Summary

An evaluation on the interrater reliability of several common radiographic parameters used for patellofemoral instability and the attempt to improve reliability through training and consensus building around measurement definitions.

Abstract

Purpose

Surgical decision making for children and adolescents with patellofemoral instability relies heavily on radiographic findings and measurements. In order to be clinically useful, radiographic parameters must demonstrate acceptable interrater reliability. The purpose of this study was 1) to evaluate the interrater reliability of several common radiographic parameters used for patellofemoral instability, and 2) to attempt to improve reliability for those measurements demonstrating unacceptable interrater reliability.

Methods

Two fellowship-trained musculoskeletal radiologists assessed 50 patients on several additional common patellofemoral instability radiographic measures including the trochlear crossing sign, Caton-Deschamps index, and congruence angle (Table 1). Intraclass Correlation Coefficients (ICC) were calculated for each measurement. For those measurements demonstrating unacceptable interrater reliability (ICC < 0.6), raters discussed consensus methods to improve reliability and examined 20 images from a separate set of subjects. If reliability was still low after the second round of assessment, the measure was considered unreliable.

Results

Of the 50 selected participants, 22 (44%) were male and the average age at the time of imaging was 14 ± 2 years. Interrater reliability for 5 radiographic and 7 MRI assessments was initially acceptable, 4 radiographic and 2 MRI assessments improved after consensus training, and 6 radiographic and 3 MRI assessments remained unacceptably unreliable even after consensus training (Table 2).

Conclusion

This study found that interrater reliability of many clinically relevant patellofemoral instability parameters was unacceptably low. Some measurements were initially unreliable but improved with training and consensus building around measurement definitions. Unreliability may persist for those measurements with a high level of difficulty or ambiguity, or for indices that are infrequently assessed.

Significance

Lack of measurement reliability can jeopardize patient care both when determining surgical indications as well as choosing an appropriate procedure based upon skeletal maturity. Physicians should ensure they are using a standardized definition or technique for radiographic measurements in children with patellofemoral instability, and unreliable measurements should be removed from clinical decision-making algorithms.