Comparison of Warrick Scoring and Goh Staging for Detecting Progression in Systemic Sclerosis–Associated Interstitial Lung Disease

Objective Interstitial lung disease (ILD) is the leading cause of mortality in systemic sclerosis (SSc). The Warrick semiquantitative scoring system and the Goh staging algorithm provide different approaches for assessing lung involvement on high-resolution computed tomography (HRCT). We compared their sensitivity for detecting radiological progression in SSc-ILD.

Methods We retrospectively analyzed 33 patients with SSc-ILD who underwent serial HRCT examinations. Each scan was evaluated using the Warrick score (0-30) and Goh staging (limited/extensive). Progression was defined as an increase of ≥ 3 points in the Warrick score or transition from limited to extensive disease by Goh staging. Comparative progression analysis was restricted to the 24 patients with baseline limited disease, as only these patients were eligible for Goh-defined progression. Progression rates were compared using the McNemar test, and receiver-operating characteristics analysis assessed Warrick score thresholds for identifying Goh-defined extensive disease.

Results The cohort included 33 patients (85% female; mean age 53.2 [SD 8.9] years), all with diffuse cutaneous SSc. At baseline, 24 (73%) had limited disease and 9 (27%) had extensive disease. Among patients with baseline limited disease, Warrick scoring identified progression in 15 (62%) compared with 8 (33%) by Goh staging (McNemar chi-square test 7.00, P = 0.01). All Goh progressors also fulfilled Warrick progression criteria, whereas 7 additional patients showed Warrick progression without stage transition. Among baseline extensive patients, 6 (67%) demonstrated further radiological worsening according to Warrick scoring. Follow-up Warrick scores showed excellent discrimination for Goh-defined extensive disease (area under the curve 0.94, 95% CI 0.86-1.00), with an optimal threshold of 20 points (sensitivity 88%, specificity 88%). Agreement between methods was substantial (Cohen κ 0.76).

Conclusion Warrick scoring was significantly more sensitive than Goh staging for detecting radiological progression in SSc-ILD and remained informative in patients already classified as having extensive disease, supporting its use in longitudinal monitoring.