Abstract
BACKGROUND: COPD ranks as the third leading global cause of mortality. Despite the widespread use of the BODE index and its variants for mortality prediction, their accuracy may be affected by factors like ethnicity, altitude and regional disparities. This study aimed to assess a new altitude-adapted prognostic index in COPD patients at moderate altitudes compared with the BODE and other mortality predictors. METHODS: We evaluated the performance of BODE and other standard prognostic indexes in an exploratory cohort of COPD patients using Cox proportional hazards models. Additionally, we developed a new risk-weighted prognostics model. We assessed its performance utilising receiver operating characteristic curves and compared it with BODE index variants. The indices were validated in a second cohort. The mean follow-up period in both cohorts was 7 years. RESULTS: In the exploratory cohort (n=416), COPD patients had a 77% 5-year survival rate. eBODE exhibited superior predictive capacity (HR 1.22, 95% CI 1.11-1.34, p<0.05). Prognostic models included oxygen saturation measured by pulse oximetry <90%, body mass index, 6-min walk test (6MWT) and current smoking as primary relevant risks (BOSEA-90), with the second model excluding 6MWT (BOSA-90). BOSEA-90 predicted a major risk per point (HR 1.41, 95% CI 1.31-1.51, p<0.001) and superior discrimination (area under the curve (AUC) >0.75) compared to other indices (AUC <0.70, p<0.001). The validation cohort (n=436) confirmed effectiveness in differentiating mild and very severe cases. CONCLUSIONS: The BODE prognostic index and its variants effectively predicted mortality, and the newly proposed BOSEA-90 and BOSA-90 indices demonstrated appropriate performance for a group of COPD patients at moderate altitude.