Objectives: To evaluate the performance of subjective self-assessment of dental care need with objective oral disease status in a sample of Hispanics/Latinos.
Methods: Data from 13,561 participants of the 2008-2011 Hispanic Community Health Study/Study of Latinos were analyzed using complex survey procedure in SAS 9.4. Subjective self-reports of types of dental care needed (check-up or cleaning, teeth filled, teeth pulled, gum treatment, denture repair) were validated against objectively determined oral disease status [dental decay, periodontitis, bleeding on probing (BOP) and count of missing teeth].
Results: Individuals who reported needing a cleaning/checkup were twice as likely to present with BOP upon oral examination (POR = 2.01, 95% CI: 1.54, 2.63). Similarly, individuals who reported needing gum treatment were more likely to present with periodontitis (POR = 1.96, 95% CI: 1.71, 2.24) and BOP (POR = 2.70, 95% CI: 2.37, 3.07) upon oral examination. In multivariable prediction modeling, demographic factors and subjective dental care measures were associated with the respective oral disease states. Moreover, the sensitivity, specificity, and area under the ROC curve for the count of missing teeth were, respectively, 77 percent, 57 percent, and 0.84, while the positive predictive value (PPV) was 26 percent.
Conclusions: In bivariate analysis, self-reported type of dental care needed appear indicative of actual oral disease state and may be of value for the surveillance of oral diseases when clinical measures are unattainable. In multivariable prediction modeling, these subjective measures had low PPVs thus limiting the generalizability of our findings. Nonetheless, validation and refinement of these constructs in other populations is warranted.
Keywords: HCHS/SOL; Hispanic Americans; Oral health; ROC curve; epidemiology; public health; self-report.
© 2020 American Association of Public Health Dentistry.