RT Journal Article
SR Electronic
T1 Population-based analysis of COPD patients in Catalonia: implications for case management
JF European Respiratory Journal
JO Eur Respir J
FD European Respiratory Society
SP PA4956
DO 10.1183/1393003.congress-2017.PA4956
VO 50
IS suppl 61
A1 Baltaxe, Erik
A1 Roca, Josep
A1 Vela, Emili
A1 Tenyi, Akos
A1 Cano, Isaac
A1 Monterde, David
A1 Claries, Montse
A1 Garcia-Altes, Anna
A1 Hernandez, Carme
A1 Escarrabill, Joan
YR 2017
UL http://erj.ersjournals.com/content/50/suppl_61/PA4956.abstract
AB Rationale: Despite substantial progresses on standard of care recommendations, management of Chronic Obstructive Pulmonary (COPD) patients show potential for improvement provided that patients’ heterogeneities are better understood.Objectives: To perform a population-based analysis of the COPD patients in Catalonia (ES) (7,5 M citizens) aiming at exploring the potential of the health registry information to enhance risk assessment and stratification in the clinical arena.Methods: The characteristics of 264,830 COPD cases registered in 2014 in the Health Surveillance System were assessed. Multiple logistic regression analysis was used to elaborate predictive models for four health indicators: (i) mortality, (ii) unplanned hospitalizations (all and COPD-related hospitalizations), (iii) multiple exacerbators, and, (iv) high users of healthcare resources. No clinical, nor forced spirometry data were available.Main Results: Multimorbidity, expressed by GMA (adjusted morbidity groups) scoring, was the covariate with highest impact in the four predictive models. This score explained a substantial percentage of interindividual variability (AUC): (i) mortality (0.763); (ii) unplanned hospitalizations (0.829); (iii) multiple exacerbators (0.766); and, (iv) higher use of healthcare resources (0.803). Patients above the 85 percentile in terms of healthcare costs per year represented 59% of the overall costs of COPD patients.Conclusions: The predictive models stress the explanatory role of the covariate multimorbidity. The results highly encourage further developments fostering interoperability between health registries and electronic health records to enhance clinical risk prediction.