Decaf score for predicting treatment escalation in exacerbations of COPD

Introduction: There is a need for a robust prediction model that identifies patients with acute exacerbation of COPD (AECOPD) at risk for treatment escalation during hospital admission. The DECAF score has been validated to predict in-hospital mortality. It is not known if the DECAF score might predict treatment escalation in AECOPD. Such a score might serve to allocate the level of monitoring on hospital admission. Methods: Prospective cohort study in a single teaching hospital in the Netherlands. The DECAF score was noted in consecutive patients admitted to the hospital with an AECOPD. De sensitivity and specificity of the DECAF score to identify treatment escalation during admission was calculated. Treatment escalation was defined as need for NRM, NIV or transfer to ICU. Results: 83 patients were included in the study. Most patients had severe COPD (mean FEV1 37%). Treatment escalation was seen in 13.1% of patients; 10.7% needed non-invasive ventilation to recover and 2.4% were transferred to the intensive care unit for treatment. No patients died during admission. A low DECAF score could not safely excluded the need for treatment escalation. The sensitivity of a low DECAF score (0-1) to exclude treatment escalation was 59% (95%-CI: 48-71%). Specificity of a low DECAF score to exclude the need for treatment escalation was similarly mediocre, 71% (95%-CI: 29%-96%). Conclusion: The DECAF score on admission cannot be used to predict treatment escalation during admission for an AECOPD.