How Accurate is the Multiplier Method in Predicting the Timing of Angular Correction After Hemiepiphysiodesis?

Abstract

The purpose of the study was to validate the accuracy of the Multiplier Method (MM) in predicting the timing of angular correction after hemiepiphysiodesis and to determine the role of using skeletal age when calculating those predictions. This retrospective study included 131 physes in 77 patients treated with hemiepiphysiodesis to gradually correct a coronal plane deformity before skeletal maturity. To compare the MM's predictions to the actual treatment duration, the "desired angular correction" was considered the actual achieved angular correction determined from the "endpoint x-ray" (last x-ray before implant removal). We measured the bone length and width of the growth plate from the preoperative x-ray and calculated the MM's prediction of the duration of treatment based on the MM formula. We compared the predicted duration to the observed duration of treatment for each case. The difference was calculated by subtracting the observed duration from the predicted duration. The result was the "absolute difference," which is the number of months over or under predicted by the MM. The mean absolute difference between the MM's predicted duration and the observed duration was 2.31 months, which was highly significant (P≤0.001). The MM's prediction agreed with the observed duration of treatment (ie, zero absolute difference) in 15% of the predictions, 69% were under predicted, and 16% were over predicted. Sixty-eight percent of the absolute differences were within 3 months regardless of the direction of error. The mean difference was relatively less in genu varum cases and was statistically significant (P=0.047). Comparing the mean difference using chronological age and skeletal age in the formula showed no statistically significant difference. The MM has a tendency to under predict. Therefore, doing a guided growth right before skeletal maturity should be started 2 to 4 months earlier than suggested by the MM. Moreover, our data did not show that the bone age gave more accurate predictions than chronological age. Level IV.