Cost-to-baby as a percentage of median household income predicts population-level access to ART: a global health economics analysis
Giuseppe Silvestri, Steve Rooks, Alejandro Chavez-Badiola, Silvana Perez, Stephanie Kuku, Alan Murray, David Sable, G David AdamsonAbstract
STUDY QUESTION
Does cost-to-baby, expressed as a percentage of median household income, predict population-level utilisation of ART across countries?
SUMMARY ANSWER
ART affordability follows a power-law relationship with population-level utilisation (R² 0.77–0.84), with highest-utilisation countries consistently achieving gross cost-to-baby below 100% and net out-of-pocket cost-to-baby below 50% of median household income.
WHAT IS KNOWN ALREADY
Cost has been consistently identified as a dominant barrier to ART access worldwide, and earlier cross-national work has shown that per-cycle IVF cost as a proportion of disposable income is associated with national utilisation. However, no prior analysis has used the cumulative cost of achieving a live birth, normalised it to median household income, tested a non-linear functional form, or separated gross from net (out-of-pocket) cost as independent predictors.
STUDY DESIGN, SIZE, DURATION
Cross-national study using secondary aggregate data from ART registries and economic/demographic agencies for reporting years 2021–2023, covering 22 countries and regions that together represent more than 95% of estimated global ART cycle volume.
PARTICIPANTS/MATERIALS, SETTING, METHODS
ART cycle volumes and live births were extracted from ICMART, ESHRE EIM, CDC NASS, CARTR, HFEA, REDLARA, ANZARD, and national registries. Median after-tax disposable household income (HHI) data were sourced from the Organisation for Economic Cooperation and Development (OECD) and World Bank, while female population aged 15–49 years and total live births were obtained from the United Nations Population Division. Gross cost-to-baby was calculated as the product of mean per-cycle cost (including embryo transfer, preimplantation genetic testing weighted by national usage, and medications) and the age-weighted number of cycles required to achieve a live birth. Net out-of-pocket cost-to-baby was derived after applying published reimbursement schedules, subsidies, and tax benefits, with structured estimates where authoritative published data were unavailable. Both metrics were expressed as a percentage of median HHI. Utilisation was measured as the percentage of total national live births achieved via ART and as ART births per 1,000 women aged 15–49 years. Power-law (log-log) regression was fitted by ordinary least squares; a subgroup analysis was performed for women aged 38–49 years. A Monte Carlo probabilistic sensitivity analysis (10,000 iterations; Program Evaluation and Review Technique (PERT) distributions with uncertainty bands of ± 5%, ±7.5%, or ± 10% calibrated to data-source quality) generated 90% simulation envelopes around each baseline regression.
MAIN RESULTS AND THE ROLE OF CHANCE
Gross cost-to-baby varied more than 12-fold across countries, ranging from 66% of median HHI in Israel to 833% in Africa (excluding Egypt, Tunisia, and South Africa); net out-of-pocket cost-to-baby ranged from 13% to 825% of median HHI. Power-law regression demonstrated strong inverse relationships between affordability and utilisation across all four model specifications (gross cost vs. ART births per 1,000 women aged 15–49: R² = 0.84; net cost vs. the same measure: R² = 0.84; gross cost vs. percentage of births via ART: R² = 0.83; net cost vs. percentage of births via ART: R² = 0.77). The model-implied elasticity indicated that halving net out-of-pocket cost-to-baby is associated with a 2.67-fold increase in ART births per 1,000 women of reproductive age and halving gross cost-to-baby with a 5.12-fold increase. Countries with gross cost-to-baby below 100% and net cost-to-baby below 50% of median HHI consistently achieved the highest utilisation (e.g. Israel 8.7% of births via ART; South Korea 11.8%; Spain 11.7%; Japan 9.3%). The power-law relationship was preserved in women aged 38–49 years (R² 0.76–0.85). Monte Carlo simulation envelopes were narrow across all four models, indicating that the affordability–utilisation gradient is robust to plausible variation in country-level input estimates.
LIMITATIONS, REASONS FOR CAUTION
This is a cross-national, population-level (ecological) analysis. Since it uses aggregated, country-level data, the observed associations cannot support individual-level causal inference, may be subject to ecological confounding, and reverse causation through volume-driven economies of scale cannot be excluded. Registry data are heterogeneous across countries in reporting standards, cycle definitions, and completeness. Unmeasured confounders include cultural and religious attitudes, regulatory restrictions, clinic capacity constraints, and cross-border reproductive care. Net out-of-pocket cost estimates for a subset of countries were derived through structured triangulation that included AI-assisted synthesis and should be interpreted accordingly; the Monte Carlo analysis quantifies uncertainty in the affordability metrics but holds utilisation outcomes fixed. The macro-level framework cannot address within-country inequities in ART access.
WIDER IMPLICATIONS OF THE FINDINGS
Cost-to-baby relative to median household income offers a standardised, policy-relevant affordability metric that builds on previous per-cycle approaches. The power-law relationship implies that affordability improvements yield the greatest marginal returns in countries where costs are highest relative to income. The independent predictive value of gross and net cost-to-baby identifies two complementary policy levers: provider-side efficiency (workforce expansion, technology adoption, and care-delivery efficiencies to bring gross cost-to-baby below 100% of median HHI) and demand-side coverage (insurance mandates, subsidies, or tax incentives to bring net out-of-pocket cost below 50%). Published generational accounting analyses indicate that the lifetime fiscal return on public ART investment exceeds its cost by 2.7- to 16-fold across 11 of 12 countries studied, supporting the underlying economic case.
FUNDING
No specific funding was received; the study was supported by Conceivable Life Sciences.
DISCLOSURES
G.S. is an employee and stock option holder at Conceivable Life Sciences. S.R. is an employee of Ontario Teachers' Pension Plan, has received consultancy fees from Conceivable Life Sciences and Ziöra, and is a Director of Simbryo Technologies. A.C.-B. is an employee, shareholder, and company officer at Conceivable Life Sciences, and a shareholder in IVF 2.0. and Hope IVF Mexico. S.P. has received consultancy fees from Conceivable Life Sciences. S.K. is an employee, stock option holder, and company officer at Conceivable Life Sciences, has received consultancy fees from the WHO and Vitra Labs, and is Board Trustee of The King's Fund. A.M. is a shareholder and company officer at Conceivable Life Sciences and holds shares in TMRW Life Sciences. D.S. is a shareholder and board member at Oova, a board member of AutoIVF, an advisory board member of Evvy, Conceivable Life Sciences, and The US Fertility venture fund, and a general partner at the Special Situations Life Sciences Fund and Life Sciences Innovation Fund I–II. G.D.A is CEO of Advanced Reproductive Care Inc. (ARC Fertility), Chair of ICMART, and President of the World Endometriosis Research Foundation, and received consultancy fees from Pendant Health. Furthermore, G.S., A.C.-B., and A.M. declare inventorship on US patents 12,178,475 B1 (automated sperm preparation) and 12,226,125 B2 (automated vitrification). A.C.-B. and A.M. declare inventorship on US patents 12,349,940 B2 (automated oocyte denudation), 12,245,793 B2 (robotic handling systems for IVF laboratories), 12,180,441 B1 (laboratory workflow automation), 12,268,418 B2 (automated oocyte preparation), 12,310,625 B2 (automated ICSI), and 12,253,516 B2 (optical systems for IVF automation), as well as pending patent applications relating to automated dish preparation technology and automated vitrification systems.
TRIAL REGISTRATION NUMBER
N/A