Most Coronal Knee Deformities of Healed Nutritional Rickets Under 8 Years of Age Remodel Spontaneously: Building Evidence for Practice Change.

Abstract

Nutritional rickets remains a global health burden. The optimal management of coronal knee deformities in children/adolescents with healed nutritional rickets as to the indications of conservative treatment versus guided growth surgery remains unsettled. The objective was to report the natural history of disease regarding the remodeling potential of coronal knee deformities and isolate potential patient- and deformity-related factors that are predictors of favorable deformity remodeling. We prospectively assessed 109 knees in 57 patients-mean age, 6.1 years (range, 2 to 12)-for spontaneous correction of their coronal knee deformities over a period of 1 year. Male/female:33/24, valgum/varum:34/23, bilateral/unilateral:52/5 patients. Outcome variables were, radiologic tibiofemoral angle and a questionnaire to assess parent satisfaction. They were correlated to independent patient- and deformity-related variables, including age, sex, deformity direction/severity, and laterality. The mean age of patients that achieved complete knee remodeling (4.9 y) was significantly lower compared with patients who experienced complete failure of remodeling (8.4 y), P =0.000, (95% CI, 4.3-5.4). In 109 knees, there was a highly significant negative Spearman correlation coefficient between age and degrees of coronal correction, patient satisfaction, and complete failure of remodeling R=-0.660, R=-0.475, R=-0.434, respectively, P <0.001. The knees of patients >8 years of age that failed to remodel (87.5%) was significantly higher compared with patients ≤8 years (29.9%), P =0.000, OR=16.4, (95% CI, 5.2-52.2). There were statistically significant correlations between deformity severity/tibiofemoral angle and the remodeling potential, with a cutoff angle of 16.5 degrees/13 degrees for valgum/varum above which deformities were less likely to remodel spontaneously, P =0.03, P =0.007. Complete knee remodeling in unilateral deformities was statistically but not clinically significant. Correlations of sex and deformity direction with remodeling outcomes were insignificant. The inter- and intrarater agreements were excellent for the radiologic tibiofemoral angle measurements (ICC=0.993, CI, 0.982-0.997, P  <0.01) and (ICC=0.994, CI, 0.987-0.998, P <0.01), respectively. Children with coronal knee deformities of healed nutritional rickets under the age of 8 years should be given the full opportunity to remodel spontaneously for at least 1 year, particularly children with moderate deformities ≤16 degrees. This is irrespective of sex, deformity direction, and laterality. Most deformities remodel spontaneously. This suggests that these deformities may be overtreated. This could reduce unnecessary surgeries and potentiate cost-effectiveness of treatment and risk-benefit profile. Level IV prognostic studies-investigating the effect of a patient characteristic on the outcome of disease.