Vacuum extraction and pelvic floor injury : a randomized controlled trial and cohort studies

Abstract

BACKGROUND AND AIMS: Pelvic floor dysfunction (PFD) affects many women, especially after complicated childbirth. The impact of a prolonged second stage on PFD is not clear, while the effects of operative vaginal birth (OVB) and obstetric anal sphincter injury (OASI) are better known, especially concerning the risk of anal incontinence (AI). Vacuum extraction (VE) and primiparity are the major risk factors of OASI. Observational studies report that OASI in VE in nulliparous women decreases if an episiotomy is performed, but this has not been confirmed in a randomized controlled trial (RCT). The aim of this thesis was to investigate factors that could improve second stage management, particularly in VE, to decrease the risk of OASI and future PFD. First, we aimed to identify risk factors for OASI in VE. Second, we aimed to explore effects of a prolonged second stage on PFD, and finally, we aimed to investigate if a lateral episiotomy, compared with no episiotomy, reduces the risk of OASI in VE in nulliparous women.

METHODS AND MAIN RESULTS: Study I is a retrospective cohort study with one-year medical records data from Danderyd Hospital. All primiparous women with a live, singleton fetus ≥34 gestational weeks, delivered by VE during 2013 were included (n=323). Primary outcome was OASI and exposure was operator category (obstetrician, gynecologist, and resident) with obstetrician as reference. OASI occurred in 57 (17.6%) women. Fifteen (11.5%) OASI occurred in VE performed by obstetricians, 10 (13.5%) by gynecologists (aOR 1.84, 95% CI 0.72-4.70), and 32 (26.9%) by residents (aOR 5.13, 95% CI 2.20-11.95). Study II and III are questionnaire and cohort studies including primiparous women with a live, single fetus in cephalic presentation ≥37 gestational weeks and second stage duration ≥3 h in the Stockholm Region during one year (2019, n=1302). Data were retrieved from electronic medical records (EMR). The oneyear follow-up questionnaire from the Swedish Perineal Laceration Register, including questions regarding symptoms of urinary incontinence (UI), anal incontinence (AI), and pelvic organ prolapse (POP) were distributed one to two years after delivery. In Study II, primary outcome was AI defined as Wexner score ≥2. Main exposure was mode of delivery with cesarean section (CS) as reference. Secondary exposures were degree of perineal injury and extended second stage duration. We found that the odds of AI were increased by VE (aOR 2.25, 95% CI 1.21-4.18) but not by spontaneous vaginal birth (SVB) (aOR 1.55, 95% CI 0.85-2.84). AI was also increased by OASI (aOR 2.03, 95% CI 1.17-3.52) and second-degree perineal injuries (aOR 1.36, 95% CI 1.03-1.81). OASI and VE combined inferred the highest odds (aOR 4.06, 95% CI 1.80-9.14) compared with CS. Extended duration of the prolonged second stage did not affect the risk of AI. In Study III, primary outcome was a composite of PFD including at least weekly symptoms of UI, AI, and POP. Exposure was intervention with VE or CS at 3-4 h or at 4-5 h respectively, compared with expectant management. The risk of PFD was increased after VE at 3-4 h (aRR 1.33, 95% CI 1.06-1.65) and 4-5 h (aRR 1.34, 95% CI 1.05-1.70), but remained unchanged after CS. The increased risk after VE was not mediated by OASI. Study IV is a multicenter RCT of lateral episiotomy compared with no episiotomy in nulliparous women with a live, single, fetus ≥34 gestational weeks, requiring VE during 2017-2023. The intervention was a lateral episiotomy. Primary outcome was OASI. The modified intention-to-treat (mITT) population included women with attempted or successful VE (n=702). In the intervention group, 21/344 (6.1%) women sustained OASI compared with 47/358 (13.1%) in the comparison group (p=0.002). The risk difference was -7.0% (96% CI -11.7% to - 2.5%). The unadjusted risk ratio was 0.46 (96% CI 0.28-0.78) and 0.47 (96% CI 0.23-0.97) adjusted for site. Number needed to treat was 14.3 to avoid one OASI. Wound infection and dehiscence were significantly increased in the intervention group, while all other outcomes were similar.

CONCLUSION: In VE in nulliparous women, the risk of OASI increased when the operator was a resident, indicating a need for increased training and supervision. In a prolonged second stage, an extended duration did not increase the risk of AI or PFD. If a SVB seems likely, it is better to wait than to intervene with VE in order to avoid future PFD. If intervention is necessary, and several risk factors for OASI are present, CS could be preferable to reduce the risk of AI. In VE in nulliparous women, a lateral episiotomy significantly reduces the risk of OASI. However, the intervention may increase the risk of wound infection and dehiscence.