Background: Multiple factors may impact a patient’s response on a depression screen during hospitalization for stroke. The EQ-5D-3L is a standardized measure for use as a health outcome, and comprises the dimension of anxiety/depression. We examined the relationship between pre-hospital history of depression and in-hospital and 3 months anxiety/depression on EQ-5D.
Methods: We captured 3-month outcomes on patients with stroke or TIA discharged home between September 2011 and April 2015. EQ-5D-3L dimension of anxiety/depression at 3 months dichotomized as “no symptoms” or “at least some symptoms” was the primary outcome. Two stepwise logistic regression models adjusting for age, history of congestive heart failure , and baseline NIHSS were generated: 1) using pre-hospital depression as the main predictor (N=124) and 2) using in-hospital depression as the main predictor (N=69 because of missing data in this variable). History of pre-hospital depression and in-hospital depression were highly correlated with each other and were not included in the same model.
Results: The sample included 124 patients (98 with ischemic stroke, 18 hemorrhagic strokes and 8 TIA). The mean age was 66.0 y (SD12.3), 45% women, 73% white, and mean NIHSS 4.0 + 5.2. Anxiety/depression was reported by 30/69 (43%) subjects in the hospital and by 23/124 (18%) at three months post-stroke. In the first model, previous history of depression was associated with 3 months EQ-5D-3L anxiety/depression (p=0.038). In the second model, in-hospital anxiety/depression was associated with 3-month EQ-5D-3L anxiety/depression (p=0.027) (table).
Conclusion: Self-reported depressive and anxiety symptoms in the hospital and history of depression were associated with 3 months post-stroke depression and anxiety. This confirms the importance of screening in subjects with acute stroke for history of depression and for in-hospital depression/anxiety.