Committee at Karolinska Institutet, Stockholm, Sweden. <span class="text-strong">Helsinki</span> Yes <span class="text-strong">Informed Consent</span> Yes