Special Article
February 28, 2011

Assessing Risk Factors as Potential Screening TestsA Simple Assessment Tool

Author Affiliations

Author Affiliations: Wolfson Institute of Preventive Medicine, Barts and the London School of Medicine and Dentistry, Queen Mary University of London, Charterhouse Square, London, England.


Copyright 2011 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2011

Arch Intern Med. 2011;171(4):286-291. doi:10.1001/archinternmed.2010.378

Most risk factors are not useful as screening tests for the diseases with which they are associated because the association is too small, even if they are large enough to have important causal implications.1 A risk factor (which may be causal or noncausal) has to be very strongly associated with a disease to be seriously considered as a possible screening test. For example, the odds ratio between the highest and lowest 20% of the population needs to be about 50 or more. Despite this, it is often suggested that a particular risk factor may be a useful screening test or disease predictor, even though it can be deduced from published relative risk estimates that this cannot be the case. For example (considered further in the subsection titled “Example 1: CRP as Possible Test for CHD” in the “Four Examples” section), we show that C-reactive protein (CRP) is not useful as a screening test, even though it has been suggested as a possible predictor of ischemic heart disease.2

First Page Preview View Large
First page PDF preview
First page PDF preview