Related Experiment Videos
Clinical prediction rules for computed tomographic scanning in senile dementia
Archives of Internal Medicine
|January 1, 1987
Summary
This study evaluated four prediction rules for head computed tomography (CT) in dementia diagnosis. One rule showed higher sensitivity and reduced uncertainty but had a high misclassification rate.
Area of Science:
- Geriatric Medicine
- Neurology
- Radiology
Background:
- The utility of head computed tomography (CT) in diagnosing dementia is debated.
- Several prediction rules exist to guide selective CT use in dementia evaluation.
Purpose of the Study:
- To validate four existing prediction rules for head CT in dementia assessment.
- To determine the most effective rule for guiding selective CT scanning.
Main Methods:
- A validation study was conducted in an outpatient geriatric assessment unit.
- Four prediction rules were assessed for diagnostic sensitivity, specificity, misclassification rates, and information content.
Main Results:
- Sensitivity ranged from 12.5% to 87.5%, specificity from 37.2% to 77.9%, and misclassification rates from 23.5% to 60.8%.
- One rule demonstrated significantly higher sensitivity and reduced diagnostic uncertainty more effectively than others.
- This superior rule was complex and associated with a high misclassification rate.
Conclusions:
- Existing prediction rules for head CT in dementia have variable performance.
- Further development is needed to create a clinically useful rule for selective CT scanning in dementia evaluation.