Format

Send to

Choose Destination
Am J Med. 1995 Mar;98(3):272-7.

Predicting mortality of patients hospitalized for acutely exacerbated chronic obstructive pulmonary disease.

Author information

1
Department of Respiratory Physiology, Catholic University, Rome, Italy.

Abstract

PURPOSE:

To identify factors affecting the short-term prognosis of patients with acutely exacerbated chronic obstructive pulmonary disease (COPD).

PATIENTS AND METHODS:

The 590 patients having COPD as primary disease who were hospitalized in the pneumology unit of a university hospital from 1981 to 1990 were studied. A standardized protocol for the treatment of acutely exacerbated COPD was adopted for all the patients. The patient records were retrospectively analyzed by two observers, and 23 clinical and laboratory variables defining the patient status on admission were collected. Age and arterial gas data were also taken into account, and the outcome mortality was recorded. Interobserver reproducibility was tested by computing the kappa coefficient and Spearman's rho for dichotomous and continuous variables, respectively. The relationship of clinical and laboratory factors to the outcome was assessed first by univariate analysis and then by a logistic regression analysis assessing the independent predictive role of variables previously shown to be univariately correlated with mortality.

RESULTS:

The mortality rate was 14.4%. The logistic regression analysis identified four independent predictors of death: age (odds ratio [OR] 1.07; 95% confidence interval [CI] 1.04 to 1.11), alveolar-arterial oxygen gradient greater than 41 mm Hg (OR 2.33; 95% CI 1.39 to 3.90), ventricular arrhythmias (OR 1.91; 95% CI 1.10 to 3.31), and atrial fibrillation (OR 2.27; 95% CI 1.14 to 4.51).

CONCLUSIONS:

Patients with acutely exacerbated COPD having a high risk of death can be identified at the time of admission. Variables reflecting heart dysfunction are important determinants of this risk. Among pulmonary function data, only alveolar-arterial oxygen gradient contributes to the predictive model.

PMID:
7872344
DOI:
10.1016/s0002-9343(99)80374-x
[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Elsevier Science
Loading ...
Support Center