Arterial carbon dioxide tension on admission as a marker of in-hospital mortality in community-acquired pneumonia

PURPOSE: Respiratory failure is the leading cause of death among patients admitted with community-acquired pneumonia. We sought to determine the association between arterial carbon dioxide tension (PaCO2) and in-hospital mortality in patients admitted with pneumonia. METHODS: We analyzed data from 2171 patients aged greater than or equal to17 years who had been admitted for community-acquired pneumonia to an acute care hospital in Edmonton, Alberta. We compared the risk of all-cause in-hospital mortality using a Cox proportional hazards model across categories of PaCO2. RESULTS: Overall, in-hospital mortality was 10% (n = 218). Compared with patients with normal PaCO2 values (40 to 44 mm Hg), in-hospital mortality was greater (adjusted odds ratio [OR] = 1.8; 95% confidence interval [CI]: 1.0 to 3.2) among patients with hypocapnia (PaCO2 <32 mm Hg). In-hospital mortality was also greater (OR = 2.6; 95% CI: 1.5 to 4.5) in patients with hypercapnia (greater than or equal to45 mm Hg). In-hospital mortality was similar in patients With PaCO2 values between 32 and 35 mm Hg (OR = 1.55; 95% CI: 0.89 to 2.79) and those with values between 36 and 39 mm Hg (OR = 1.42; 95% CI: 0.77 to 2.61). CONCLUSION: Among patients admitted with community-acquired pneumonia, in-hospital mortality was greater in those with hypocapnia or hypercapnia. These data suggest that measurement of PaCO2 adds prognostic information to standard prediction rules and should be used for clinical and epidemiologic purposes to risk-stratify in-hospital patients with community-acquired pneumonia. (C) 2005 Elsevier Inc. All rights reserved.