The purpose of this thesis is to analyse the treatment provided to adult medical card holders eligible under the Dental Treatment Services Scheme (DTSS), and to compare it with their need for care as assessed by the National Survey of Adult Oral Health (NSAOH), carried out between October 2000 and June 2002. The study also examines the hypothesis that the structure of the DTSS may provide incentives for patients to over-consume and for dentists to over-provide dental services, as patients within the scheme are provided with dental services at zero monetary cost to themselves, and dentists are remunerated on a fee-per-item basis through a third-party payments system.
A unique dataset is constructed by combining a third-party payments database with the Medical Card Register, which provides individual-level data on demographic, socio-economic, and access-related characteristics of both patients and providers. In a first research application of this dataset, factors influencing the utilisation of the DTSS were determined in a principal-agent framework using a two-part model, logit and 0-truncated negative binomial. The logit estimates that age, gender, parental status and disease levels were significant in the decision to contact the dentist. The 0-truncated negative binomial estimates that age and marital status were significant in determining the amount of services consumed, and that dentist density, age of dentist and type of practice were significant determinants of service provision. Large variations in patterns of service provision were encountered across health boards. Above average treatment content per visit was encountered in health boards with the greatest number of dentists per capita.
Estimates of treatment need were determined from the NSAOH (2000/02). The main findings were that tooth retention had increased and that caries levels and edentulousness had decreased. Medical card holders had the greatest need for care, particularly those aged 65 and over. The predictors of high-risk factors for caries prevalence, estimated by ordered logistic regression, were found to be age, female, regular dental visits, anxiety about the dentist, and primary education level. It was also established that smoking increased the likelihood of having high DMFS scores, in all age groups surveyed.
The comparison between utilisation and need found a disparity between the dental services that medical card holders need and receive. Those who visit the dentist regularly and frequently tended to have their need met and sometimes over-provided, whilst the others have substantial levels of unmet need. There is little evidence of over-consumption of dental services; with low utilisation rates (20%) compared to insured workers (45%). What appears as over-provision, estimated to be about 10% annually, may be explained (assuming a perfect principal-agent relationship) by methodological differences between the survey and actual dental practice behaviour This suggests that the survey has underestimated need by 10%.
However, methodological differences do not explain the variation across health boards, and these variations coupled with a positive association between dentist density and utilisation is suggestive of supplier inducement. From our analysis of need, there is little evidence that the above-average service provision in health boards with high dentist to population ratios is in response to any additional need for care. The patterns of service provision suggest that supplier inducement may exist, but we cannot reasonably assess its scale, or even be conclusive as to its existence. However, the structure of the DTSS is conducive to dentists acting in response to economic incentives. It is recommended that a system of probity be introduced to prevent, detect, and deter the provision of services - other than those based on need.