Preoperative Gallows traction as an adjunct to hip open reduction surgery: Is it safe and is it effective?

We read with interest the above publication by Uren et al. ¹ and present our uncertainty about the existence of the presented cohort and the claim that the study provides the clearest evidence of the beneficial effect of preoperative traction.

We discuss a series of publications^{2 –6} from the same unit, reflecting the evolvement of Professor Clarke’s hip dysplasia practice, which initially consisted of performing closed or open reductions (OR) as the primary procedure for failed Pavlik harness treatment or late presentations and secondary corrective surgeries (mainly pelvic osteotomies (PO)) at a later stage where necessary. ² Bolland et al. ² reported on 134 hips that had Gallows traction and OR between 1988 and 2003 (18 hips underwent a femoral shortening osteotomy), of which 4 required a revision OR and 24 a PO later. The mean acetabular index (AI) improved from 38° to 28° in the PO and 24° in the non-PO group, 14 hips had Kalamchi and MacEwen (KM) avascular necrosis (AVN) grades II–IV (6x II, 1x III, 7x IV) at a mean follow-up (FU) of 9 years. Severin types were not provided.

In 2005, Clarke et al. ³ reported on 22 hips that underwent an OR (FU ≥ 3 years), of which 9 required secondary procedures and 11 showed signs of AVN (KM: 8x I, 2x III, 1x IV), with the following Severin Grades: 15x I, 4x II, 2x III, and 1x not gradable.

The treatment strategy presented by Bolland et al. ² and Clarke et al. ³ was replaced by combining open reduction, capsulorraphy, and acetabuloplasty (ORCA) after traction in September 2004 ⁴ because of the residual dysplasia seen with traction and OR only. In 2014, Carsi et al. ⁴ published the early ORCA results performed on 55 hips between September 2004 and February 2011 (mean age: 13 months) and a FU of 2–8 years. The mean AI improved from 38° to 21° at the final FU. Severin types were not provided. Eight patients developed KM grades of II–IV (5x II, 1x III, and 2x IV), and two patients underwent secondary procedures. In 2015, Carsi and Clarke ⁵ published a comparison between traction and OR (historic group of 27 hips) and traction and ORCA (27 hips) and reported that the latter eliminated the residual acetabular dysplasia seen with the former strategy. No ORCA hip required a later pelvic osteotomy compared to 10 hips of the historic group, and 11 hips of either group developed AVN (KM: ORCA 4x II, 6x III, 1x IV and historic group 1x I, 8x II, 2x III). Severin types were 15x I and 12x II for the ORCA group and 15x I, 12x II, and 2x III for the historic group.

The senior author of the current study ¹ presented the long-term ORCA FU data (4–12 years) for 100 hips as a podium presentation in September 2022 ⁶ as the standard treatment strategy used in their unit, not mentioning the existence of the strategy reported by Uren et al. ¹ Age at the time of the surgery ranged from 11 to 30 months, with a mean preoperative AI of 40° and 18° at a mean FU of 8 years, compared to 23° and 15°, respectively, for the contralateral normal hip. Ninety-four percent of hips were a Severin type 1 or 2, 2% of hips re-dislocated, and 2% required a later pelvic osteotomy.

The previous publications^{2 –6} suggest that there was no cohort treated with Gallows traction and open reduction only between 2010 and 2020 ¹ but that the patients had undergone the ORCA procedure instead, which seemed to be the unit’s standard procedure until at least 2018, considering the 4-year minimum FU reported by Aarvold et al. ⁶

Uren et al. ¹ reported AVN in 25 hips (KM: 55x 0 or I (not differentiated), 11x II, 9x III, 5x IV) and Severin’s types I x40, II x37, III x2, and IV x1. Uren et al’s. ¹ by far worse AVN (31.2%: KM II–IV) and Severin’s type I rate (50%) compared to the historic study from the same unit ³ with AVN and Severin’s type I rates of 13.6% and 68%, respectively, indicates that the former authors’ ¹ treatment approach is not as excellent as they claim.

In conclusion, Uren et al.’s unit ¹ has not performed a study to compare the effect of Gallows traction and no traction on the outcomes of open reduction only or ORCA since at least 1988. Therefore, without comparative data, Uren et al.’s ¹ claim that their study is the clearest evidence of the beneficial effect of preoperative traction is not valid. The authors also failed to identify those hips where the International Hip Dysplasia Institute grade remained unchanged despite traction having worse outcomes compared to those where the grade had improved.

The discussed studies^{2 –6} raise the question of why Uren et al. ¹ claimed that all patients had traction and an open reduction only, despite the unit having reported to have replaced traction and OR only with the ORCA procedure since 2004.^{4 –6}