|
Activity Number:
|
474
- New Advances in Modeling Survey Data
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 6, 2020 : 10:00 AM to 2:00 PM
|
|
Sponsor:
|
Survey Research Methods Section
|
|
Abstract #314078
|
|
|
Title:
|
Applying Bayesian Methods to Estimate Human Papillomavirus (HPV) Vaccination by Public Health Districts (PHD) in Georgia
|
|
Author(s):
|
David Yankey* and Michael P. Chen and Laurie D. Elam-Evans and Benjamin Fredua and Cassandra S. Pingali and Mary Ann K. Hall and James A. Singleton and Shannon Stokley
|
|
Companies:
|
CDC and Centers for Disease Control and Prevention and Centers for Disease Control and Prevention and Leidos Health Incorporated and Centers for Disease Control and Prevention and Eagle Global Scientific and Centers for Disease Control and Prevention and Centers for Disease Control and Prevention
|
|
Keywords:
|
National Immunization Survey;
Vaccination;
Public health District;
HPV;
Bayesian;
MCMC
|
|
Abstract:
|
We applied Bayesian hierarchical methods and spatial effects at the Public Health District (PHD) and county levels to estimate human papillomavirus (HPV) vaccination initiation (?1 HPV dose) coverage in Georgia’s 18 PHDs. We used 2015–2018 National Immunization Survey-Teen (NIS-Teen) data and the 2018 Area Health Resource File data. Bayesian methods allow indirect estimation with small sample sizes, missing values, and covariates. We computed estimates using two different Bayesian approaches for model fitting. The Markov chain Monte Carlo (MCMC) method uses simulation methods while the Integrated nested Laplace approximation (INLA) uses approximation methods within the class of latent Gaussian models. Crude estimates for HPV vaccination initiation coverage range from 44.4 (95% CI: 28.8–61.1) in Waycross/Southeast PHD to 78.4 (67.2–86.6) in Rome/Northwest PHD; average coverage in Georgia was 63.1 (59.9–66.3) in 2015–2018 combined. We compared estimates from these two approaches to the crude estimates to determine whether one method outperformed the other. This study will also elucidate ?1 HPV coverage disparities between PHDs and can inform PHD-specific intervention needs.
|
Authors who are presenting talks have a * after their name.
