Health Services Research & Development

Veterans Crisis Line Badge
Go to the ORD website
Go to the QUERI website

2006 HSR&D National Meeting Abstract


3044 — Brain MRI in Acute Ischemic Stroke (IS): Do New Approaches Diffuse Equally to Blacks and Whites?

Author List:
Levine DA (Deep South Center on Effectiveness)
Allison JJ (Deep South Center on Effectiveness)
Kirk KA (Deep South Center on Effectiveness)
King PH (Birmingham VA Medical Center)
Houston TK (Deep South Center on Effectiveness)
Safford MM (Deep South Center on Effectivness)
Williams LS (Roudebush VA Medical Center)
Kiefe CI (Deep South Center on Effectiveness)

Objectives:
Blacks hospitalized for acute ischemic stroke (IS) may receive radiologic imaging less frequently than whites. The use of brain magnetic resonance imaging (MRI) in IS patients is increasing, without VA guidelines for its use. We assessed whether there is significant racial variation in brain MRI of veterans hospitalized for IS.

Methods:
During FY 2001-03, we identified 8503 consecutive non-transferred black or white patients aged 40-85 years discharged from any VA medical center with a primary diagnosis of IS (ICD-9 codes 434.xx and 436.xx) using DSS National Data Extracts and Medical SAS Datasets (PTF and OPC). The 18 VISNs with race identified in >88% patients were included. We compared the frequency of in-hospital brain imaging by race. With multivariable logistic regression, we adjusted associations between brain imaging and race for age, sex, length of stay (LOS), prior cerebrovascular disease, hypertension, hyperlipidemia, diabetes, and VISN (region) and we assessed race-VISN interaction.

Results:
Blacks (n=2265) were more likely than whites (n=6238) to be younger than age 65 (54% vs 40%, P<0.0001) and have greater LOS, mean 9.7 (SD13) vs. 7.9 (SD 17) days. Hypertension was more common in blacks (91% vs. 87%, P<0.0001), hyperlipidemia less common (53% vs. 63%, P<0.0001), but prior cerebrovascular disease (30%) and diabetes (45%) had similar prevalence by race. Brain MRI was performed more often in blacks than whites (64% vs. 53%, P<0.0001) as was head CT (80% vs 77%, P=0.0004). Racial differences in brain MRI varied significantly by VISN (P<0.0001) with the likelihood of brain MRI for blacks increased in 12 VISNs and decreased in 4 VISNs. However, ignoring VISN-race interaction, blacks still received brain MRI significantly more than whites, adjusted odds ratio 1.5, 95%CI (1.3-1.7).

Implications:
Black veterans with IS receive brain MRI significantly more than whites. Further work is warranted to understand the reasons for these racial differences in brain MRI.

Impacts:
Research is necessary to determine whether black IS patients benefit from increased brain MRI testing and to understand how the distribution of imaging resources impacts stroke care delivery. Guidelines and studies regarding the appropriate use of specific brain imaging modalities in IS also are needed.