Out-of-pocket payment and patients’ treatment choice for assisted reproductive technology by household income: a conjoint analysis using an online social research panel in Japan

As more people postpone parenthood [1, 2], the use of assisted reproductive technology (ART) is increasing worldwide [3]. Despite the growing need and effectiveness of this treatment, economic factors have contributed to huge disparities in access to ART within and between countries [4‐6]. To improve financial accessibility to ART treatments, governments of many countries provide various types of public funding, such as health insurance coverage, subsidies and tax refunds [7]. The proportions of reimbursement (i.e. full or partial) and eligibility criteria (e.g. clinical or demographical) vary substantially depending on regional and financial factors [7].

Japan provides universal health insurance coverage to the entire population [8]. In response to prolonged low fertility rates in Japan, ART treatments have been covered since April 2022 [9]. Prior to that date, only eligible couples received partial subsidies for ART treatment [10, 11]. Specifically, at the time of this survey in 2020, ART subsidies were available only to low- or middle-income patients. For high-income couples, the out-of-pocket payment for ART was approximately 400,000 − 500,000 JPY per cycle, or 3,300 − 4,200 euros using the 2020 exchange rate of 1 euro = 120 JPY [10, 11].

In Japan, eligibility rules for this and other subsidies (e.g. child allowance) often prioritize lower-income citizens [12]. Similar criteria were set up for ART subsidies in other East Asian countries, such as Korea [13] and Taiwan [14]. Yet, ART treatment is generally highly used among older and higher-income couples [15]. If these couples delay fertility treatment until they can better afford the associated cost, then the lack of subsidization for high-income couples could contribute to decreased rates of pregnancy [16]. In consideration of such potential adverse effects, the rationale for the eligibility criteria based on annual household income should be discussed.

Previous research has assessed the relationship between price and demand for ART [17], but few studies have assessed the quantitative association between demand and uptake of ART treatment and out-of-pocket payments by income level [4, 5]. The optimal amount of financial support by income class is unknown. We thus aimed to evaluate the probability of patients receiving ART treatment based on out-of-pocket payment and income class, using a conjoint analysis (CA).

In our CA study, we found that the out-of-pocket payment was the most influential determinant of ART treatment choice among participants, though all six attributes (e.g. pregnancy rates and kindness of staff) significantly influenced treatment preference. Higher-income patients were more likely to receive ART treatment even at a high cost, but their ineligibility for financial support due to their high income might discourage some from receiving treatment. This quantitative evaluation suggested that patients’ willingness to receive ART treatment could change substantially according to public funding of ART treatment. Many countries have sought to improve treatment accessibility in the context of limited financial resources, leading to enormous fluctuations in the number of treatments that patients opt to receive [4, 23, 24]. CA could be a feasible way to help policymakers identify the most appropriate programs for public funding under fiscal constraints.

Among the various factors affecting fertility care utilisation (e.g. sociocultural acceptability, availability), financial accessibility is a dominant one [6, 11]. Our findings corroborate that patients consider their out-of-pocket payment to be the most important attribute in their fertility decisions (Fig. 2). In addition, our prediction is consistent with previous findings on the relationship between cost and utilisation of ART treatment. For example, in an ecological study of 30 high-　and　upper-middle-income countries, Chambers et al. [5] find that a 1 percentage point decrease (based on annual disposable income for a single person with no dependents) in the cost of a treatment cycle predicts a 3.2% increase in utilisation. In our study, the probability of participants choosing ART treatment increased by 54% when the out-of-pocket payment decreased by 800,000 JPY (Fig. 3), corresponding with a 2.3% increase in demand per 1 percentage point decrease (based on disposable household income) in cost [25]. However, the slope differed depending on both income and out-of-pocket payment.

A study on ART treatment copayments in Germany demonstrated arc price elasticity of demand of -0.36 [17]: in other words, when the copayment for fertility treatment increased from free to 50% (1,500–2,000 euros), ART treatments dropped by 53% [23]. We observe a smaller elasticity (-0.13) but comparable inelasticity when the cost of ART treatment increases from 0 (free) to 200,000 JPY (1,670 euros), which is similar to results for other medical services, generally in the range of -0.1 to -0.3 [26]. Although the choices made in a CA would not completely match actual treatment choices, we confirmed that the CA could reasonably predict treatment choices based on arbitrary treatment costs.

As shown in Fig. 3, higher-income participants consistently opted for ART even at a higher cost: 27% of those in the High-income group were willing to pay up to 800,000 JPY (6,770 euros). For out-of-pocket payments exceeding 500,000 JPY (4,170 euros), we observed significant interactions between out-of-pocket payments and the Upper-middle- or High-household income groups (Table 4), suggesting an important effect of income: The negative impact of cost on ART demand was lower among higher-income populations, compared to lower-income populations.

Yet, the probability of receiving ART treatment among High-income group could fall below that of other income groups, depending on their subsidy eligibility. Among all five income groups, to attain a 47% probability of opting for ART treatment costing 400,000 JPY (as we observed for the High-income group), the ideal subsidy amounts are 200,000 JPY for Low-, 180,000 JPY for Lower-middle, 140,000 JPY for Middle- and 120,000 JPY for Upper-middle income groups. In 2020, the Japanese government provided a 150,000 JPY income-based subsidy for each of the second through sixth ART treatments, which was close to our suggested ideal amount. In contrast, the 300,000 JPY subsidy for the first application may induce a sense of “unfairness” among High-income patients.

In response to the country’s prolonged low fertility rate, the Japanese government removed the income eligibility criteria at the end of 2020 [10]. Since April 2022, up to six (three) cycles of ART treatments have been covered by health insurance for all legally or virtually married women aged < 40 years (40–42 years). Under the health insurance scheme, out-of-pocket payment for a fresh ART treatment cycle is estimated to be approximately 150,000 JPY (e.g., when six to nine eggs are retrieved, and two to five embryos are cultured and cryopreserved) [9]. This policy change helps to eliminate the perceived “unfairness” among High-income couples and to increase the total number of fertility treatments that women receive. Our model calculated that the probability of High-income couples opting for a fresh ART treatment cycle would increase nearly 1.3 times, from 47 to 61%, if the out-of-pocket payment decreased from 400,000 JPY(i.e., without a subsidy) to 150,000 JPY(i.e., covered by health insurance). This potentially large fiscal impact should be monitored to ensure that public funding for ART remains sustainable.

Interestingly, we found that Japanese patients preferred to seek ART treatment when the clinic staff was perceived as friendly (Fig. 2). Our findings agree with studies in China [19] and Europe [27] indicating that patients value physician’s attitude toward patients as much as they value the treatment’s success rate, whereas physicians underestimate the importance of patient-centred care. Although we could not assess a variety of aspects of patient-centredness, such as physician continuity or information provision of treatment [28], this is the first study to quantify the effects of patient-centred fertility care on treatment choice of patients in Japan. The fact that patients attached great importance to staff attitudes should promote clinicians’ understanding of the care they seek.

Another important finding of this study relates to women engaged in paid work for 40 h or more per week, who were more likely to choose treatment if they could visit outpatient clinics at night or on weekends. Aligned with rises in parental age and women’s labour force participation [2], 60% of women receiving fertility care have concerns about missing work [29], and about 17% of women in Japan resign from their jobs after starting fertility treatment [30]. We found that medical institutions offering consultation during nights and weekends were preferred by participants working outside the home. Since ART procedures require frequent and sometimes unpredictable visits based on the menstrual cycle [30], providing flexible clinic hours and a supportive work environment would help employees balance their fertility treatments with their work schedules [31].

This study has some limitations. First, attributes not included in our CA scenarios might be relevant in real-life settings. We selected six attributes based on surveys in Japan [21] and abroad [19, 20] and on our interviews of patients and clinicians. To maintain a reasonable number of scenarios presented to participants, we did not include all possible attributes (e.g. geographical access to the institution, amount of time required to see doctors). However, we ensured through the pilot survey that the questionnaire was understandable and that no vital information was missing from the scenarios. A future qualitative study including in-depth interviews in Japan might help construct a CA questionnaire. Second, decisions in a CA may not exactly match those during actual treatment. Empirical studies on patients’ actual behaviours before and after the subsidy policy revision and based on health insurance coverage could help confirm the validity of this study. Third, the use of social research panels could have caused selection bias associated with higher education [32, 33]. To recruit eligible samples efficiently, the market research company invited a large sample to the pre-screening, accepted responses in order of arrival and invited randomly selected participants to the survey. However, the sociodemographic distribution of our participants was similar to that of previous clinical settings in Japan [30], and patients with fertility problems tend to have higher education levels. Thus, our data should not be heavily influenced by potential selection bias. Finally, this study focused on ART treatment in Japan. However, CA is a generic approach that can be used widely. Future research should assess cultural relevance in perceptions of fertility information.

We found that out-of-pocket payments were the most influential determinant for patients in Japan seeking ART treatment. Higher-income patients opted for treatment even at a higher cost, but their income-based ineligibility for government financial support discouraged some from receiving treatment. Although the setting of the present study was in Japan, CA could be a feasible way to discuss appropriate programs for public funding in many countries.