Splint Duration and Not the Mode of Anesthesia Is the Main Factor Influencing Avascular Necrosis After Closed Reduction for Developmental Dysplasia of the Hip in Kosovo

Abstract

The aim of this study was to determine whether the use of analgesia and sedation (AS) as opposed to general anesthesia (GA) for closed reduction and spica casting of children with severe developmental dysplasia of the hip (DDH) influenced the long-term incidence of avascular necrosis (AVN). In a prospective, randomized, single-blinded clinical trial we investigated 100 pediatric patients with DDH type IIIa, IIIb, and IV (according to Graf classification), who were randomly assigned into the group receiving AS, and the group receiving GA. Baseline demographics, splint duration, and type of DDH were carefully assessed. The presence of AVN was assessed at the follow-up visits at 1 and 7 years after the end of treatment. The AS-group consisted of 50 patients (46 girls) with 76 hips affected (n = 11/Type-IIIa, n = 32/Type-IIIb, and n = 33/Type-IV). The GA-group consisted also of 50 patients (44 girls) with 78 hips involved (n = 15/Type-IIIa, n = 34/Type-IIIb, and n = 29/Type-IV). At 7-years follow-up, AVN was diagnosed in 9 of 154 hips (5.8%), 5 hips in the AS-group and 4 hips in the GA group. The logistic regression model showed no significant difference in AVN incidence between the AS and GA groups at 7-years follow-up (p = 0.27). The multivariate regression analysis showed that neither the type of DDH nor the age at diagnosis influenced the incidence of AVN (p = 0.48 and p = 0.28, respectively). Splint duration was identified as the only significant factor for the long-term incidence of AVN in the treatment of severe DDH. For every month of longer splint duration, the odds of AVN at 7-years follow-up increased by a factor of 3.81 (95%CI: 1.35-13.73, p = 0.02). Closed reduction and spica casting of children with severe DDH under AS can be considered a feasible alternative to management under GA. All efforts must be made to diagnose patients with DDH as early as possible and shorten the duration of splint treatment to prevent the development of AVN. Level of Evidence. Level II-1.

Keywords: analgesia and sedation; closed reduction; developmental dyspalsia of the hip; general anesthesia; splint duration.

FIGURE 1

A 4.5-months-old baby-girl with left hip dislocation after closed reduction and spica casting.

FIGURE 2

The CONSORT flowchart shows the selection and randomization of the study subjects in two different groups of the clinical trial.

FIGURE 3

A baby girl with left hip dislocation undergoing closed reduction under analgesia and sedation (A–C), and a baby girl with bilateral hip dislocation undergoing closed reduction under general anesthesia (D–F): (A) Ultrasound examination of the hips at the age of 4.9 months showing left hip dislocation type IIIb according to Graf’s classification; (B) radiography of the hips at the age of 20 months, showing AVN type I of the left femoral head according to Ogden and Bucholz classification- hypoplastic epiphysis of the proximal femur; (C) radiography of the hips at the age of 7.9 years showing complete improvement of ossification of the left femoral head. (D) Ultrasound examination of the hips at the age of 3.9 months showing bilateral hip dislocation type IV according to Graf’s classification; (E) Radiography of the hips at the age of 2.4 years showing AVN type II of the left femoral head according to Ogden and Bucholz classification with overgrowth of greater trochanter into valgus, lateral metaphysis shows evidence of injury; (F) radiography of the hips at the age of 7.6 years showing AVN type II of the left femoral head according to Ogden and Bucholz classification – overgrowth of greater trochanter into valgus.

FIGURE 4

The logistic regression model at 1-year follow-up. The univariate analysis showed no significant difference in the development of AVN between AS and GA groups and identified age at diagnosis and splint time as substantial risk factors. In the multivariate analysis, none of the independent variables showed a significant influence on the development of AVN at 1-year follow-up. AVN, avascular necrosis; AS, analgesia and sedation; GA, general anesthesia; N, number of affected hips.

FIGURE 5

The logistic regression model at 7-years follow-up. The univariate analysis showed no significant difference in the development of AVN between AS and GA groups and identified age at diagnosis and splint time as substantial risk factors. In the multivariate analysis, only splint duration showed a significant influence on the development of AVN at a 7-years follow-up. For every month of longer splint duration for the treatment of DDH, the odds of AVN at 7-years follow-up increased by a factor of 4.03 (95%CI: 2.13–9.13, p < 0.001) in univariate analysis, and 3.81 (95%CI: 1.35–13.73, p = 0.02) in multivariate analysis. AVN, avascular necrosis; AS, analgesia and sedation; GA, general anesthesia; N, number of affected hips.

FIGURE 6

The linear regression model for splint duration as a dependent variable. Both univariate and multivariate analyses showed a highly significant influence of age at diagnosis and the degree of DDH on the splint duration. Furthermore, splint duration was –0.66 months (95%CI: –1.03 to –0.3, p < 0.001) shorter in the GA group than in the AS group in the univariate analysis, and –0.39 months (95%CI: –0.63 to –0.15, p = 0.002) in multivariate analysis. Interestingly, affection of both hips by DDH led to a 0.27 months longer splint duration (95%CI:0.01–0.54, p < 0.05) in the multivariate analysis. AS, analgesia and sedation; GA, general anesthesia; N, number of affected hips.