Comparison of self-reported fecal occult blood testing with automated laboratory records among older women in a health maintenance organization.

Screening guidelines for colorectal cancer recommend annual fecal occult blood (FOB) testing for adults aged 50 years and older. Self-reported history of screening is frequently the sole source of data available to researchers and clinicians. This study validated FOB testing in a sample of 1,021 older women. Testing rates based on self-reported data exceeded rates based on computerized laboratory records by 13.9%. Agreement was moderate (kappa = 0.52; 95% confidence interval 0.47, 0.58). Sensitivity was 0.92 and specificity 0.58. Logistic regression analysis showed that older age and physician encouragement for FOB testing were associated with accurate recall (p<0.05). Self-report is the most commonly available information about the occurrence and timing of cancer detection procedures. These data suggest cautious use of self-reported screening by FOB for clinical decision making and for research and surveillance.