Purpose: We assessed the accuracy of human papillomavirus (HPV) vaccination status based on adult proxy recall and household immunization records for adolescent females in the United States.
Methods: We used data from the 2010 National Immunization Survey-Teen for females aged 13 to 17 years. The accuracy of HPV vaccination status (≥1 dose) based on adult proxy recall (unweighted n = 6868) and household immunization records (unweighted n = 2216) was assessed by estimating the sensitivity, specificity, and corresponding 95% confidence limits (CL) of these measures with provider-reported HPV vaccination status as the reference standard. Our analyses accounted for the complex survey design and population weights.
Results: The sensitivity and specificity of adult proxy recall were 83.9% (95% CL: 81.2%, 86.6%) and 90.4% (95% CL: 88.9%, 92.0%), respectively. Conversely, the sensitivity and specificity of household immunization records were 74.2% (95% CL: 69.1%, 79.2%) and 98.0% (95% CL: 96.8%, 99.1%), respectively. The accuracy of both measures varied by race/ethnicity, proxy respondent, and maternal education.
Conclusions: Our results suggest that adult proxy recall and household immunization records have reasonable accuracy for classifying HPV vaccination status for females aged 13 to 17 years in the United States, but these measures present a trade-off between sensitivity and specificity.
Copyright © 2013 Elsevier Inc. All rights reserved.