Abstract
BACKGROUND: The risk of exacerbation in chronic obstructive pulmonary disease (COPD) depends on the severity of disease and other less well known factors. Predictive models of exacerbation are more accurate than the forced expiratory volume in one second (FEV1). The objective was to design a model that predicts the risk of exacerbation in COPD. METHODS: Retrospective cohort study with data from the electronic medical records of patients diagnosed with COPD in the province of Lleida (Spain). A total of 2501 patients were followed during 3 years. The dependent variable was acute exacerbation; independent variables were: clinical parameters, spirometry results, severity of disease, influenza and 23-valent pneumococcal immunisation, comorbidities, smoking and history of exacerbation. The association of these variables with disease exacerbation was measured by the adjusted odds ratio using a logistic regression model. RESULTS: Mean age at the start of the study was 68.38 years (SD = 11.60) and 74.97% patients were men; severity of disease was considered mild in 50.82% of patients, moderate in 35.31%, severe in 9.44% and very severe in 4.44%. During the three year study period up to 83.17% of patients experienced at least one exacerbation. Predictive factors in the model were age, gender, previous exacerbations, influenza and 23-valent pneumococcal immunisations, number of previous visits to the General Practice and severity (GOLD), with an area under the ROC curve (AUROC) of 0.70. CONCLUSIONS: This model can identify patients at high risk of acute exacerbation. Preventive measures and modification of treatment in these high-risk patients would improve survival.