Association of service use with subjective oral health indicators in a freedom of choice pilot

Abstract Objectives A freedom of choice pilot provided access to private oral health care services without queuing and with fixed public service‐fees for participants in Tampere region, Finland in 2018–2019. The aim of this study was to investigate how use of oral health care services differed by demographics, socioeconomic status, dental fear, and self‐reported oral health in this pilot. Material and methods SMS‐messages including a link to online questionnaire were sent to participants who had booked an appointment, and to those who had not booked an appointment despite registering to pilot. We categorized participants to (1) those who had booked their first appointment before receiving SMS (visitors), (2) those who booked an appointment after receiving the SMS‐message (late‐visitors), and (3) those who had not booked an appointment during pilot (nonvisitors). We used regression analysis to estimate the association of age, gender, dental fear, economic situation, Oral Health Impact Profile‐14‐severity (oral health‐related quality of life [OHRQoL]), self‐reported oral health and need for oral health care (exposures) with oral health care service use during the pilot (outcome). Results Out of 2300 participants, 636 (28%) responded. Late‐visitors were more likely older and reported more likely need for oral health care, poorer oral health and OHRQoL than visitors or nonvisitors. Nonvisitors were younger and had better OHRQoL than the others. The differences in the service use by gender, economic situation, and dental fear were small. Conclusions Service use during the pilot depended on the subjective oral health. Our findings highlight the potential of reminders in increasing care use among those with perceived need for services.


| INTRODUCTION
In Finland the whole population has been entitled to use subsidized oral health care services since December 2002. Since then, the entire Finnish population has been entitled to use public oral health care services (POHCS). Alternatively, they are entitled to receive reimbursements from the National Health Insurance for their private oral health care costs. From a user perspective, POHCS, which are maintained by municipalities (309 in 2021, in Finland), provide much cheaper services with nationally fixed out-of-pocket fees for adults. In addition to the basic fee, there are separate procedure fees for treatments and examinations depending on complexity class. In the private sector, client fees are not fixed and the reimbursement level from the National Health Insurance has long been decreasing.
Additionally, private services are found mainly in the large municipalities, whereas POHCS are available in every municipality. As adults have freedom to choose the service sector they use, unsurprisingly the demand for and use of subsidized oral health care services has been uneven between the sectors since the reform. In many municipalities, POHCS has been congested which has led to long treatment queues and prolonged treatment periods. Meanwhile, in the private sector there is considerable overcapacity in certain areas.
However, the majority of private dentists feel that the demand for their care is reasonable, thus guaranteeing significantly faster access to services in addition to better availability of specialist care (Raittio, 2016).
To meet the demand for services and to explore different oral health care provision strategies, the government introduced a freedom of choice oral health care pilot (FCOHC-pilot) during 2018-2019. It included access to private adult oral health services (excluding acute, implant, and fixed prosthodontic care) with the fixed fees of POHCS. The project was partly financed by the government, partly by the municipalities and out-of-pocket payments. The private caregivers in Tampere region showed interest in the project, since 13 service providers joined the pilot. Regular measurements of subjective oral health and patient experience were also part of the pilot.
The connection between the use of oral health care services and oral health is complex. On the one hand, the oral health care service use seems to affect clinical and subjective oral health status, and on the other hand, clinical and subjective oral health status seem to affect the service use (Torppa-Saarinen et al., 2018, 2019, 2021. It is also well known that out-of-pocket payments, dental fear and queues affect the oral health care service use (Cooray et al., 2020;Manning et al., 1985;Mueller & Monheit, 1988;Raittio et al., 2014;Suominen et al., 2017). For instance, considerable proportion of Finnish adults report that out-of-pocket payments and queues hinder their oral health care service use (Suominen et al., 2017).
The conducted FCOHC-pilot improved access to oral health care services without queuing for the pilot participants with the fixed fees of POHCS. Thus, with the available information on the pilot participants' subjective oral health, the pilot served as a fruitful base for investigating how subjective oral health measures and other determinants are connected with the oral health care service use.
Particularly, we investigated how those who did or did not book appointments during the pilot differed by their subjective oral health, oral health-related quality of life (OHRQoL), self-assessed treatment need, dental fear, and perceived financial situation. In the FCOHC, previsit SMS-messages, which included a link to online questionnaire were sent to pilot participants who booked their first visit before February 13, 2019. In addition, it was sent to those who had not booked an appointment until date February 13, 2019 (reminder SMS-message).

In
With these questionnaires we gathered information about subjective oral health, dental fear, and economic situation. Subjective oral health measures included self-assessed treatment need (yes/no), perceived oral health (good/rather good/moderate/rather poor/poor) and the 14-question version of Oral Health Impact Profile  for the OHRQoL. Economic situation was queried with a question: "How would you describe the current balance between income and expenditure in your household?" with reply alternatives: "We have more than enough money to cover our needs," "There is enough money to cover our needs," "We have, to some extent, to compromise when deciding what we do with our money," "We have to compromise considerably in our consumption but we can manage on our income," "We have to make major compromises in our consumption and, despite that, we do not manage on our own income" and "I cannot say/it is hard to estimate." Dental fear was elicited with question: "Do you think that visiting a dentist is…" using reply alternatives "not frightening at all," "somewhat frightening" and "very frightening." Information about age, gender, and appointments during pilot were collected from the oral health care data system.
To measure service use during the pilot, we categorized participants to (1) those who had booked their first appointment RAITTIO ET AL. | 135 before receiving SMS (visitors), (2) those who booked an appointment after receiving the SMS-message (late-visitors), and (3) those who had not booked an appointment during pilot (nonvisitors).

| Statistical analyses
We used regression analysis to estimate the association of age, gender, dental fear, economic situation, and perceived oral health variables (exposures) on oral health care service use during the pilot (outcome). We first drew our assumptions about the causal relationships of variables in a directed acyclic graph (DAG) using DAGitty (Supporting Information: Figure 1) (Tennant et al., 2021). Because perceived oral health and OHIP are very closely related (Kaprio et al., 2012), perceived oral health and OHIP shared same node in our DAG.
Based on our DAG, we identified minimum adjustment sets to estimate total effect (association) of each exposure (node) on the outcome with DAGitty (Table 1). As resulting with the use of this procedure, for each exposure-outcome pair, we identified all confounders which should be adjusted for and all mediators and colliders which should not be adjusted for (Tennant et al., 2021).
Statistical analyses were done in R environment (version 4.0.3) with packages nnet (Venables & Ripley, 2002), DAMISC (Armstrong, 2022), ggplot2 (Wickham, 2016) and tableone (Yoshida & Bartel, 2021). We performed multinomial regression analysis for each exposure-outcome pair with the adjustment variables. In addition to models with service use as outcome for age, gender, dental fear, economic situation, and perceived oral health care treatment need, we ran separate models for OHIP-14 severity, seven dimensions of OHIP and perceived oral health with same adjustment variables, and therefore there were 14 regression models in total. Then, we generated average effects for each exposure-outcome pair. We visualized the average effects of each exposure on outcome. In addition, we produced a table of basic descriptive statistics of our variables.

| RESULTS
During 2018, 5419 adults were registered to the pilot (Figure 1). Of those, 1700 who had booked their appointment received a questionnaire, and one-fourth (26%) completed it (visitors). Of 5419, 600 did not book their first appointment, but received the T A B L E 1 Minimum adjustment sets according to exposure variables based on a directed acyclic graph (DAG) The late-visitors were on average older (57.4 years) and the nonvisitors younger (51.1 years) than the visitors (54.0 years) ( Table 2).
The groups were similar according to gender, perceived economic situation, and dental fear. Higher proportion of the late-visitors reported the need for oral health care or had poor or rather poor perceived oral health than of the visitors or nonvisitors. The latevisitors also reported poorer OHRQoL than the other groups, and rather consistently also in all dimensions of OHRQoL.
T A B L E 2 Descriptive statistics of data in total and stratified by oral health seeking When the relationships between the exposures and outcome were adjusted for available confounders (Table 1) with multinomial regression analyses (Figures 2 and 3), the associations were mostly similar to bivariate differences shown in the  Figures 2 and 3).
F I G U R E 2 Average effect of each exposure on oral health care service use from multinominal regression analyses

| DISCUSSION
Our findings indicate that there were some differences in the oral health care service use by age, self-assessed treatment need, subjective oral health and OHRQoL among the FCOHC-pilot participants. Those who sought care after the reminder and questionnaire (late-visitors) were more likely older and reported more likely need for oral health care, poorer oral health, and OHRQoL than those who visited or did not visit at all during the pilot. Those who did not visit during the pilot were younger and had better OHRQoL than the others. The differences in the service use by gender, economic situation and dental fear were small.
Taking in to account the literature on the determinants of the oral health care service use in Finland (Nguyen et al., 2005;Suominen et al., 2017;Torppa-Saarinen et al., 2019) and elsewhere (Hajek et al., 2021), we detected only quite small differences in the service use by self-reported oral health, dental fear, economic situation, and demographics. Some reasons can be postulated. First, the sample in our study was very selected, because it included only those who were able to register themselves to the pilot and who also responded to the questionnaire. For instance, it may be that those, who have had economic problems and for that reason postponed their visits, did not register to the pilot because the fees were same as in the POHCS already (i.e., one could not receive any cheaper care through the pilot). Second, the highly selected sample may have affected the studied associations because the investigated oral health care service use and its determinants may have affected the registration and responding to the questionnaire and thus the sample selection process, which may in turn have induced biased estimates in our study due to so-called collider bias (Cole et al., 2010).
Interestingly, despite the highly selected group of individuals, the response rate was quite low. This is consistent with the findings related to quite low general response rates to health care questionnaires directed to patients (Booker et al., 2021). Solutions to increase the response rates for this kind of questionnaires should be looked for in future because they are considered important in evaluation of quality of (oral) health care (Riordain et al., 2021).
Our findings were in line with research (Guy et al., 2012;Kannisto et al., 2014;Schwebel & Larimer, 2018)  appointment. For instance, Scottish (Perry, 2011), Indian (Prasad & Anand, 2012) and Australian (Storrs et al., 2016) studies have indicated that appointment reminders can considerably decrease missed appointments in oral health care setting. However, some contradicting evidence also about the effects of reminders on the attendance to dental care have been found; the authors have speculated that it has been to high proportion of vulnerable populations and no financial consequence for failing to attend (Bellucci et al., 2017;Stormon et al., 2022). SMS-messages have also been successfully used to improve compliance to various medical interventions and self-care, including oral hygiene (Schwebel & Larimer, 2018). Our findings indicate that those who reacted after receiving the reminder (and questionnaire) were older and reported more likely need for oral health care, poorer oral health and OHRQoL.
The potential of reminders in increasing the services should not be overlooked as a way of increasing the oral health care service use among those for whom the services are directed to.
Our study is not without limitations. First, the sample size was quite small and based on low response rate, which may undermine validity, reliability, and generalizability of our findings. In addition, the generalizability of our findings is also weak due to the uniqueness of the pilot.
Second, we missed some important data. We had only data about care seeking, not about number or content of visits. We also missed important confounders, such as clinically assessed oral health and history of service use. However, they are to some extent represented by the available variables (such as self-assessed treatment need and subjective oral health measures). On the other hand, we gathered systematically collected data and implemented well-grounded statistical analyses to answer the question whether those who did or did not book appointments during the pilot differed from each other by their subjective oral health, OHRQoL, self-assessed treatment need, dental fear, and perceived financial situation.
In conclusion, our study showed that the differences in the use of oral health care services by gender, economic situation, and dental fear were small among the FCOHC-pilot participants. However, those who sought care after the reminder and questionnaire (late-visitors) were more likely older and reported more likely need for oral health care, poorer oral health, and OHRQoL than those who visited or not visited at all during the pilot. Those who did not visit at all during the pilot were younger and had better OHRQoL than those who visited with or without receiving a reminder. Despite uncertainty caused by the low response rates to the questionnaires, our findings highlight potential of reminders in increasing care seeking among those with perceived need for services.