The Utilization of Dental Services Among Indonesian Youth: A Nationwide Study Based on the Expanded Andersen Behavioral Model

Abstract

Background The purpose of this study was to examine factors associated with dental service utilization among Indonesian youth using the Expanded Andersen Behavioral Model (EABM). Methods This cross-sectional study assessed data from 3332 people aged 15 to 24 from the 2023 Indonesian Health Survey. Using EABM, variables were classified as predisposing, enabling, need, and psychosocial domains. Dental service utilization within the previous year was the outcome variable. A series of appropriate statistical analysis, such as chi-square, binary logistic regression, multivariable logistic regression, and principal component analysis (PCA) with varimax rotation was conducted to analyse the 2023 Indonesian Health Survey data. Logistic regressions were reported as odds ratios (ORs) with 95% confidence intervals (CIs). PCA was conducted to examine clustering patterns among the included variables and to assess whether these patterns were broadly consistent with the conceptual domains. Results Overall, only 7.6% of respondents reported using dental services within the previous year. Female (adjusted OR [AOR] [95% CI] = 1.655 [1.236-2.154]), urban residence (AOR [95% CI] = 2.274 [1.689-3.058]), and health insurance ownership (AOR [95% CI] = 1.985 [1.383-2.849]) were significantly associated with higher utilization rates. Factors that showed the strongest associations were oral pain (AOR [95% CI] = 12.630 [7.747-20.582]) and caries experience (AOR [95% CI] = 2.342 [1.467-3.741]). In contrast, psychosocial factors, including depression and psychological distress, were not significantly associated with utilization. PCA identified four components corresponding to psychosocial, need, predisposing, and enabling domains, which explained 60.24% of total variance, supporting the conceptual structure of EABM. Conclusion Dental service utilization among Indonesian youth remains low and is primarily driven by clinical need and structural access factors rather than psychosocial conditions. While psychosocial variables were in line with a coherent latent construct, they did not translate into observable utilization behaviour.