Valuing Insurance Coverage for Infertility Treatment: A Double-Bounded Contingent Valuation Study in Kermanshah, Iran

Abstract

Background

Infertility imposes substantial medical, psychological, and financial burdens. Despite recent expansion of insurance coverage for infertility services in Iran, evidence on public willingness to contribute to pooled insurance arrangements remains limited. This study estimated willingness to pay (WTP) for infertility insurance among adults in Kermanshah, western Iran, and identified associated determinants.

Methods

A cross-sectional contingent valuation study was conducted among 630 adults recruited via convenience sampling in Kermanshah. WTP for a hypothetical package covering 70% of infertility treatment costs (up to two cycles/year) was elicited using a double-bounded dichotomous choice (DBDC) format (bids: 4,000–16,000 thousand IRR). Mean WTP was estimated via maximum-likelihood interval-data regression. Sensitivity analyses evaluated gender composition and response certainty.

Results

Of 626 valid respondents, 85.5% expressed positive WTP. Acceptance declined monotonically with increasing bids, supporting construct validity. Mean monthly WTP was 11,107 thousand IRR (95% CI: 10,212–12,003; base) and 11,266 thousand IRR (95% CI: 10,317–12,216; n=451, adjusted). Higher education (β=3,145.56, P=0.011) and socioeconomic status (β=5,254.04, P<0.001) predicted greater WTP; government-funding expectations (β=−5,046.92, P=0.004) and basic insurance (β=−3,697.27, P=0.050) predicted lower WTP. Certainty-calibrated and gender-balanced analyses confirmed robustness.

Conclusion

Adults in Kermanshah expressed substantial support for infertility insurance. Given the single-city convenience sample, generalization requires caution. Findings suggest that affordability, socioeconomic gradients, and public expectations regarding government responsibility should be considered in benefit-package and financing design.

Introduction

Infertility is a prevalent reproductive health condition with substantial medical, psychological, and social consequences.^1–4 Affecting approximately one in six people during their reproductive lifetime, it is recognized by the World Health Organization (WHO) as a disease of the reproductive system and an important component of sexual and reproductive healthcare.⁵ Despite this recognition, access to infertility services remains highly unequal, particularly in low- and middle-income countries (LMICs), where financial barriers and limited health-system capacity often restrict access to diagnosis and treatment.⁶

Advances in assisted reproductive technologies (ART), including in vitro fertilization (IVF), intracytoplasmic sperm injection (ICSI), and frozen embryo transfer (FET), have substantially improved the chances of conception for individuals experiencing infertility.⁷ However, these technologies are costly and often require repeated cycles to achieve pregnancy.⁸ In many settings, infertility services are only partially covered—or excluded entirely—from health insurance benefit packages, resulting in considerable out-of-pocket expenditure.^6,9 Such costs may deter treatment use, weaken financial protection, and widen inequalities in access to care.^10–12 As a result, decisions about insurance coverage for infertility services must consider not only clinical effectiveness, but also affordability, equity, and societal preferences regarding the use of collective healthcare resources.^6,13

These issues are particularly relevant in Iran. A recent national meta-analysis estimated lifetime and current infertility prevalence in Iran at 11.3% and 3.7%, respectively, with no substantial change in burden between 1990 and 2017, confirming that infertility remains a persistent public health concern in the country.¹⁴ Infertility carries important social and cultural implications in a Iranian context where childbearing remains highly valued.^15,16 At the same time, Iran is facing declining fertility and population ageing.^17–19 In response, the government enacted the Law on Supporting the Family and the Youthful Population in 2021, followed by policies expanding insurance coverage for selected infertility services.^6,20,21 Nevertheless, important policy questions remain unresolved, including the appropriate scope of coverage, cost-sharing arrangements, and the long-term sustainability of financing mechanisms.

It is important to distinguish conceptually between WTP for direct infertility treatment and WTP for insurance coverage. WTP for direct treatment reflects out-of-pocket payment for a specific cycle, typically elicited among patients with active treatment needs. In contrast, WTP for insurance coverage captures the recurring premium individuals would pay for pooled financial protection, independent of current fertility status. This distinction is fundamental for benefit-package design but remains underexplored in LMIC infertility literature.^22–25

Incorporating societal preferences into health financing decisions is increasingly important in economic evaluation and policy design.^26,27 Contingent valuation (CV), a widely used stated-preference method, estimates WTP for non-market health goods and services. In the context of health insurance, WTP captures the value individuals place on coverage and their willingness to contribute financially to pooled financing arrangements, thereby providing useful evidence for premium setting, subsidy design, and benefit-package development.^25,28–30

Although several studies have examined WTP for fertility care or infertility treatment in countries such as Greece, China, Canada, and the Netherlands, most have focused on infertile patients or treatment users rather than the general public.^22,31,32 Evidence on societal willingness to finance infertility services through insurance mechanisms remains limited, particularly in LMICs.

Our study extends the existing literature in three important ways. First, whereas most prior WTP studies in reproductive health have valued direct treatment costs among patients actively seeking fertility care,^22,31,33,34 we value a prospective insurance premium among the general population—a fundamentally different and more policy-relevant object for benefit-package design. Second, we contribute to the sparse LMIC evidence base on societal support to finance infertility services through pooled financing mechanisms. Third, to our knowledge, no previous study has estimated the Iranian general public’s WTP for infertility insurance coverage.

Therefore, this study estimates WTP for infertility treatment insurance coverage among adults in Kermanshah, western Iran, using a CVM with a double-bounded dichotomous choice DBDC format.³⁵ We hypothesized that (i) adults in Kermanshah would express positive WTP for infertility insurance coverage; (ii) WTP would be positively associated with education and socioeconomic status; and (iii) WTP would be negatively associated with stronger expectations of full government financing. By quantifying societal preferences for infertility coverage, the findings aim to inform benefit-package design, premium setting, cost-sharing policy, and broader resource-allocation decisions in reproductive healthcare.

Methods

Study Design and Setting

We conducted a cross-sectional contingent valuation study to estimate public WTP for insurance coverage of infertility services in Kermanshah, western Iran. Data were collected from September 23 to November 21, 2025. The contingent valuation component was embedded within a broader discrete choice experiment (DCE),³⁶ with WTP elicitation questions integrated into the same survey instrument administered to all participants.

Participants

Eligible participants were adults aged ≥18 years in Kermanshah, recruited through convenience sampling from 15 purposively selected sites across five geographical districts, approximately proportional to district population: District 1 (n=138, 22.0%), District 2 (n=125, 20.0%), District 3 (n=119, 19.0%), District 4 (n=127, 20.3%), District 5 (n=117, 18.7%). Sites included primary health centers, community spaces, and public locations; recruitment spanned weekdays, weekends, and morning/afternoon shifts to maximize diversity. The sole exclusion criterion was inability to comprehend the questionnaire; no exclusions were based on infertility history.

Sample Size

Sample size was determined based on the parent DCE using the regression-based method of Assele et al (2023),³⁷ yielding a minimum requirement of 402 respondents (α = 0.05, two-sided; 90% power); no separate calculation was performed for the CVM component. We recruited 630 participants to account for incomplete or invalid responses. Although not independently powered for CV estimation, the achieved sample size (626 with valid DBDC responses; 451 with complete covariate data) yielded relatively narrow 95% confidence intervals around the primary WTP estimates (10,211–12,003 thousand IRR, base model), supporting adequate precision at the population level; precision for subgroup-specific estimates may be more limited.

Survey Instrument and Data Collection

Participants completed self-administered paper-based questionnaires following standardized instructions. Trained research assistants remained at each site to provide standardized, non-directive clarification; approximately 5–10% of respondents raised questions, most commonly on the insurance package scope or certainty scale. In addition to the WTP elicitation questions, the instrument collected information on sociodemographic characteristics, health status, insurance coverage, infertility experience, and attitudes toward government financing of infertility treatment.

The validity of the questionnaire was assessed qualitatively by a panel of six experts (two health economists, one obstetrician-gynecologist, one biostatistician, one expert in health management, and one in health policy) as well as participants’ feedback. A pilot study with 30 participants (15 females, 15 males; age 22–58 years) was conducted to evaluate the appropriateness of bid levels, identify ambiguous items, and ensure clarity, comprehensibility, and cultural suitability across subgroups defined by age, education, and socioeconomic status; this confirmed that the four bid levels (4,000; 8,000; 12,000; 16,000 thousand IRR) spanned an appropriate range (>80% acceptance at the lowest; <20% at the highest).³⁶ Pilot participants were excluded from the final sample.

Contingent Valuation Scenario and WTP Elicitation

WTP was elicited using the CVM with a DBDC format. CVM is a widely used stated-preference technique for valuing non-market health goods by presenting respondents with a hypothetical scenario and eliciting their maximum WTP.^{29,30,35,38,39} The DBDC format enhances statistical efficiency by incorporating a follow-up bid contingent on the initial response, generating narrower confidence intervals compared with single-bounded or open-ended formats.^40,41

Respondents were presented with the following hypothetical scenario:

Suppose a new health insurance benefit, entitled “Infertility Insurance Coverage”, is introduced. Under this scheme, 70% of the costs of all infertility-related diagnostic and treatment services would be covered for up to two treatment cycles per year. Services would be available through contracted public and private providers. To finance this scheme, all insured individuals—not only those seeking fertility treatment—would pay an additional monthly premium.

Following the scenario, each participant received a randomly assigned initial bid and responded “yes” or “no” to the stated amount. A follow-up bid was then presented, adjusted according to the initial response: affirmative responses triggered a doubled follow-up bid, while negative responses triggered a bid reduced by 50%.⁴² An open-ended question also followed the DBDC asking maximum WTP of participants.^40,43

Bid Design and Bias Mitigation

Four initial bid amounts were used (4,000; 8,000; 12,000; and 16,000 thousand IRR), randomly assigned across respondents with approximately equal allocation. Bid levels were informed by pilot testing and calibrated to reflect plausible monthly premiums for complementary health insurance in Iran, ensuring coverage of a policy-relevant valuation range.^25,44 The use of multiple balanced starting bids was intended to reduce starting-point bias.

To address hypothetical bias, respondents were asked to rate their certainty regarding each WTP response on a 10-point Likert scale. Certainty calibration was subsequently applied by restricting the analytic sample to respondents who reported certainty scores at or above a pre-specified threshold, thereby yielding more conservative and policy-relevant estimates.^45–47

Econometric Framework

Under the DBDC design, each respondent’s two bounded responses define an interval within which their true WTP lies, based on the initial bid (t₁) and follow-up bid (t₂) amounts and the corresponding response pattern:^25,39,40

1. Yes–No (t₂ > t₁): t₁ ≤ WTP < t₂

2. Yes–Yes (t₂ ≥ t₁): t₂ ≤ WTP < ∞

3. No–Yes (t₂ < t₁): t₂ ≤ WTP < t₁

4. No–No (t₂ < t₁): 0 < WTP < t₂

WTP was modeled as a linear function of explanatory variables:

where is a vector of respondent characteristics, is a vector of parameters to be estimated, and is a normally distributed error term.

The parameters and were estimated by maximizing the following log-likelihood function:

where , , , and are indicator variables that take the value of one or zero depending on the relevant case for each individual, that is to say, a given individual contributes to the logarithm of the likelihood function in only one of its four parts, and denotes the standard normal cumulative distribution function.^25,48,49 Respondents reporting zero WTP during debriefing were retained in the No–No category and contributed through the term; no separate point mass at zero was specified. Although the normal latent-WTP specification theoretically permits negative values, their probability was negligible given the estimated mean WTP relative to its standard deviation.

Mean WTP was estimated via maximum-likelihood interval-data regression in both a base specification (intercept only) and an adjusted model with sociodemographic, insurance, and attitudinal covariates. The adjusted model used complete-case analysis. Missingness was primarily driven by economic status, education, and government financing attitudes. All analyses were performed in Stata 17.0 (StataCorp LLC, College Station, TX, USA).

Results

In total, 630 individuals participated in the study. The largest age group was 18–29 years (36.5%), and 65.4% of respondents were female. More than half of the participants were single (55.7%) and held an associate or bachelor’s degree (51.3%). Most respondents were covered by basic health insurance (91.9%), mainly through Social Security Insurance (54.9%), while 41.8% reported having supplementary insurance. The majority reported middle economic status (43.1%), and over half had no formal income (51.3%), a category that aggregated students, homemakers, and unemployed individuals. Additionally, 53.1% of participants reported having no children (Table 1).

Table 1.

Demographic and Socioeconomic Characteristics of the Participants (N = 630)*

Variable	Group	Number	Percentage (%)
Age	18–29 years	230	36.51
	30–39 years	181	28.73
	40–49 years	139	22.06
	50 years and older	80	12.70
Gender	Female	412	65.40
	Male	218	34.60
Marital Status	Single	344	55.66
	Married	274	44.34
Education Level	Illiterate/Primary/Lower secondary/High school	46	7.35
	High school/Diploma	113	18.05
	Associate degree and Bachelor’s	321	51.28
	Master’s degree and above	146	23.32
Basic Health Insurance	Uninsured	50	8.12
	Health Insurance Organization	172	27.92
	Social Security Insurance	338	54.87
	Armed Forces Insurance	46	7.47
	Other insurance schemes	10	1.62
Supplementary Insurance	No	349	58.17
	Yes	251	41.83
Economic Status	Low	182	29.26
	Middle	268	43.09
	High	172	27.65
Employment Status**	Government employee/Retired	196	31.31
	Private sector employee/Self-employed	109	17.41
	No formal income	321	51.28
Number of Children	No children	315	53.12
	One child	87	14.67
	Two children	131	22.09
	Three or more children	60	10.12

Notes: *The number of observations varies across variables because of missing data; therefore, the denominator for percentages may differ from the total sample size (N = 630). **The “no formal income” category includes students, homemakers, and unemployed individuals.

Most participants reported good self-rated health (73.0%) and no history of chronic disease (85.8%). Over half had no personal or indirect experience with infertility; however, a substantial proportion reported infertility experience among relatives or acquaintances. Awareness of infertility treatment methods was predominantly low to moderate (86.8%). Overall, strong support was observed for government financing of infertility treatments through health insurance, with 89.0% of respondents agreeing or strongly agreeing (Table 2).

Table 2.

Health Status, Infertility Experience, and Attitudes Toward Insurance Coverage (N = 630*)

Variable	N	Percentage (%)
Self-rated Health Status
Poor	48	7.63
Fair	122	19.40
Good	459	72.97
History of Chronic Disease
Yes	89	14.19
No	538	85.81
History of Infertility Problems
No	355	56.80
Yes, respondent	22	3.52
Yes, spouse	10	1.60
Yes, relatives	125	20.00
Yes, acquaintances	113	18.08
Physician Consultation for Infertility
No	381	60.86
Yes, respondent	18	2.88
Yes, spouse	9	1.44
Yes, relatives	116	18.53
Yes, acquaintances	102	16.29
Experience of Infertility Treatment
No	394	63.14
Yes, respondent	18	2.88
Yes, spouse	9	1.44
Yes, relatives	103	16.51
Yes, acquaintances	100	16.03
Awareness of Infertility Treatment Methods
Very low	151	24.04
Low	162	25.80
Moderate	232	36.94
High	67	10.67
Very high	16	2.55
Attitude Toward Government Financing of Infertility Treatment Through Health Insurance
Strongly agree	369	58.66
Agree	191	30.37
Neutral	55	8.74
Disagree	7	1.11
Strongly disagree	7	1.11

Notes: *The number of observations varies across variables because of missing data; therefore, the denominator for percentages may differ from the total sample size (N = 630).

Of the 630 enrolled participants, 626 provided valid responses to the DBDC bidding questions and were included in the primary analysis. The open-ended follow-up question on maximum WTP was completed by 540 respondents, of whom 60 stated a zero maximum WTP and 480 stated specific positive amounts.

Of 626 respondents with valid DBDC responses, 535 (85.5%) reported positive WTP by accepting at least one of the two offered bids, 66 (10.5%) rejected both bids (including some who explicitly indicated zero WTP in the open-ended follow-up), and 25 (4.0%) provided missing or invalid WTP responses (eg, incomplete bid-response sequences) and were excluded from the WTP analysis.

Acceptance rates decreased systematically with increasing bid amounts. The proportion of respondents accepting both the initial and follow-up bids (Yes/Yes) declined from 36.9% at the 4,000 thousand IRR bid to 8.0% at the 16,000 thousand IRR bid, while the proportion rejecting both bids (No/No) increased from 5.4% to 37.2%. These patterns demonstrate price sensitivity consistent with economic theory, providing evidence of the construct validity of the willingness-to-pay responses (Table 3).

Table 3.

Distribution of Responses to DBDC Questions by Initial Bid Amount (N = 626**)

Initial Bid (IRR Thousands*)	F/U8*** if Yes if Answer to Initial Bid = “yes”	F/U if No if Answer to Initial Bid = “No”	Yes/Yes (%)	Yes/No (%)	No/Yes (%)	No/No (%)
4000	8000	2000	36.94	35.14	22.52	5.41
8000	16,000	4000	17.71	33.33	32.29	16.67
12,000	24,000	6000	22.32	26.79	32.14	16.96
16,000	32,000	8000	7.96	30.97	23.89	37.17

Notes: *1 US$ ≈ 1,120,000 IRR at the time of the study. **The number of observations varies across variables because of missing data; therefore, the denominator for percentages may differ from the total sample size. ***Follow-up bid.

Higher education (β=3,145.56, p=0.011) and higher socioeconomic status (β=5,254.04, p<0.001) were significantly associated with greater WTP for infertility insurance coverage. Conversely, basic insurance coverage was associated with lower WTP at the margin of statistical significance (β=−3,697.27, p=0.050), and stronger endorsement of government-funded coverage was significantly associated with lower WTP (β=−5,046.92, p=0.004). Other demographic and health-status variables—including age, sex, marital status, having children, infertility history, supplementary insurance, and self-rated health—were not statistically significant predictors (all p>0.05). These results indicate that WTP is primarily driven by economic capacity, existing insurance arrangements, and attitudes toward collective financing rather than individual health or demographic characteristics (Table 4).

Table 4.

Influence of Explanatory Variables on Mean WTP for Infertility Insurance Coverage Based on the DBDC Interval-Data Regression Model

Variable	Coefficient	Standard Error	z-Value	P-Value	95% Confidence Interval
					Lower	Upper
Gender (Reference: Female)
Male	−324.05	1083.72	−0.30	0.765	−2448.10	1800.01
Age Group (Reference: 18–29 Years)
30–39 Years	830.16	1356.02	0.61	0.540	−1827.59	3487.92
40–49 Years	712.11	1711.04	0.42	0.677	−2641.46	4065.68
50 Years and Above	2382.11	2026.88	1.18	0.240	−1590.49	6354.72
Marital Status (Reference: Single)
Married	−544.38	1483.12	−0.37	0.714	−3451.25	2362.48
Education Level (Reference: Non-University)
University Education	3145.56	1230.73	2.56	0.011	733.36	5557.75
SES (Reference: Poor)
Middle	1404.63	1194.12	1.18	0.240	−935.81	3745.07
High	5254.04	1391.27	3.78	<0.001	2527.21	7980.88
Basic Insurance (Reference: No Basic Insurance)
Has Basic Insurance	−3697.27	1942.93	−1.90	0.050	−7505.34	110.80
Supplementary Insurance (Reference: No Supplementary Insurance)
Has Supplementary Insurance	−299.87	1074.17	−0.28	0.780	−2405.21	1805.47
Attitude Towards Government Support (Reference: Opposed)
In Favor	−5046.92	1752.25	−2.88	0.004	−8481.26	−1612.57
History of Facing Infertility (Reference: No History)
Has a history	394.32	1005.00	0.39	0.700	−1575.44	2364.07
Health Status (Reference: Poor to Moderate)
Good	378.52	1040.70	0.36	0.720	−1661.21	2418.24
Having Children (Reference: No Children)
Has Children	−1163.59	1599.95	−0.73	0.470	−4299.43	1972.26
_Cons	11961.35	4209.53	2.84	0.005	3710.82	20,211.88

Notes: Number of observations = 451 (extended model); base model N = 534. The reduced sample size in the extended model is due to listwise deletion of observations with missing covariate data, primarily on economic status, education, and attitude toward government financing. Complete-case analysis was used. Wald χ²(14) = 44.12, Log-likelihood = −650.22, Prob > χ² = 0.0001. All values are in thousands of IRR. Bold values indicate statistical significance at P ≤ 0.05.

Table 5 reports the estimated mean WTP from both the base specification (n=534) and the model incorporating explanatory variables (n=451). The results are highly consistent across specifications, with only a marginal increase in WTP when covariates are included (11,107 vs 11,266 thousand IRR), indicating model stability. In both cases, WTP estimates are statistically significant at the 1% level with relatively narrow confidence intervals, suggesting precise estimation. Overall, the minimal difference between the two models implies that observable sociodemographic characteristics do not materially affect aggregate WTP estimates, thereby reinforcing the robustness of the findings for policy interpretation (Table 5).

Table 5.

Mean WTP in the Base Model, the Model with Explanatory Variables, and Sensitivity Analyses

Model	N	Estimated WTP (Thousand IRR)	SE	z-Value	P-Value	95% CI (Thousand IRR)
Base Model (full sample)	534	11,107.5	456.9	24.31	<0.001	10,211.5–12,002.5
Adjusted Model (full sample)	451	11,266.5	484.3	23.26	<0.001	10,317.2–12,215.7
Certainty-calibrated base (score ≥7)	392	11,443.6	542.6	21.09	<0.001	10,380.1–12,507.1
Certainty-calibrated adjusted (score ≥7)	336	11,470.6	560.1	20.48	<0.001	10,372.8–12,568.3
Gender-balanced base (50/50)	375	10,937.3	548.4	19.94	<0.001	9862.4–12,012.1
Gender-balanced adjusted (50/50)	313	11,082.0	582.9	19.01	<0.001	9939.6–12,224.4

Of 566 respondents with valid certainty scores (89.8%), 446 (78.8%) reported certainty ≥7. In this calibrated subsample, mean WTP was 11,444 thousand IRR (95% CI: 10,380–12,507; n=392) in the base model and 11,471 thousand IRR (95% CI: 10,373–12,568; n=336) in the adjusted model, differing from full-sample estimates by 1.8–3.0%. All major coefficient directions were preserved, though some lost significance due to reduced power (education p=0.056; government attitudes p=0.181). These small shifts indicate negligible hypothetical bias, and full-sample estimates were retained as primary results.

As a further robustness check for the female-dominated sample (65.4% female), a gender-balanced subsample (218 females, 218 males; n=436) was constructed via random downsampling of female respondents. Mean WTP was 10,937 thousand IRR (95% CI: 9,862–12,012; n=375) in the base model and 11,082 thousand IRR (95% CI: 9,940–12,224; n=313) in the adjusted model—1.5–1.6% below full-sample estimates. All major covariates retained direction and significance (education β=3,434.73, p=0.017; high SES β=5,413.30, p=0.001; government attitudes β=−4,493.84, p=0.032), confirming that female overrepresentation did not materially bias results.

Discussion

This study estimated WTP for infertility insurance coverage among 626 adults in Kermanshah, western Iran, using a DBDC contingent valuation design. Unlike most previous WTP studies that valued infertility treatments directly,^{22,31,32,34,50} we assessed valuation of an insurance premium for a package covering 70% of infertility-related diagnostic and treatment costs for up to two cycles per year—a design more directly relevant for benefit-package policy. To the best of our knowledge, no studies are directly comparable to our study. However, our results align with the limited comparable findings from other published studies.^22,31,32

Three findings are particularly notable First, 85.5% of respondents expressed positive WTP, indicating broad public support. Second, mean monthly WTP was stable across specifications: 11,107 thousand IRR in the base model and 11,266 thousand IRR in the adjusted model. At the time of data collection, this corresponded to approximately US$10 per month at the parallel market exchange rate of 1,120,000 IRR/USD and was equivalent to approximately 11% of the monthly minimum wage of 104 million IRR as established by the Ministry of Cooperatives, Labor, and Social Welfare (MCLSW) for the year of the study. Third, WTP was associated primarily with university education, socioeconomic status, basic insurance coverage, and attitudes toward government financing, whereas most demographic and health characteristics were not significant predictors.

These findings suggest that infertility insurance is perceived as a legitimate component of health protection rather than a discretionary service. Comparisons with patient-level WTP estimates should be interpreted with caution, as the valuation objects differ fundamentally: our study values a prospective insurance premium among the general population, whereas patient-level studies value immediate treatment costs among individuals with current fertility needs. These differences in valuation object, respondent population, and temporal framing limit direct numerical comparability.

Nonetheless, our interpretation aligns with the broader international evidence. In Canada, Poder et al found that women’s WTP for ovulation induction averaged CAD$4,805—substantially exceeding treatment cost—and concluded that such care was highly socially desirable.³² In Spain, Palumbo et al reported that women were willing to pay a median of €800 per cycle for ovarian stimulation, with 35.5% willing to pay €101–300 more for a 1–2% effectiveness improvement.⁵⁰ These studies confirm that fertility services carry value beyond immediate clinical outcomes due to their psychological and social meaning.

However, WTP is also strongly shaped by affordability. In Greece, Ntais et al found that women were willing to pay only approximately €2,000 per IVF cycle—roughly one-third of the average cost of €6,000.³¹ In China, Li et al reported mean WTP for IVF of RMB 30,163 (US$4,259), equivalent to 0.34 times GDP per capita.²² In Brazil, Lima et al documented mean WTP of R$18,720 (US$4,800), with positive correlations to household income and education.³⁴ Earlier Iranian evidence from Darvishi et al showed mean WTP for IUI and IVF of 15,941,061 and 28,870,833 IRR per cycle respectively—both below actual treatment costs, with financial constraints identified as the decisive factor.³³

Our findings fit this pattern. The estimated WTP reflects valuation of an insurance premium, not direct treatment purchase. This distinction is critical: bounded WTP should not be interpreted as low perceived value of infertility care, but rather as a preference for collective financing of high-cost services through risk pooling and public subsidy. At 11% of minimum wage, the monthly premium represents a substantial relative burden for Iranian households, particularly lower-income groups.

The bid-response pattern supports construct validity. Acceptance of the initial bid declined from 72.1% at the 4,000 thousand IRR bid (Yes/Yes 36.9% + Yes/No 35.1%) to 38.9% at the 16,000 thousand IRR bid. Acceptance of both the initial and follow-up bids (Yes/Yes) declined from 36.9% to 8.0%, while double rejection (No/No) increased from 5.4% to 37.2%. This price sensitivity demonstrates economically coherent preference formation.³⁶ The minimal difference between base and adjusted models (11,107 vs 11,266 thousand IRR) indicates robustness to specification. Certainty calibration yielded only marginally higher estimates (11,428 and 11,539 thousand IRR), suggesting negligible hypothetical bias.

Respondents with university education reported 3,146 thousand IRR higher WTP (p=0.011), while high socioeconomic status increased WTP by 5,254 thousand IRR (p<0.001) compared to low status. This nearly 47% difference has important equity implications. To illustrate these equity implications—as a hypothetical example rather than an empirical estimate from our dataset—a monthly premium of 11,000 thousand IRR would represent approximately 11% of monthly income for a household earning the minimum wage (~104 million IRR), compared with approximately 5.5% for a hypothetical household earning 200 million IRR per month. This illustrative comparison underscores the regressivity of flat-rate premiums and the potential rationale for income-adjusted contributions, consistent with equity principles underlying social health insurance.^51,52

A particularly policy-relevant finding was the negative association between WTP and both basic insurance coverage (β=−3,697 thousand IRR, p=0.05) and favorable attitudes toward government financing (β=−5,047 thousand IRR, p=0.004). This should not be interpreted as weaker support for coverage. Rather, these respondents likely expect infertility services to be financed through existing pooled arrangements or public subsidy rather than additional household payments. Lower stated WTP may therefore reflect stronger expectations of solidarity-based financing. This interpretation is supported by Lima et al, who found that previous use of Brazil’s public health system was associated with lower WTP despite clear perceived value of treatment.³⁴ In settings where collective health financing is normatively established, WTP captures preferences over both service value and financing source. However, we acknowledge that alternative explanations are plausible and cannot be distinguished with the present data. Respondents strongly favoring government financing may perceive an additional premium as duplicative of tax-funded provisions, or may register a principled refusal of personal contribution as a form of protest response. Qualitative follow-up research is needed to disentangle these mechanisms.

Most other factors—age, sex, marital status, having children, infertility history, supplementary insurance, and health status—showed no significant association with WTP. In contrast, WTP was significantly associated with university education, high socioeconomic status, basic health insurance coverage), and favorable attitudes toward government financing. This suggests that support for infertility insurance extends beyond those directly affected. Many previous studies sampled infertile patients or treatment-seeking couples, whose valuations may be shaped by immediate need.^22,31,34 Our general population perspective is more appropriate for insurance design decisions in publicly financed systems.

Rather than prescribing specific financing parameters, our findings suggest several considerations for policy design: (1) the high level of public support (85.5% positive WTP) provides a social mandate for including infertility services within the benefit package; (2) the estimated mean WTP (~11,000 thousand IRR per month) provides an empirical reference point for premium-setting discussions, though actual premium levels must be informed by actuarial analysis, budget-impact assessment, and fiscal sustainability modeling; (3) the significant association between WTP and socioeconomic status suggests that income-sensitive contribution mechanisms may be warranted to protect lower-income households; and (4) the negative association between WTP and expectations of government financing indicates that public preferences regarding the source of funding should be explicitly considered. International evidence from countries with established infertility coverage mandates suggests that mixed public–private financing arrangements can improve access and reduce socioeconomic disparities in ART utilization.^6,53–57The specific parameters of any Iranian scheme, however, require dedicated actuarial and budget-impact studies beyond the scope of the present WTP analysis.

This study has several strengths, including an efficient DBDC elicitation format, pre-tested bid vectors, certainty calibration, and multiple sensitivity analyses, providing—to our knowledge—the first assessment of public WTP for pooled infertility insurance coverage in Iran.

Several limitations warrant acknowledgment. First, single-city convenience sampling introduces selection bias and limits generalizability; the sample’s younger, more educated, and more insured profile likely biased WTP upward. Second, as with all stated-preference methods, several interrelated biases may have affected responses: hypothetical bias (mitigated via certainty calibration, which suggested negligible distortion), residual anchoring despite balanced bid rotation, social desirability given the cultural salience of infertility, and potential context effects from embedding WTP questions within a broader discrete choice experiment.^58,59 Third, predominantly low-to-moderate treatment awareness (86.8%) may have introduced information bias, while aggregation within the “no formal income” (51.3%) and “university-educated” categories—without cross-tabulation against employment status—limits subgroup interpretation and should be refined in future surveys. Fourth, complete-case analysis led to the exclusion of participants with missing data on some variables (n = 451, reduced from 626); comparison of included and excluded respondents revealed no significant differences in key characteristics, and sensitivity analyses showed only marginal differences in mean WTP estimates, suggesting limited impact on our findings. Fifth, the sample size was determined for the parent discrete choice experiment and was not independently powered for the CV component; although the narrow confidence intervals around our primary WTP estimates support adequate precision at the population level, precision for subgroup-specific estimates may be more limited. Sixth, the normal distribution assumed for latent WTP theoretically permits negative values and does not accommodate point-mass zero responses; log-normal, Weibull, or spike models merit future exploration. Finally, insurance-premium valuations are not directly comparable with per-cycle treatment WTP estimates.

Despite these limitations, this study offers important evidence on public valuation of infertility insurance in a middle-income setting, demonstrating that support for coverage is substantial but shaped jointly by perceived benefit, economic capacity, and expectations about collective financing—implying that policy design must balance fiscal sustainability with equity and social acceptability.

Conclusion

This study estimated WTP for infertility insurance coverage among adults in Kermanshah, Iran, and tested three hypotheses: positive WTP, positive associations with education and socioeconomic status, and negative association with government-financing expectations. All three were corroborated: 85.5% expressed positive WTP (mean ~11,107 thousand IRR/month; ~US$10, ~11% of the national minimum wage), with the predicted directional associations confirmed in multivariable analysis and robustness supported by sensitivity analyses. These findings indicate considerable public support for infertility coverage and suggest that affordability and financing-source preferences should inform benefit-package design. Given the single-city convenience sample, nationally representative studies incorporating actuarial and budget-impact modeling are needed to translate these preference data into specific, fiscally sustainable financing mechanisms.

Funding Statement

The authors received no financial support for this study.

Data Sharing Statement

The datasets used and/or analyzed during the current study are not publicly available due to privacy and confidentiality considerations but are available from the corresponding author upon reasonable request.

Ethics Approval and Consent to Participate

This study received approval from the Ethics Committee of Kermanshah University of Medical Sciences (IR.KUMS.REC.1403.074). All procedures were performed in accordance with the Declaration of Helsinki. Informed consent was obtained from all participants.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors declare no conflict of interest.