Geographic variation in poststroke depression among VA stroke patients

Jia H, Ried LD, Wang X, Chuang H, Damush TD, Young LJ, Williams LS. Geographic variation in poststroke depression among VA stroke patients. Poster session presented at: VA QUERI National Meeting; 2008 Dec 11; Phoenix, AZ.

BACKGROUND: Depression after stroke is common. It is estimated that poststroke depression (PSD) occurs among approximately 25-40% of stroke survivors. PSD impedes patients’ functional recovery and quality of life, and it is associated with increased morbidity, mortality, and healthcare use. Currently, there is no documentation available regarding geographic variation in PSD detection. OBJECTIVES: This study is to compare geographic variation in poststroke depression (PSD) detection among a national sample of VA patients with acute stroke across 8 U.S. geographic regions. METHODS: Our study cohort consisted of 5825 VA patients diagnosed with acute stroke in 2001. VA medical and pharmacy data as well as Medicare data were used. ICD-9 depression codes and antidepressant medication dispensing with guideline recommended minimum daily dosage were applied to define patients’ PSD status 12 months post index stroke. Logistic regression models were fit to compare VA PSD diagnosis and overall VA-Medicare PSD detection between the 8 regions as defined by U.S. Bureau of Economic Analysis. STATUS: These findings suggest that PSD in veterans is commonly detected outside the VHA system. Although no regional difference in diagnosis within VA was found, combining VA and Medicare data suggested that veterans in the Plains were more likely to be diagnosed with PSD compared to veterans from other regions of the country. Further research is needed to identify factors associated with these regional differences in PSD detection. FINDINGS: Using VA inpatient and outpatient data alone (ICD-9 codes) may underestimate the rate of PSD detection among the patients. The rate was doubled if VA medical data were used alongside Medicare data and VA pharmacy data: VA medical data alone (21%), VA Medical plus pharmacy data (36%), and VA medical, pharmacy and Medicare (41%). Our logistic regression analyses showed no significant geographic variation in PSD diagnosis within the VA system. Nevertheless, significant geographic variation was present in overall PSD detection (VA plus Medicare), after adjusting for patient demographics and clinical factors. Specifically, compared with patients in Plains, patients in Mid-east (OR 0.6, p < 0.01), South-east (OR 0.8, p < 0.05), and Far-west (OR 0.8, p < 0.05) were less likely to be diagnosed with PSD. IMPACT: VA policy makers need to consider the use of services outside the system by its enrollees with stroke when conducting program evaluation. VA Clinical providers may need to queri patients on diagnoses and prescriptions received outside of the VA system. Thus, clinical tools (e.g., CPRS) may need to be revised to reflect these findings. Future research on PSD among veteran patients should include multiple sources of administrative data in order to more completely describe clinical care and outcomes.