To Wean or Not to Wean: A Randomized Controlled Trial of Pavlik Harness Weaning in Infantile Developmental Dysplasia of the Hip

Background: After successful treatment of developmental dysplasia of the hip (DDH) with a Pavlik harness, controversy exists about whether it’s best to continue harness use for an additional “weaning” period or simply terminate treatment. The purpose of this randomized controlled trial (RCT) was to compare radiographic outcomes between patients with stable hip dysplasia who were weaned (W) or not weaned (NW) after Pavlik harness treatment. Methods: This was a single-center RCT of infants with stable ultrasonographic dysplasia in whom a Pavlik harness was initiated at <3 months of age. After 23 hours/day of harness treatment and normalization of ultrasound indices, patients were randomized into W or NW groups. W patients continued harness use for an additional 4 to 6 weeks, progressively decreasing daily use. Primary outcome was acetabular index (AI) on AP radiographs of the pelvis at 6 months of age. Secondary outcomes were number of hips with AI ≥ 30 degrees at 6 months (our threshold for prescribing part-time abduction bracing for residual dysplasia) and AI at 1 year of age. Results: Seventy-two patients (36 W, 36 NW) met inclusion criteria. There were no differences between W and NW cohorts in age at Pavlik initiation or baseline alpha angle or percent femoral head coverage. At time of ultrasonographic normalization, there was no significant difference in mean time spent in the harness between W and NW cohorts (35.0 vs. 34.9 d, P =0.84). At 6 months, mean AI was significantly lower in the W cohort (24.8±3.9 deg.) compared with the NW cohort (26.9±3.4 deg.) ( P =0.02). Fewer braces were prescribed for residual dysplasia at 6 months in the W versus NW cohort (17% vs. 22%), but this was not statistically significant ( P =0.56). AI at 1 year of age was not significantly different between groups (W 24.6±3.4 deg.; NW 25.3±2.3 deg., P =0.84) but this likely was affected by the use of additional bracing. Conclusions: A Pavlik weaning protocol resulted in improved AI at 6 months of age versus immediate termination of treatment. At 1 year, AI was not significantly different between W and NW cohorts, likely due to the use of additional part-time bracing in those who were dysplastic at 6 months. Level of Evidence: Level I—randomized controlled trial.