Accuracy of Self-reported Abnormal Pap Smears Among Reproductive-age African-American Women.

Abstract

Self-reported history of abnormal Pap smear is used in large public health surveys. However, literature on the accuracy of this reporting is limited. We sought to assess the validity of self-reported abnormal Pap history in a community-based sample of African-American women 24-37 years of age in the Detroit, MI, recruited from 2010 to 2012. We compared self-reported data on 2-year history of abnormal Pap smear to medical Pap record data (the gold standard) obtained from eligible participants. Sensitivity and specificity were evaluated to summarize the accuracy of the self-reported data. We also explored the sensitivity by severity of abnormality. We identified 345 participants with 480 eligible Pap records. Thirty-five percent of abnormal Pap results were not identified by self-report (sensitivity/specificity: 65%/91%; kappa = 0.54). The sensitivity of reporting an abnormal Pap tended to be higher for those with a more severe abnormality. A large proportion of abnormal Pap smears were not identified by self-report in this sample of African-American women. Public health studies utilizing self-reported abnormal Pap information should be interpreted with caution.