Perineal Assessment and Repair Longitudinal Study (PEARLS): a matched-pair cluster randomized trial

Table 6 Assessing sustainability: comparison of group A clusters in phase 3 with group B (late intervention) clusters in phase 2

	Mean Difference (95% CI)^a,b	P-value^c
Sutures removed, %	1.9% (−0.8% to 8.1%)	0.18
Perineal wound infection since birth, %	1.0% (−4.0% to 6.1%)	0.66
Use of continuous non-locking suturing technique for vaginal wall, %	−3.5% (−17.2% to 10.1%)	0.57
Use of continuous non-locking suturing technique for muscle layer, %	5.1% (−9.3% to 19.5%)	0.44
Use of subcuticular suturing technique for perineal skin, %	−5.8% (−17.8% to 6.1%)	0.30
Use of evidence-based management technique for all layers, %^d	1.7% (−16.0% to 19.4%)	0.84
Use of rapidly absorbable polyglactin suture, %	−9.4% (−32.4% to 13.7%)	0.38
Women who received post-natal leaflet, %	−34.5% (−54.2% to −14.8%)	0.003

^aMean difference = mean in B clusters – mean in A clusters. Negative mean differences indicate values were on average higher in the group A (early intervention) clusters.
^bOne cluster had no data at phase 3, and so this cluster and its pair were excluded (9 degrees of freedom).
^cPaired t-test.
^dFor this variable one additional cluster had no data, and so this cluster and its pair were excluded (8 degrees of freedom).

ISSN: 1741-7015