Information on the proportion of adults and mean number of teeth with estimated treatment need for extractions and restorations was obtained from a clinical examination conducted as part of an epidemiological national survey, and information on treatment provided amongst matched age groups was obtained from administrative (claims) data. An advantage of using claims databases to measure treatment provided is that the information represents real-world dentistry. Each dentist has his/her own approach to treatment, and patients have different perceived needs and lifestyle preferences. These data represent the true complexity of what occurs daily in dental surgeries . The similarities in mean number of teeth present between the claims data and the survey data for employed adults instil confidence in the representativeness of the survey sample.
As in Wanman and Wigren, this was essentially an evaluation of treatment need from two points of view . The first was a professional assessment on a random sample of employed and less well-off adults who claimed to attend the dentist regularly, and was based on an examination made by independent dentists (epidemiologically estimated need), where their only consideration was the subject’s oral status. The second was the treatments provided to employed and less well-off adults who used the DTBS and DTSS schemes (evaluated need). For the latter, factors such as aesthetics, cost, and patients’ perceived needs and preferences were also considered in treatment planning.
There was a lack of agreement between mean estimated treatment need and mean treatment provided in all age groups, especially among 16-24 and 65+ year-old less well-off adults and 35-44 year-old employed adults. Although treatment provided was greater than estimated need in some cases, it is important to note that the dentist providing the service had recourse to radiographs, and could therefore offer a more thorough clinical examination than for the national survey. In the national survey, estimated need for advanced restorations included endodontics, crowns, bridges and veneers. However, the only advanced restoration covered by the DTSS (the scheme for less well-off adults) was endodontic treatment for anterior teeth. As endodontics for other teeth were not covered, these less well-off adults may have chosen extraction of compromised teeth rather than incur the expense of advanced restorations. This may explain why the proportions of less well-off 16-24 and 35-44 year-olds who had teeth extracted, was greater than estimated as needed. In addition, Millar and Locker  found that people in low-income households were less likely than those in high-income households to mention preventive reasons for visiting a dentist. Extractions have been found to be more likely when the reason for a visit is pain , whereas visiting the dentist for a check-up, instead of when in need or pain, is associated with increased retention of natural teeth .
There are two possible measurement reasons for the gap between estimated need for restorations and the mean number of restorations provided in the schemes. First, mean need for restorations was calculated regardless of number of surfaces involved, however, patients may have been provided with restorations on different surfaces at several visits to a dentist during the period of analysis. Second, restoration repair may have been recorded as ‘other’ in the survey (as explained in the methods section) and as a ‘restoration’ in the claims databases.
Other studies also found disparities between assessment of dental treatment need and the treatment actually provided [8–10]. Nuttall  found a large discrepancy between need for dental treatment recorded in an epidemiological survey and the clinical treatment that was subsequently provided (for the same subjects) in the General Dental Service in Scotland. He suggests that the results “cast doubt upon the usefulness of the epidemiological survey as a tool for predicting restorative treatment” . Naegele and colleagues found that more teeth were treated by fee-for-service dentists, based on a thorough routine dental check-up, than predicted as needed by salaried dentists (within six months) . Wanman and Wigren  also question the validity of epidemiological assessment of treatment needs. They compared professionally assessed treatment need in an epidemiological survey with treatment provided in the Public Dental Service in Sweden, and found a significantly higher frequency of restorative treatments provided than the assessed need, especially among 65+ year-olds . Similar results were found in this study, where more restorations were provided than estimated as needed in the survey, across all age groups.
Clarkson and colleagues suggest that the lack of agreement between what dental epidemiologists observe and the treatment that dentists provide may be due to the more complex nature of treatment decisions made by dentists compared to the diagnostic criteria used in conventional epidemiological studies . According to Sheiham and colleagues, a more realistic assessment of treatment needs should include “the functional and social dimensions of dental disease, and an assessment of the social motivational factors which predispose people towards dental ill health and influence the effectiveness of treatment and health education” . In evaluating need in the Irish national survey, no consideration was given to the patient’s financial situation or whether he/she wanted treatment, whereas both patient and oral health factors were considered in the provision of treatment.
According to Schonfeld, gaps between treatment need and treatment provided may indicate requirements for additional manpower, an increase in productivity of existing manpower, or a change in the pattern of dental care . Grembowski and colleagues suggest that under fee-for-service reimbursement, dentists’ efforts to build financially successful practices may encourage over-treatment . Where dental services are provided at zero monetary cost to the patient, as for less well-off adults in Ireland, there may be an incentive for patients to over-consume or dentists to over-provide treatments. According to Woods , if there is evidence of either over-consumption and/or over-provision of services, for particular treatments, or to certain groups, then resources should be diverted from areas of excess provision to groups with greatest need.
Although the gap between epidemiologically estimated need and treatment provided seems to be greatest for less well-off adults, we do not know if the differences are related to dentist or patient factors, and therefore we cannot determine whether over- or under-treatment occurred. As in health care generally , variations in dental treatment arise from the interaction between supply and demand, which depend on the preferences and perceptions of both patients and dentists; therefore, any differences are probably due to several factors.
This study is concerned with the relative validity of epidemiologically assessed treatment need for adults using two different dental care delivery schemes, one a ‘free’ service for less well-off adults and the other a co-payment scheme for employed adults. The differences between epidemiologically estimated need and treatment provided could also provide an indication of accessibility to dental services for the three age groups in the two schemes. However, differences by socio-economic groups do not automatically reflect inequities . Those in equal need and with equal opportunities to access health care may not make equal use of those opportunities. Nonetheless, an unacceptable reason for differences in use of health care would be that some individuals may be less capable of taking advantage of health care services .
In accordance with traditional demand theory, demand for oral health depends on its price per unit, constraining income, the price of all other commodities, and the value people place on oral health as a source of consumption benefit . Income level is associated with utilisation of dental care services . Consumers must allocate their income between buying dental care and other commodities . The price of dental care consists of an out-of-pocket payment and other costs such as travel costs, opportunity cost of the time devoted to dental care, and non-monetary costs (such as time and psychological costs) . Sintonen and Maljanen refer to these as the ‘shadow price of dental care’ . Although treatments are provided at a subsidised rate to employed adults in Ireland, people may still feel the cost is prohibitive, especially for advanced restorations such as crowns. This may explain why the proportion of employed adults receiving advanced restorations was significantly less than the epidemiologically estimated need for the 35-44 and 65+ age groups.
Perceived need has been found to be a stimulus for regular attendance . The large gap between mean number of teeth estimated as needing extractions and teeth extracted among less well-off 65+ year-olds could reflect a difference between need and demand for treatment. People may not feel they need treatment (low perceived need), and those found as needing treatment in the survey may not have visited a dentist under the scheme during the period. According to Holm-Pedersen and colleagues, professionally assessed need for dental treatment, based solely on clinical diagnosis, often leads to an “overestimation of the true need for treatment, especially among frail and functionally dependent elderly people, some of whom do not want treatment, either because there is no perceived need or no expressed demand” . Perceived need may be increased or decreased through, for example, health education programs or changing financial incentives to seek services . We should encourage people to visit their dentist for a check-up rather than waiting until they feel pain. Future surveys should include questions on perceived dental treatment needs, as this would provide further insight into the gap between epidemiologically estimated need and treatment provided.
In this study, we compared adults who said they were regular users in a survey with treatment provided to those who used the schemes. Although the most recent national survey of adult oral health in the Republic of Ireland was conducted 10 years ago, by comparing it with utilisation data of the same time, we feel that our findings are still relevant today. We were unable to compare treatments provided to the same people who were examined in the survey; this would have enabled us to measure whether there was unmet need or variance with provision. However, given the confidentiality issues and difficulties obtaining agreements to link survey and administrative data, we feel that the method used in this paper was a valuable alternative.
Another limitation is that we do not know why the treatments were provided, for example, whether a restoration was provided for aesthetic reasons or due to caries. The DTBS and DTSS databases could be improved by the inclusion of a field, on the claim form, for reasons for provision of treatments; this would provide a more accurate indication of dental health. Administrative databases are a largely untapped resource for analysis of treatments provided, and this study demonstrates their utility. Holtz and colleagues recommend that comparisons across survey and administrative data sources be encouraged, rewarded, and funded so that limitations can be reduced or removed .