Jonathan D Mahnken1, Daniel H Freeman, Anthony R DiNuzzo, Jean L Freeman. 1. Department of Preventive Medicine and Public Health, Center for Biostatistics & Advanced Informatics, University of Kansas Medical Center, MSN 1008, 3901 Rainbow Blvd, Kansas City, KS 66160, USA. jmahnken@kumc.edu
Abstract
OBJECTIVES: To determine the accuracy of self-reported mammography use and generate a correction factor that adjusts for reporting errors among Mexican-American women. DESIGN: Self-reported mammography use was compared with a medical chart review. PARTICIPANTS: A random sample of 199 Mexican-American women aged 50-74 in southeast Texas. MAIN OUTCOME MEASURES: Self-reported mammography use for the preceding 2 years and mammography use determined by medical chart review. RESULTS: Positive and negative predictive values of mammography self-reports in the previous 2 years were 73.9 and 86.0%, respectively. The correction factor was 79.6%, or approximately 80% of the self-reported mammograms within the preceding 2 years could be verified through medical chart reviews. CONCLUSIONS: Mammography use rates among Mexican-American women were not as high as self-reports indicated. Estimates need to be adjusted downward by approximately 20%.
OBJECTIVES: To determine the accuracy of self-reported mammography use and generate a correction factor that adjusts for reporting errors among Mexican-American women. DESIGN: Self-reported mammography use was compared with a medical chart review. PARTICIPANTS: A random sample of 199 Mexican-American women aged 50-74 in southeast Texas. MAIN OUTCOME MEASURES: Self-reported mammography use for the preceding 2 years and mammography use determined by medical chart review. RESULTS: Positive and negative predictive values of mammography self-reports in the previous 2 years were 73.9 and 86.0%, respectively. The correction factor was 79.6%, or approximately 80% of the self-reported mammograms within the preceding 2 years could be verified through medical chart reviews. CONCLUSIONS: Mammography use rates among Mexican-American women were not as high as self-reports indicated. Estimates need to be adjusted downward by approximately 20%.
Authors: Vanessa B Sheppard; Judy Wang; Bin Yi; Toni Michelle Harrison; Shibao Feng; Elmer E Huerta; Jeanne S Mandelblatt Journal: J Gen Intern Med Date: 2008-10-07 Impact factor: 5.128