View in EDS

Estimating county-level dental care utilization among adults in California using multilevel modeling with raking approach

Saved in:
Bibliographic Details
Title: Estimating county-level dental care utilization among adults in California using multilevel modeling with raking approach
Authors: Yilan Huang, Honghu Liu
Source: Archives of Public Health, Vol 83, Iss 1, Pp 1-11 (2025)
Publisher Information: BMC, 2025.
Publication Year: 2025
Collection: LCC:Public aspects of medicine
Subject Terms: Small area Estimation, Raking, Multilevel model, Dental care utilization, Public aspects of medicine, RA1-1270
Description: Abstract Background Regular dental visits are essential for oral health, yet disparities between regions exist due to socioeconomic and geographic factors. While national surveys provide valuable data on dental care utilization, they generally lack sufficient sample sizes at the local level to generate reliable county-level estimates. Small area estimation techniques, such as multilevel regression and post-stratification (MRP), can help address this gap by producing robust estimates for smaller geographic areas. However, the MRP approach relies on detailed population data in the form of joint distributions and cannot be applied when only marginal distributions are available. Methods This paper introduces a hybrid approach combining multilevel modeling with the raking procedure. We used individual-level data from the 2018 Behavioral Risk Factor Surveillance System (BRFSS) and census data from American Community Survey to estimate county-level dental care utilization among adults in California. Results The county-level dental care utilization in California ranged from 52.5 to 73.1%, with a median of 63.1%. Our model-based estimates matched direct BRFSS estimates at metropolitan and micropolitan statistical area levels. Furthermore, we found significantly positive correlations between our model-based estimates and direct estimates from the California Health Interview Survey for 41 counties (Pearson coefficient: 0.801, P
Document Type: article
File Description: electronic resource
Language: English
ISSN: 2049-3258
Relation: https://doaj.org/toc/2049-3258
DOI: 10.1186/s13690-025-01673-6
Access URL: https://doaj.org/article/5581491c0a474b5cb42959f87d1bcb76
Accession Number: edsdoj.5581491c0a474b5cb42959f87d1bcb76
Database: Directory of Open Access Journals
Description
Abstract:Abstract Background Regular dental visits are essential for oral health, yet disparities between regions exist due to socioeconomic and geographic factors. While national surveys provide valuable data on dental care utilization, they generally lack sufficient sample sizes at the local level to generate reliable county-level estimates. Small area estimation techniques, such as multilevel regression and post-stratification (MRP), can help address this gap by producing robust estimates for smaller geographic areas. However, the MRP approach relies on detailed population data in the form of joint distributions and cannot be applied when only marginal distributions are available. Methods This paper introduces a hybrid approach combining multilevel modeling with the raking procedure. We used individual-level data from the 2018 Behavioral Risk Factor Surveillance System (BRFSS) and census data from American Community Survey to estimate county-level dental care utilization among adults in California. Results The county-level dental care utilization in California ranged from 52.5 to 73.1%, with a median of 63.1%. Our model-based estimates matched direct BRFSS estimates at metropolitan and micropolitan statistical area levels. Furthermore, we found significantly positive correlations between our model-based estimates and direct estimates from the California Health Interview Survey for 41 counties (Pearson coefficient: 0.801, P
ISSN:20493258
DOI:10.1186/s13690-025-01673-6