DESIGN: Two-centre, randomised, controlled trial with concealed allocation, blinded assessors and intention-to-treat analysis.

PARTICIPANTS: Seventy-two adults who had undergone cardiac surgery via a median sternotomy were included.

INTERVENTION: Participants were randomly allocated to one of two groups at 4 (SD 1) days after surgery. The control group received the usual advice to restrict their upper limb use for 4 to 6 weeks (ie, restrictive sternal precautions). The experimental group received advice to use pain and discomfort as the safe limits for their upper limb use during daily activities (ie, less restrictive precautions) for the same period. Both groups received postoperative individualised education in hospital and via weekly telephone calls for 6 weeks.

OUTCOME MEASURES: The primary outcome was physical function assessed by the Short Physical Performance Battery. Secondary outcomes included upper limb function, pain, kinesophobia, and health-related quality of life. Outcomes were measured before hospital discharge and at 4 and 12 weeks postoperatively. Adherence to sternal precautions was recorded.

RESULTS: There were no statistically significant differences in physical function between the groups at 4 weeks (MD 1.0, 95% CI -0.2 to 2.3) and 12 weeks (MD 0.4, 95% CI -0.9 to 1.6) postoperatively. There were no statistically significant between-group differences in secondary outcomes.

CONCLUSION: Modified (ie, less restrictive) sternal precautions for people following cardiac surgery had similar effects on physical recovery, pain and health-related quality of life as usual restrictive sternal precautions. Similar outcomes can be anticipated regardless of whether people following cardiac surgery are managed with traditional or modified sternal precautions.

METHODS: A randomized trial was performed in a university hospital in Malaysia. Participants were nulliparas at term with unripe cervixes (Bishop Score ≤ 5) admitted for IoL who were randomized to digital or speculum-aided transcervical Foley catheter insertion in lithotomy position. Primary outcomes were insertion duration, pain score [11-point Visual Numerical Rating Scale (VNRS)], and failure. All primary outcomes were recorded after the first insertion.

RESULTS: Data from 86 participants were analysed. Insertion duration (with standard deviation) was 2.72 ± 1.85 vs. 2.25 ± 0.55 min p = 0.12, pain score (VNRS) median [interquartile range] 3.5 [2-5] vs. 3 [2-5] p = 0.72 and failure 2/42 (5%) vs. 0/44 (0%) p = 0.24 for digital vs speculum respectively. There was no significant difference found between the two groups for all three primary outcomes. Induction to delivery 30.7 ± 9.4 vs 29.6 ± 11.5 h p = 0.64, Cesarean section 25/60 (64%) vs 28/64 (60%) RR 0.9 95% CI p = 0.7 and maternal satisfaction VNRS score with the birth process 7 [IQR 6-8] vs 7 [7-8] p = 0.97 for digital vs. speculum arms respectively. Other labor, delivery and neonatal secondary outcomes were not significantly different.

CONCLUSION: Digital and speculum insertion in nulliparas with unripe cervixes had similar insertion performance. As digital insertion required less equipment and consumables, it could be the preferred insertion method for the equally adept and the insertion technique to train towards.

OBJECTIVES: To assess the effects on mother and baby of a policy of selective episiotomy ('only if needed') compared with a policy of routine episiotomy ('part of routine management') for vaginal births.

SEARCH METHODS: We searched Cochrane Pregnancy and Childbirth's Trials Register (14 September 2016) and reference lists of retrieved studies.

SELECTION CRITERIA: Randomised controlled trials (RCTs) comparing selective versus routine use of episiotomy, irrespective of parity, setting or surgical type of episiotomy. We included trials where either unassisted or assisted vaginal births were intended. Quasi-RCTs, trials using a cross-over design or those published in abstract form only were not eligible for inclusion in this review.

DATA COLLECTION AND ANALYSIS: Two authors independently screened studies, extracted data, and assessed risk of bias. A third author mediated where there was no clear consensus. We observed good practice for data analysis and interpretation where trialists were review authors. We used fixed-effect models unless heterogeneity precluded this, expressed results as risk ratios (RR) and 95% confidence intervals (CI), and assessed the certainty of the evidence using GRADE.

MAIN RESULTS: This updated review includes 12 studies (6177 women), 11 in women in labour for whom a vaginal birth was intended, and one in women where an assisted birth was anticipated. Two were trials each with more than 1000 women (Argentina and the UK), and the rest were smaller (from Canada, Germany, Spain, Ireland, Malaysia, Pakistan, Columbia and Saudi Arabia). Eight trials included primiparous women only, and four trials were in both primiparous and multiparous women. For risk of bias, allocation was adequately concealed and reported in nine trials; sequence generation random and adequately reported in three trials; blinding of outcomes adequate and reported in one trial, blinding of participants and personnel reported in one trial.For women where an unassisted vaginal birth was anticipated, a policy of selective episiotomy may result in 30% fewer women experiencing severe perineal/vaginal trauma (RR 0.70, 95% CI 0.52 to 0.94; 5375 women; eight RCTs; low-certainty evidence). We do not know if there is a difference for blood loss at delivery (an average of 27 mL less with selective episiotomy, 95% CI from 75 mL less to 20 mL more; two trials, 336 women, very low-certainty evidence). Both selective and routine episiotomy have little or no effect on infants with Apgar score less than seven at five minutes (four trials, no events; 3908 women, moderate-certainty evidence); and there may be little or no difference in perineal infection (RR 0.90, 95% CI 0.45 to 1.82, three trials, 1467 participants, low-certainty evidence).For pain, we do not know if selective episiotomy compared with routine results in fewer women with moderate or severe perineal pain (measured on a visual analogue scale) at three days postpartum (RR 0.71, 95% CI 0.48 to 1.05, one trial, 165 participants, very low-certainty evidence). There is probably little or no difference for long-term (six months or more) dyspareunia (RR1.14, 95% CI 0.84 to 1.53, three trials, 1107 participants, moderate-certainty evidence); and there may be little or no difference for long-term (six months or more) urinary incontinence (average RR 0.98, 95% CI 0.67 to 1.44, three trials, 1107 participants, low-certainty evidence). One trial reported genital prolapse at three years postpartum. There was no clear difference between the two groups (RR 0.30, 95% CI 0.06 to 1.41; 365 women; one trial, low certainty evidence). Other outcomes relating to long-term effects were not reported (urinary fistula, rectal fistula, and faecal incontinence). Subgroup analyses by parity (primiparae versus multiparae) and by surgical method (midline versus mediolateral episiotomy) did not identify any modifying effects. Pain was not well assessed, and women's preferences were not reported.One trial examined selective episiotomy compared with routine episiotomy in women where an operative vaginal delivery was intended in 175 women, and did not show clear difference on severe perineal trauma between the restrictive and routine use of episiotomy, but the analysis was underpowered.