Sex Differences in Veterans Admitted to the Hospital for COPD Exacerbation.

RATIONALE: As chronic obstructive pulmonary disease (COPD) prevalence in women has outpaced that in men, COPD-related hospitalization and mortality are now higher in women. Presentation, evaluation, and treatment of COPD differ between women and men. Despite higher smoking rates in Veterans, little work has characterized differences in Veterans with COPD by sex.

OBJECTIVES: We determined risk factors for 30-day readmission amongst Veterans hospitalized for COPD exacerbations and how they differed by sex.

METHODS: We performed a retrospective observational analysis of Veterans receiving primary care in Veterans Health Affairs (VHA) facilities. We included VA-based hospitalizations for Veterans with a COPD exacerbation (identified by International Classification of Disease, 9th revision (ICD-9) codes) who survived to discharge between fiscal years 2012-2015. Primary outcome was 30-day readmission. Predictors ascertained prior to hospitalization included smoking status (current, former, never), pulmonary function testing (PFT), pulmonary medication prescriptions, and medical and psychiatric comorbidities (identified by ICD-9 codes). We created combined and sex-stratified multi-variate logistic regression models to identify associations with 30-day readmission.

RESULTS: Our sample included 48,888 Veterans (4% women). Compared to men, women Veterans were younger, more likely to be non-white, and differed in smoking status. Women were more likely to have asthma, drug use, and several psychiatric comorbidities. Before hospitalization, women were less likely to have PFTs (76% vs. 78%, p=0.01) or be treated with anti-muscarinic (43% vs. 48%) or combined long-acting bronchodilator/inhaled corticosteroid (61% vs. 64%) inhalers. Women were more likely to receive nicotine replacement therapy (all p<0.01). Women had shorter length of stay (median days: 2 vs. 3; p=0.04) and lower 30-day readmission rate (20 vs. 22%; p=0.01). In adjusted models including both sexes, age, anti-muscarinic use, comorbidities, and diagnosis of drug or alcohol use were associated with readmission; there was no association with sex and readmission risk. In models stratified by sex, associations were similar between women and men.

CONCLUSION: This study suggests differences between women and men hospitalized for COPD regarding presentation, evaluation, and management. Readmission is strongly influenced by comorbidities, suggesting individualized and comprehensive case management may reduce readmission risk for women and men with COPD.