An overview of the methodological aspects and policy implications of willingness-to-pay studies in oral health: a scoping review of existing literature

In joint decision making between patient and health care providers, preferences of patients for proposed procedures are at least as important as clinical norms [1, 2]. On the other hand, demands for dental services are beyond the capacities of healthcare systems particularly in a majority of low and middle income countries [3].

Policymakers and health care managers must be informed about the different types of health care benefits perceived by patients, as well as factors influencing their services utility, in order to allocate optimal resources and generate favorable incentives within the healthcare systems [4, 5]. Patient valuations should include factors beyond service effectiveness, such as time span, discomfort, pain and anxiety [2].

“Willingness-to-pay” (WTP) is a systematic and trustworthy method in monetary terms to measure the benefit of a health care intervention [4]. WTP seeks to measure the preference strength of an individual for any desired intervention by calculating the maximum amount of money they would sacrifice [6]. This is considered the manifestation of “direct democracy” in public policymaking [7]. The technique is primarily seen as an aid to place monetary value on health care programs and to compare them particularly with programs beyond the health sector [8, 9]. WTP results can serve as the benefit wing of the economic evaluation (cost–benefit analysis) of a desired service, compared to other alternatives, which is strongly recommended as a critical input at the public level for decisions of allocation [10]. Therefore, WTP can help tailor dental treatment to individual patient valuations in a clinical practice, or help policymakers make informed decisions regarding resource allocation in the public sector and a priority setting across patient groups [8, 9].

WTP can be measured in two main ways: First, the “revealed preferences” approach which focuses on consumer behavior in the market and can be measured based on the information acquired from actual real market purchase of individuals and second, “stated preferences”; an indirect technique in which consumers are asked to explicitly state their WTP [11, 12]. “Contingent Valuation” (CV) is a WTP survey asking participants to state their maximum WTP for a hypothetical item [11].

WTP can be elicited through interview or questionnaire but to achieve a more valid result it is recommended to use the face-to-face interview method [13]. WTP is versatile and can be used for inquiring about a public service such as water fluoridation for a state, or it can be a non-public service such as tooth filling service provisions in a private clinic. There are several methods to elicit the WTP amounts: (a) the “open-ended” questions in which respondents are asked to freely state their maximum amount of WTP, (b) the “Take-It-Or-Leave-It” (TIOLI) in which respondents agree or disagree with one proposed value, (c) the “bidding game” which starts with a single bid and increases or decreases in accordance with the respondent agreement till the maximum WTP is reached, (d) the “payment cards” in which respondents decide among cards presenting various values (If the cards are presented randomly the technique is called “shuffled payment cards”), and (e) “payment scales” in which respondents should select a range of values which consist of their desired maximum WTP [14]. These techniques can be used alone or combined with another elicitation method. The “open-ended” questions may lead to inaccurate answers because of “strategic bias” (this occurs when respondents behave strategically to influence the provision or funding of the asked item instead of expressing their true WTP amounts) [15]. Although the “bidding game” provides a “market-like” situation, it could suffer from “starting-point” bias (when the first presented amount affects the true maximum WTP of respondents) [15]. Need for a larger sample size and being susceptible to “starting-point” bias are disadvantages of TIOLI [15]. Although the “payment cards” method may be affected by “range bias” (effect of the range of amounts printed on the cards on the WTP amounts) it may not be suitable to be used in rural areas. Regarding the oral health field, some authors have recommended the “shuffled payment card method” as the most appropriate method to be applied for WTP studies [16].

Johannesson and Meltzer have recommended that investigation of societal WTP for health care should be a research priority [17]. WTP has also been suggested as the most appropriate method to measure patient preferences in dental care programs; both in publicly funded health care programs and in private insurance based plans [18‐20].

Since 1999 when the first WTP study in dentistry was published, the number of such studies dramatically increased globally over two decades. A similar trend in CV studies on general health has been documented over a broader time span [21]. Therefore, it seems to be the proper time to have a critical overview of the existing WTP studies in oral health care to search for any implication of the results for oral health policymaking, and in the future help undertake more robust practical research. Although a valuable critical review probed methodological aspects of WTP studies related to clinical services [22], in the current review both clinical and non-clinical studies are included. In addition, this study aims to conduct a scoping review on the existing oral health literature on WTP studies with a particular focus on policy implications and to examine areas of shortcomings in this field.

We chose a scoping review methodology on WTP studies in the area of dentistry to explore all aspects of oral health services to derive diverse methodologies and policy implications. Our methods align with the 5-step methodological framework recommended by Arksey and O’Malley, albeit the first two steps are amalgamated [23]. The last search was done on March 15, 2020.

Although representative samples of the general population are recommended in WTP studies, especially for consequential allocation decisions, most studies have used convenience samples. More opportunities may have arisen to undertake participant face-to-face active communication interviews accompanied by more detailed information regarding the desired service(s). Some studies elicited parental WTP for paediatric services [26, 34, 36, 37, 46, 52, 53]. It is worth mentioning that these case results cannot be indicative of patient (child) perceptions, feelings and utilization of services which is an important aspect of the WTP method. Moreover, lack of standard sample size calculation formulas in dentistry has remained a common pitfall in WTP studies. Pretesting was not done in a majority of the studies (60%) which is consistent with the findings of the previous review study [22].

Cultural as well as health system attributes (including the predominant payment method) are two critical points to be undertaken in designing WTP studies in any setting. For example, although the ‘shuffled payment card’ is more likely to elicit true WTP amounts [16], it has been shown that the ‘bidding-game’ may be more suitable in developing countries [72]. In most of the studies, due to the unclear or inadequate details about contextual attributes in WTP reports, we are not aware of such considerations given in the design of these studies. Another limitation of this review was not identifying, by critical appraisal, the high-quality studies. According to the included studies in this review, the “bidding game” is the predominant method for eliciting WTP in the oral health field. This finding is consistent with previous oral health review articles [22].

In the majority of previous review article studies (57%) which analyzed the association between income and WTP amounts, have reported a statistically significant association; indicating a direct correlation between ‘ability’ and ‘willingness to pay’ [22]. A possible advantage of WTP studies that researchers can investigate is whether or not people in higher income groups tend to prefer one option more frequently than do those in lower income groups [7]. If WTP results are to be used as inputs for equitable resource allocation decisions, they should be adjusted according to the income differences, otherwise WTP estimations may lead to skewed resource allocations favoring the higher income levels of society [73]. Reporting the WTP amounts as a percentage of income is recommended for this purpose [74].

In all 13 studies which reported a statistically significant association between WTP values and education level, higher education was associated with higher WTP amounts. In the health field, the positive effect of education level on WTP values was previously shown in a literature review on WTP studies regarding diagnostic technologies in healthcare [75]. In a majority of the included studies which reported a statistically significant association between gender and WTP amounts (5 of 7), females exhibited higher WTP amounts This finding is consistent with the results of the previous review study [22].

The net benefit of a service to a society is evaluated by comparing it to its costs [11]. Only 31 studies of the studies reviewed have made such comparisons between benefits (WTP amounts) and real market prices or service costs.

Some authors presumed WTP as an amount which will be precisely paid by participants in real-life circumstances [10, 33, 34, 37, 47, 50, 58], whereas many studies have shown a significant overstatement in WTP figures due to hypothetical bias (potential difference between individual decisions in the real market and a hypothetical situation) [76‐80]. Some studies have shown the potential of understatement in WTP amounts, especially in case of private goods and services [81].

A few authors have suggested that services be funded from public resources through either direct provision of that service by the public sector (government), or its inclusion within a benefit package through a national or social health insurance [32, 38, 39, 45, 45, 47]. Researchers should be aware of the part-whole bias (different values elicited for an item depending on whether it is valued solely, or as a part, in a more inclusive package) [82] in WTP studies which can magnify the utility of the service [83]. Even if the biases are fully prevented and we view WTP as a true proxy of population demand, still a suggestion needs further necessary criteria for it to be taken under consideration. For example, cost burden for a household, relative cost-effectiveness and socio-economic status of the potential consumers are just three of the important factors in deciding whether to finance a service under insurance coverage [84].

Some authors concluded from WTP results that health literacy should be improved within the population [24, 42, 59]. It seems that in these cases, the WTP method is reduced to an attitude assessment tool, whereas the WTP method could have more practical implications.

In this study Scopus and Web of science databases were not used and access to grey literature or personal communications was nonexistent. In future reviews of WTP studies in oral health, quality appraisal of conducted studies can unveil more details about the methodological aspects.

In order to apply WTP studies to allocation purposes (as the benefit wing of cost–benefit analysis), there should be a consistency in the design of WTP studies [11], otherwise, WTP results are reduced to comparisons between desired services, when at least two choices have been considered, and this seriously limits the generalizability of the studies. Though no systematic and defined quality appraisal was done in this review, it appears that the majority of WTP studies in the field of oral health suffer from several deficits in some parts of the design such as sample size, representing samples, dealing with potential biases and pretesting. According to the current body of literature, the generalizability of oral health WTP results remains questionable, particularly in higher level decision making. It is felt that researchers who are interested in conducting WTP surveys for dental services need to pay more attention not only to the methodological aspects of the WTP studies, but to health policy knowledge as well, in order to conduct well-built studies in connection with meaningful and pragmatic policy considerations. It can be deliberated that accompanying research-minded policymakers, policy-minded researchers are also needed.