Agreement in quantifying the extension of autoimmune-associated interstitial lung disease using the Goh method

Abstract

Introduction

Interstitial lung disease (ILD) is one of the leading causes of mortality in autoimmune diseases. The extent of the disease is a determining factor in the prognosis and treatment initiation and monitoring. Quantification using the Goh method is the most commonly used method; however, it is subjective. So far, no studies have evaluated the level of agreement among various readers.

Objective

The study's objective is to determine the interobserver and intra-observer variability in using ILD quantification among physicians from various specialties and levels of experience.

Methods

Images from chest computed tomography of patients with rheumatoid arthritis (RA) or systemic sclerosis (SSc) and ILD were collected. The five necessary cuts described by Goh were extracted to be evaluated by pulmonologists, rheumatologists, radiologists, fellows, and a thoracic radiologist (gold standard). The interobserver and intra-observer variability values were calculated using the intraclass correlation coefficient test or Cohen's Kappa test, depending on the nature of the variable, between each group of medical specialties and in comparison with the gold standard.

Results

Seventy-nine patients were selected, primarily women, 56% having SSc. A total of 1098 CT scans were performed. The intraclass correlation coefficient was .75 (95% CI: .67–.81), including all nine readers. The best correlation with the gold standard was found among pulmonologists (CCI .83) and rheumatologists (CCI .81). According to severity (more significant or less than 20% extension), the Kappa coefficient was .64 among the nine readers. The intraclass correlation coefficient for the average intra-observer correlation of all readers was .89 (95% CI: .81–.93), and the Kappa coefficient was .82.

Conclusion

The Goh method is valuable and highly correlated among a diverse group of specialties that manage ILD, making it a practical tool for assessing the extent of the disease.