Table 2.
Individual and community characteristics associated with dental service utilisation.
| Variables | Model 0 a | Model 1 b | Model 2 c |
|---|---|---|---|
| Fixed-effects (OR with 95% CI) | - | - | - |
| Individual-level variables | - | - | - |
| Age | - | - | 1.007 (1.003–1.011) |
| Sex (ref: female) | - | - | - |
| Male | - | - | 0.844 (0.735–0.969) |
| Current marital status (ref: not married) | - | - | - |
| Married | - | - | 0.718 (0.643–0.801) |
| Educational attainment (ref: elementary school or below) | - | - | - |
| Junior high school | - | - | 1.449 (1.236–1.698) |
| Senior high school | - | - | 2.065 (1.780–2.396) |
| Higher education | - | - | 4.189 (3.532–4.969) |
| Economic status (ref: quintile 1, poorest) | - | - | - |
| Quintile 2 | - | - | 1.207 (0.999–1.458) |
| Quintile 3 | - | - | 1.414 (1.173–1.704) |
| Quintile 4 | - | - | 1.764 (1.467–2.120) |
| Quintile 5 (richest) | - | - | 2.034 (1.686–2.453) |
| Health insurance (ref: no) | - | - | - |
| Yes | - | - | 1.302 (1.177–1.440) |
| Dental pain (ref: no) | - | - | - |
| Yes | - | - | 3.320 (2.959–3.726) |
| Self-reported mouth ulcers (ref: no) | - | - | - |
| Yes | - | - | 1.284 (1.143–1.443) |
| Self-rated health status (ref: unhealthy) | - | - | - |
| Healthy | - | - | 0.963 (0.852–1.089) |
| Unmet healthcare needs (ref: no) | - | - | - |
| Yes | - | - | 0.819 (0.713–0.941) |
| Smoking status (ref: never) | - | - | - |
| Former | - | - | 1.048 (0.828–1.327) |
| Current | - | - | 0.858 (0.737–0.999) |
| Community-level variables | - | - | - |
| Residential area (ref: rural) | - | - | - |
| Urban | - | 2.118 (1.748–2.565) | 1.475 (1.249–1.742) |
| Cognitive social capital (ref: low) | - | - | - |
| High | - | 1.002 (0.845–1.188) | 1.048 (0.905–1.212) |
| Structural social capital (ref: low) | - | - | - |
| Medium | - | 1.238 (1.004–1.527) | 1.173 (0.981–1.404) |
| High | - | 1.679 (1.339–2.104) | 1.449 (1.194–1.759) |
| Random effects | - | - | - |
| Community random variance (SE) | 0.585 (0.068) | 0.346 (0.046) | 0.196 (0.034) |
| VPC | 0.151 | 0.095 | 0.056 |
| PCV (%) | Reference | 40.9% | 66.5% |
Note: Results were based on multilevel logistic regression analyses, with non-users as a reference group. Bold values represent significant associations. The significance level was set at a p-value < 0.05. OR, odds ratio; CI, confidence interval; SE, standard error; VPC, variance partition coefficient; PCV, proportional change in variance. a Model 0 was fitted without explanatory variables. b Model 1 was adjusted with community variables. c Model 2 was adjusted with individual and community variables.