Search | Search by Center | Search by Source | Keywords in Title
Abrams T, Rosenthal GE, Vaughan-Sarrazin MS. Methods for Assessing the Impact of Anxiety and Depression on Chronic Obstructive Lung Disease. Poster session presented at: Society of General Internal Medicine Annual Meeting; 2010 Apr 30; Minneapolis, MN.
Objectives: Studies of hospitalized patients indicate that comorbid depression and anxiety increase the risk of adverse outcomes. While studies examining depression and chronic obstructive pulmonary disease (COPD) adverse outcomes are more robust, studies examining the impact of anxiety among hospitalized patients with COPD lack agreement. Moreover, many of these studies lack generalizability due to the limited sampling sites. Thus, this study was completed on a national cohort of hospitalized COPD patients to test the effects pre-existing comorbid anxiety and depression on COPD related mortality using a national cohort identified using administrative data. Methods: This study used a retrospective cohort design to identify 23,306 consecutive patients admitted for COPD (mean age, 69.0; 97% male) admitted to 129 VHA hospitals during October 2006 to September 2007. COPD related admissions were identified using the principal admission ICD-9-CM codes 491.21 and 518.81 - 84. Pre-existing anxiety and depression were identified using diagnoses from outpatient encounters in the prior 12 months. Primary outcomes were 30- and 365-day mortality. Generalized estimating equations (GEE) were used to adjust mortality for demographics, medical comorbidities, admission laboratory abnormalities, receipt of mechanical ventilation, and to account for clustering within hospitals. Results: Six percent (n = 1,351) of veterans had existing anxiety and 9.5% (n = 2,216) had existing depression. Unadjusted 30-day mortality was higher in patients with anxiety (8.4% vs. 4.6%; p = < .001) and depression (7.2% vs. 4.6%; p < .001). Unadjusted 365-day mortality was higher in patients with anxiety (11.4% vs. 8.9%; p = .003) but was similar in patients with depression (9.8% vs. 9.1%; p = .28). In GEE analyses, adjusted odds of 30-day death were higher for patients with anxiety (OR, 1.66; 95% CI, 1.37 - 2.02; [p < .001]) and depression (OR, 1.52; 95% CI, 1.24 - 1.86; [p < .001). Additionally, adjusted results for 365-day death were higher for anxiety (OR, 1.30; 95% CI, 1.11 - 1.51; [p = .001]) as well as for depression (OR, 1.20; 95% CI, 1.04 - 1.39; [p = .01]. Conclusion: Consistent with extant literature, our results confirm that depression likely contributes to significant elevations of both short- and long-term mortality risks among veterans admitted for COPD related illness. Moreover, this study used a large cohort to identify patients with pre-existing anxiety and found that these patients are likely subject to similar mortality risks. These findings indicate that comorbid anxiety and depression may have an adverse impact on COPD hospital prognosis or may be a marker for unrecognized severity of illness. Efforts should be made to efficiently identify and manage these disorders among hospitalized patients with COPD.