Cost-effectiveness analysis of dental sealant using econometric modeling

Abstract

Dental sealants have been shown to be effective in reducing caries. Most believe sealants are still underused, particularly among children who need sealants the most. Because the caries prevalence continues to decline, and the disparities in children's dental heath and dental care still exist, more efficient sealant placement strategies should be implemented based on the scientific information from cost-effectiveness analysis (CEA) of dental sealant at the community level. Previous CEAs of sealant using observational datasets were unreliable because they did not address the self-selection problem. The objectives of this study are to examine the utilization of dental sealants and its determinants, evaluate the incremental effectiveness and expenditure associated with sealant placement after correcting the potential selection issue, and explore the differences in sealant's cost-effectiveness among subpopulations. This study mainly utilized enrollment data and encounter data from a large Health Maintenance Organization in Minnesota. The study sample included 3,700 children aged 6 to 17 years during 1997 to 2001 and were continuously enrolled for 5 years. They all had a caries risk assessment (CRA), which was conducted at the beginning of the observation period, and no prior caries record for their included first permanent molars (FPMs). The CRAs were classified into three scores: low, moderate and high risk. Information on the 64 dentists who participated in the study was linked to the encounter data to identify those who conducted CRAs. Outcome variables included discounted effectiveness, as measured by the duration of caries-free state (healthy months) of a FPM, and the discounted cost associated with caries treatments, within the study period. The key independent variables included demographic variables (e.g., age, gender, race), caries risk level, socio-economic status, sealant placement, and preventive care utilization. Bivariate analysis and logistic analysis were performed to examine the pattern of sealant utilization and identify the determinants affecting sealant placement decision. Econometric models including classic Tobit model, selectivity-corrected Tobit model, classic two-part model, and selectivity-corrected two-part model were used to examine the selection issue and obtain unbiased marginal effects of sealant on caries-free duration and caries-related treatment cost. The working experience of the dentists who conducted the initial CRAs was used as the instrumental variable. The bootstrap method was used to obtain standard errors and confidence intervals for the incremental cost-effectiveness ratios. Sensitivity analysis and subgroup analysis were performed. In this study, approximately 77% of the sample had one or more FPM sealed during the entry period, more than half of them had all four FPMs sealed. Children aged 6 to 8 were more likely to receive sealant than children aged 9 or older. Children at relatively high caries risk, as well as children who visited dentists for preventive care more than once a year, had greater odds of receiving sealants. Non-white children or those from families with low incomes or low education level were more likely to receive sealant. This study also identified some dentists' characteristics, such as age, gender, and working experience as the good predictors of sealant decision. After 5 years, the sealant group had more individuals (83.9%) and more FPMs (94.3%) that stayed healthy compared with the non-sealant group in which 83.1% of the sample individuals and 91.8% of FPMs stayed healthy. A sealed FPM was associated with $56.84 expenditure (initial sealant charge was $39.00) over 5 years, and an unsealed FPM was associated with $13.13 expenditure. The sample-average incremental cost-effectiveness ratio (ICER) was $38/caries-free month for each FPM. Based on the results from econometric models, sealants were associated with a lower probability of having any caries, longer caries-free duration, an increased probability of using any resource, and less resources consumption. The final ICER indicates that sealant cost $42.16 more than non-sealant treatment to get one more caries-free month for each FPM. The 95% CI was $22.64 to $85.40 per one more caries-free month for each FPM. Significant selection or endogeniety issue was not found in either the effectiveness or cost analysis based on the whole sample, but it existed when analyzing sealant effects among certain subgroup children. The results from subgroup analysis show that sealing children at high risk for caries appears to be highly cost effective. In contrast, sealing children at low risk for caries would be much less cost effective. Sealing the FPMs of infrequent utilizers of preventive care appears to be more cost effective than frequent utilizers of preventive care. There is no significant difference in ICERs between sealing younger children and sealing older children. In conclusion, sealant application is not always cost effective. A uniform and fixed sealant utilization goal may not be appropriate. Sealant application should be increased among the high risk populations, such as those with previous caries or low dental care utilizers, or those directly deemed at high caries risk by dentists. The caries risk assessment procedure can improve clinical decisions on sealant application and increase efficient sealant delivery.