Is the Caton-Deschamps index a valuable ratio to investigate patellar height in children?

Abstract

Background: Patellar height variations can be responsible for many functional disorders in children. The Caton-Deschamps index to measure patellar height has been described for adults. Our goal was to determine Caton-Deschamps index values in a pediatric population.

Methods: Lateral radiographs of the knee were analyzed in a cohort of 300 healthy patients. All radiographs were done to evaluate the patients after minor trauma and all were reported to be normal by a senior radiologist. The cohort was divided into ten groups on the basis of the age of the child. Radiographic measurements were done by a computer-assisted technique. Intraobserver and interobserver reliability studies were performed prior to the descriptive analysis of the data.

Results: Mean patellar length (and standard deviation) was 33.39 ± 7.4 mm. Mean patellar tendon length was 34.57 ± 6.7 mm. The mean Caton-Deschamps index was 1.06 ± 0.21. Patellar and patellar tendon length significantly increased with age, whereas the Caton-Deschamps index significantly decreased. Intraobserver and interobserver reliability studies showed excellent reliabilities with an intraclass correlation coefficient that was between 0.930 and 0.944 (95% confidence interval).

Conclusions: The Caton-Deschamps index is a simple and reliable index for evaluating patellar height in children as well as adults. It is an alternative to the Insall-Salvati index measurement, in which reproducibility is poor due to difficulties in determining the distal point of the patellar tendon, and to the Koshino index, which is complex to use. In our study, there was a correlation between the Caton-Deschamps index and age, due to the progressive patellar ossification that begins at the proximal part of the patella. The Caton-Deschamps index is a pertinent and reliable ratio to evaluate patellar height in children and adolescents. To make an accurate diagnosis of patellar disorders in children, the normal, age-based Caton-Deschamps values need to be considered.