Date of Award

1-10-2003

Document Type

Open Access Thesis

Degree Name

Medical Doctor (MD)

First Advisor

Walter Kernan

Abstract

To identify risk factors for delayed diagnosis of subarachnoid (SAH) or intracerebral (ICH) hemorrhage in younger patients (ages 18 - 49), we performed a subsidiary case-control study among 702 subjects from the Hemorrhagic Stroke Project (HSP). Case subjects were the 54 HSP patients (7.7%) who did not receive an appropriate diagnostic evaluation (brain CT scan, and if CT negative, LP) within 24 hours of consulting a physician for symptoms consistent with hemorrhagic stroke. For each case subject with a delayed diagnosis, we identified two subjects from the HSP with a prompt diagnosis, all successfully matched on recruitment site. We calculated odds ratios (ORs) for the association between risk factors and delayed diagnosis. Four features were associated with risk for delayed diagnosis (criteria: OR =2.0 or =0.5): initial evaluation in a physician's office (OR=23.1), absence of alarm symptoms (photophobia, loss of consciousness, focal weakness) (OR=4.6), no effortful activity preceding focal time (OR=5.5), and Hispanic ethnicity (OR=2.4). Risk factors associated with delayed diagnosis were different in patients who presented to the hospital compared to patients who presented to an office. In a separate analysis, risk factors associated with delayed diagnosis were different for patients with SAH compared with patients with ICH. In conclusion, patients with SAH or ICH are at greater risk for a delay in their diagnosis if they do not have alarm symptoms, especially if they present to an office rather than a hospital emergency department. Efforts to reduce delayed diagnosis should be directed at primary care physicians who see low-risk patients with milder symptoms. Our findings also indicate that there are different risk factors associated with delayed diagnosis of hemorrhagic stroke based on location of presentation and on stroke type. Future research to understand and prevent delayed diagnosis must consider location of presentation and stroke type.

Comments

This is an Open Access Thesis.

Download
Open Access

This Article is Open Access

Share

COinS