C-reactive protein and other predictors of poor outcome in patients hospitalized with exacerbations of chronic obstructive pulmonary disease

Abstract

Background and objective: CRP is elevated in patients with acute exacerbations of COPD (AECOPD), but there is little information on whether this biomarker can help to identify adverse short-term clinical outcomes. Methods: A 6-month prospective study of all patients with AECOPD requiring hospital admission. Clinical, laboratory (including plasma CRP levels at admission) and functional data were recorded. The outcome variable (the adverse outcome) consisted of: (i) death in hospital or within 15 days of discharge, (ii) transfer to the intensive care unit, or (iii) development of acute heart failure during hospitalization. Results: Data from 147 patients with a total of 160 admissions were recorded. During follow up, 38 (23.7%) adverse outcomes were observed, including 13 (8.8%) and 8 (5.4%) patients who died during hospitalization or within 15 days of discharge, respectively. CRP at a level of 50 mg/L was related to an adverse outcome (OR 4.9, 95% CI: 1.92-12.6, P < 0.01), although by itself it was neither sensitive nor specific (area under the receiver operating characteristic curve (AUC) 0.69, 95% CI: 0.60-0.77). However, a risk score derived from the combination of CRP with other variables, such as 'current smoker', 'at least two comorbidities' and 'confusion,' at admission showed good predictive ability to identify an adverse outcome (AUC of 0.80, 95% CI: 0.72-0.88). Conclusions: Plasma CRP in combination with other variables obtained at admission may assist identification of high-risk patients with AECOPD. © 2008 The Authors.