The effect of marital status on the presentation and outcomes of elderly male veterans hospitalized for pneumonia

Background: Although marital status has been shown to affect the outcomes of many conditions, there are limited data on the relationships between marital status and the presentation and outcomes of pneumonia. Methods: We used Veterans Affairs administrative databases to identify a retrospective cohort of male veterans age ≥ 65 years hospitalized for pneumonia between 2002 and 2007. We assessed unadjusted and adjusted associations between marital status and mortality, hospital length of stay, and readmission to the hospital using generalized linear mixed-effect models with admitting hospital as a random effect and adjusted for baseline patient characteristics. Results: There were 48,635 patients (26,558 married and 22,077 unmarried) in the study. Married men had a slightly higher Charlson comorbidity score (3.0 vs 2.8, P < .0001) but were less likely to require ICU admission, ventilator support, and vasopressor treatment during the first 48 h of hospitalization. Married patients had significantly lower crude and adjusted in-hospital mortality (9.4% vs 10.6%; adjusted OR, 0.87; 95% CI, 0.81-0.93) and mortality during the 90 days after hospital discharge (14.7% vs 16.0%; adjusted OR, 0.92; 95% CI, 0.88-0.98). Their adjusted incidence rate ratio length of stay was also lower (0.92; 95% CI, 0.91-0.92). Conclusions: Unmarried elderly men admitted to the hospital with pneumonia have a higher risk of in-hospital and postdischarge mortality, despite having a lower degree of comorbidity. Although marital status may be a surrogate marker for other predictors, it is an easily identifiable one. These results should be considered by those responsible for care-transition decisions for patients hospitalized with pneumonia.