Lower Than Expected In-Hospital Mortality of COPD Exacerbations

Abstract & Commentary

Synopsis: In this study of a large nationwide database, mortality during hospitalization for acute exacerbation of COPD was 2.5%, which is substantially lower than that reported in previous studies.

Source: Patil SP, et al. In-hospital mortality following acute exacerbations of chronic obstructive pulmonary disease. Arch Intern Med. 2003;163:1180-1186.

The purpose of this study using a large administrative database was to obtain generalizable estimates of in-hospital mortality in patients admitted for an acute exacerbation of COPD, and also to identify risk factors for death during the hospitalization. Patil and associates used data from the 1996 Nationwide Inpatient Sample, a 20% sample of all admissions to acute-care, nongovernmental hospital beds (roughly 6 million patients), and identified patients with COPD exacerbations using ICD-9 discharge codes. Outcome variables examined were in-hospital mortality, length of stay, total charges, mechanical ventilation, and discharge disposition. Demographic data on the patients included age, gender, race, and income (as estimated from residence zip codes).

A total of 71,130 patients (1.1% of the total database) had the 429.21 ICD-9 code for acute COPD exacerbation and were older than 40 years, and were thus used in the data analysis. Median patient age was 69.9 years; 56.4% were female; 86.6% were white; and the majority was below the national median for annual income. Median length of hospital stay was 5 days (interquartile range, 3-7 days), 3% of the patients required mechanical ventilation during the hospitalization, and 2.5% died (99% confidence interval, 2.4-2.7%). Among patients who underwent mechanical ventilation, the mortality rate was 27.8%. Patients who died following acute exacerbation of COPD were older (74.1 ± 9.3 years vs 69.8 ± 11.2 years), had a higher level of comorbid illness, had longer length of stay, and generated higher charges than those who were alive at discharge. By multivariate analysis, older age, male sex, higher income, nonroutine admission sources, and more comorbid medical conditions were all independent risk factors for mortality during the hospitalization. Race was not related to outcome.

Comment by David J. Pierson, MD

The 2.5% in-hospital mortality for patients admitted with an acute COPD exacerbation in the present study is markedly lower than virtually all previously reported mortality rates. The discrepancy is most likely due to selection bias in previous studies, which have been institution- or ICU-based or have examined other non-representative segments of the general population. This study is the closest yet to an assessment of the "average" American patient, in that it included all acute-care hospital admissions other than those to government (ie, VA and military) hospitals, nationwide, during 1996.

Most previous studies that have attempted to identify risk factors for adverse outcomes of COPD exacerbations have used physiologic indices, such as measures of oxygenation and acid-base status and acute physiology scores. The data available from the database used by Patil et al were confined to administrative information, yet clear predictors of adverse outcomes could be identified.

Not surprisingly, older patients and those with more comorbid medical conditions fared less well than their younger and healthier counterparts in this study. Despite a larger overall number of women in the database, men were more likely to die during hospitalization. Patil et al speculate that this may be because men are more likely to delay seeking medical care when they are sick than women, and thus tend to present later in the course of illness. Somewhat surprisingly, patients with annual incomes less than $25,000, as assessed from the postal zip code of residence, were significantly less likely to die during the hospitalization than patients with incomes of more than $35,000. This is at odds with what most studies of COPD prognosticators have found. Here, Patil et al speculate that, with more access to outpatient health care, individuals with higher incomes might have delayed presentation to the hospital when they became seriously ill, whereas lower-income patients with fewer resources presented to the hospital earlier in the course of illness.

Like death certificate diagnoses, discharge ICD-9 codes are a pretty shaky source for accurate information on what was wrong with a patient. However, despite this limitation, the findings of this study are of considerable interest. The degree to which its lower than previously reported mortality rate can be ascribed to improvements in patient assessment and management since earlier studies were performed is impossible to say. However, the fact that more than one-fourth of all patients who required mechanical ventilation during hospitalization for an acute exacerbation of COPD died should remind us that acute-on-chronic ventilatory failure remains a life-threatening event and stimulate us to continue to look for ways to improve its outcome.

Dr. Pierson is Professor of Medicine University of Washington Medical Director Respiratory Care Harborview Medical Center Seattle.