Introduction: Accurate outcomes data and predictors of outcomes are fundamental to the effective care of patients with COPD and in guiding them and their families through end-of-life decisions.
Design: We conducted a retrospective cohort study of 166 patients using prospectively gathered data in patients with COPD who required mechanical ventilation for acute respiratory failure of diverse etiologies.
Results: The in-hospital mortality rate for the entire cohort was 28% but fell to 12% for patients with a COPD exacerbation and without a comorbid illness. Univariate analysis showed a higher mortality rate among those patients who required > 72 h of mechanical ventilation (37% vs 16%; p < 0.01), those without previous episodes of mechanical ventilation (33% vs 11%; p < 0.01), and those with a failed extubation attempt (36% vs 7%; p = 0.0001). With multiple logistical regression, higher acute physiology score measured 6 h after the onset of mechanical ventilation, presence of malignancy, presence of APACHE (acute physiology and chronic health evaluation) II-associated comorbidity, and the need for mechanical ventilation > or = 72 h were independent predictors of poor outcome.
Conclusions: We conclude that among variables available within the first 6 h of mechanical ventilation, the presence of comorbidity and a measure of the severity of the acute illness are predictors of in-hospital mortality among patients with COPD and acute respiratory failure. The occurrence of extubation failure or the need for mechanical ventilation beyond 72 h also portends a worse prognosis.