Hip arthrogram parameters predict radiographic outcomes of patients with developmental dysplasia of the hip treated...

This study aimed to investigate the relationship between intraoperative hip arthrogram parameters and residual acetabular dysplasia (RAD) and avascular necrosis of the femoral head (AVN) in children with developmental dysplasia of the hip (DDH) treated by closed reduction.

We retrospectively analyzed the data of 102 patients (110 hips; mean age, 14.6 months ± 4.7 months) with DDH treated by closed reduction. A hip arthrogram was routinely performed during the operation. The femoral head coverage rate (FHC), medial pool distance of the hip (MPD), labral inversion, and reduction quality classification were evaluated under the hip arthrogram. The presence of RAD and AVN was assessed on radiographs at the last follow-up. The relationship between each arthrogram parameter and RAD as well as AVN was investigated using a t-test, chi-square test, and logistic regression.

The overall FHC and medial pool distance of the hip (MDP) averaged 42.2% ± 12% and 8.1% ± 11.7%, respectively. There were 80 hips (72.7%) with labral inversion and 30 hips (27.2%) without. The reduction quality was type A in 57 hips (51.8%), type B in 28 hips (25.4%), and type C in 25 hips (22.7%). A total of 32 hips (29%) were in the RAD group, and 78 hips (71%) were in the recovered group according to whether pelvic osteotomy was performed or not and according to the last Severin grade. The FHC was significantly higher in the recovered group than that in the RAD group (P = 0.014). No significant difference was observed in sex, age at reduction, side, preoperative acetabular index, International Hip Dysplasia Institute classification, follow-up time, quality of reduction, MDP, and proportion of labral inversion between the recovered and RAD groups. Logistic regression analysis showed that only the FHC was a risk factor for RAD. The incidence of AVN above type II was 11.8% in this group of patients, and the incidence of AVN was significantly higher in patients with labral inversion (23.2%) than that in those without (7.5%; P = 0.041). Logistic regression analysis showed that labral inversion was a risk factor for AVN.

The FHC measured under arthrogram can predict the occurrence of RAD after closed reduction of DDH, whereas MDP, reduction quality classification, and labral inversion are of little significance. Labral inversion is a risk factor for AVN.