Comparative analysis of intra- and interobserver reliability and validity of five basic classifications used to determine the stage of avascular necrosis of the femoral head.

Abstract

Background: Avascular necrosis (AVN) of the femoral head lacks a standardized classification system, with multiple systems used in clinical practice by orthopedic surgeons, radiologists, and clinicians treating this condition. Differences in the reliability and diagnostic criteria across existing classification systems can give rise to conflicts in the treatment decisions made. Therefore, identifying the most reliable and uniform classification system is crucial for optimal patient care and treatment planning. This research aimed to identify the most reliable and uniform classification system across all therapeutic results. It also considered precision and dependability as diagnostic standards among these systems.

Methods: This retrospective observational study was conducted between January 2021 and January 2023, involving 20 patients who had avascular necrosis of the femoral head confirmed by magnetic resonance imaging. The reliability of diagnosis was evaluated using five main classifications (Ficat and Arlet, Steinberg, Mitchell, ARCO, Marcus Enneking) at two different time points: initially and after one month, assessed by two orthopedic surgeons and one radiologist. Key measures included the intraclass correlation coefficient (ICC) for interobserver agreement/intraobserver consistency, Spearman's rho for MRI progression correlation, and staging outcome analysis of variance.

Results: The findings showed that classification systems have different diagnostic precision. Furthermore, the Marcus Enneking scoring system recorded the best diagnostic performance (AUC = 0.65, p < 0.01) and the strongest association with MRI progression (rho = 0.65) whereas Mitchell's staging caused significant stage discrepancies between them. Marcus Enneking also had good intraobserver reliability based on kappa values, which was not the case for other techniques.

Conclusions: Our study demonstrates varying levels of reliability and diagnostic precision among different classification systems for femoral head avascular necrosis. The Marcus Enneking system seems to perform better as compared to the other systems by exhibiting a strong diagnostic accuracy and correlation with the MRI progression, this indicates that it might be more appropriate for clinical practice. Additional studies are necessary to determine how these differences in classification systems directly influence treatment outcomes in clinical practice.

Level of evidence: Level III (Retrospective observational study).