Validation of the codex index in patients with COPD enrolled in randomized clinical trials: A post-hoc analysis of the trilogy, trinity and tribute studies

Abstract

Introduction: The CODEX (age-adjusted Charlson comorbidity score, Obstruction, Dyspnea, and severe EXacerbations) index has shown to be useful at predicting survival and COPD readmissions, with a prognostic capacity superior to other indices. Its validation in randomized clinical trials has not been performed.

Aims and Objectives: To explore the performance of the CODEX in a cohort from pooled multinational randomized pharmacological intervention clinical trials, in comparison to the BODEX, DOSE and ADO indices.

Methods: TRILOGY, TRINITY, and TRIBUTE were phase III randomized clinical trials that evaluated 1-yr treatment with extra fine BDP/FF/GB versus other maintenance therapies in symptomatic patients with severe COPD and a history of exacerbations. Patients from the intent-to-treat population of each study were considered, those with missing data for any of the variables involved in the derivation of the COPD multivariate indices were excluded.

Results: 5588 patients were considered, 5584 entered the analysis. 565 (10.1%) of patients experienced death or a severe exacerbation. The mean (SD) CODEX index was 5.7 (1.15), with 82.6% of patients scoring between 5-9 on a 0-10 scale. For patients with a score ≥5, the Hazard Ratio (HR) for either death or a severe exacerbation was 2.6 (95%CI: 1.9 – 3.5, p<0.0001). ROC curves indicated a predictive capacity for the CODEX that was higher than the ADO and DOSE and similar to the BODEX.

Conclusion: A high CODEX index is significantly associated with mortality or severe exacerbation outcomes in a pooled cohort from randomized interventional clinical trials.