Background
The United States has become a major destination for cross-border reproductive care, yet limited national data exist on international assisted reproductive technology users and their outcomes.
Objective
To characterize cross-border reproductive care in the United States and compare assisted reproductive technology utilization patterns and outcomes between US and non-US residents.
Study Design
Retrospective cohort study using data from the Society for Assisted Reproductive Technology Clinic Outcome Reporting System characterizing temporal trends, treatment characteristics, and live birth outcomes among 2,275,167 assisted reproductive technology cycles from 2014 to 2022, including 59,246 (2.6%) cycles from non-US residents representing 182 countries. Outcomes were stratified by oocyte source and plurality (for preterm birth) and analyzed using multivariable log-binomial regression with generalized estimating equations.
Results
The number and percentage of assisted reproductive technology cycles among non-US residents rose steadily from 2014 to 2019 (2.8% to 3.4%) before declining in 2020. China accounted for the highest volume of international patients (19,718 cycles), nearly 3-fold higher than the next closest country, Canada (6990 cycles). Compared to US residents, non-US patients were more likely to use intracytoplasmic sperm injection (93% vs 84%), preimplantation genetic testing (68% vs 43%), donor oocytes (52% vs 9%), and gestational carriers (45% vs 2%). Live birth rate was slightly higher among non-US residents for autologous (48.3% vs 44.2%; adjusted risk ratio, 1.02; 95% confidence interval, 1.00–1.04) and donor oocyte cycles (55.5% vs 48.4%; adjusted risk ratio, 1.10; 95% confidence interval, 1.09–1.13).
Conclusion
Cross-border reproductive care accounts for a growing proportion of assisted reproductive technology cycles in the United States, with non-US residents more frequently using advanced and third-party reproductive technologies. Despite differences in assisted reproductive technology treatments, non-US residents experienced slightly better outcomes. These findings highlight persistent global disparities in assisted reproductive technology access and underscore the United States’s role as a key destination for complex fertility care.
Introduction
Assisted reproductive technology (ART), including in vitro fertilization (IVF), accounted for approximately 2.6% of all infants born in the United States in 2023, reflecting its growing role in family building. Globally, the use of ART has also expanded, contributing to the rise of cross-border reproductive care (CBRC), a practice in which individuals or couples travel internationally to access autologous and/or donor fertility services.
AJOG AT A glance
Why was this study conducted?
To characterize cross-border reproductive care in the United States and compare assisted reproductive technology (ART) utilization patterns and outcomes between US and non-US residents.
Key findings
Non-US ART patients more commonly used donor oocytes, gestational carriers, and genetic testing than US patients. Non-US residents had slightly higher live birth rates than US residents.
What does this add to what is known?
Cross-border reproductive care accounts for a meaningful and growing proportion of ART cycles in the United States, with non-US residents more frequently using advanced and third-party reproductive technologies. These findings highlight persistent global disparities in ART access and underscore the United States’s role as a key destination for complex fertility care.
CBRC is driven by multiple factors, including access to care unavailable in one’s home country and circumvention of restrictive laws, including limitations based on age, marital status, sexual orientation, or gender identity. In addition, some specialized services, such as third-party reproduction and surrogacy, are unavailable in certain countries. , Financial factors also influence destination choice, as affordability varies across countries. Beyond legal limitations, individuals seek care internationally due to long wait times and safety concerns. As ART becomes more globalized, countries with advanced technologies, favorable legal frameworks, and large donor or surrogacy networks, such as the United States, have emerged as primary CBRC destinations despite higher costs.
Despite the growing relevance of CBRC, especially in the United States, comprehensive data on this population remain limited. Previous reports have largely relied on European registry data or surveys of patients and clinics. ,, To date, only 2 studies have assessed CBRC using US national ART data. The primary study evaluating US CBRC data was a retrospective study of all ART cycles (n=1,271,775 cycles) reported to the Centers for Disease Control’s National ART Surveillance System between 2006 and 2013 that focused on fresh embryo transfer cycles and found that non-US resident use of ART in the United States increased steadily during this period and that nonresidents made greater use of specialized treatment techniques such as preimplantation genetic diagnosis or screening. A second study using all ART cycles in the United States reported to the Society for Assisted Reproductive Technology Clinic Outcome Reporting System (SART-CORS) that included an embryo transfer to a gestational carrier (GC) from 2014 to 2020 (n=40,177 embryo transfers), focusing on international use of US gestational surrogacy, and, while they too demonstrated annual growth in the use of this service, the analysis was limited to GC cycles. Both studies predated considerable clinical and global shifts, including the COVID-19 pandemic, changes in Chinese reproductive policy, and legal reforms regarding ART access in Europe and elsewhere.
To address this gap, this report examines the frequency and trends in CBRC use in the United States from 2014 through 2022, the geographic distribution of non-US patients, variation in ART treatment types by residency status, and the outcomes of these treatment cycles to better understand the CBRC in a rapidly evolving global fertility landscape.
Materials and methods
Study population
Data for this study were obtained from the SART-CORS. Data are voluntarily reported by SART member clinics, verified by SART, and reported to the Centers for Disease Control and Prevention in compliance with the Fertility Clinic Success Rate and Certification Act of 1992 (Public Law 102-493). More than 90% of all ART cycles in the United States are performed at SART-member clinics. SART conducts annual on-site validation visits of a subset of clinics using a blinded selection process to verify cycle designation, outcomes, and reporting accuracy. Clinics with systematic reporting errors undergo data correction. Certain data fields, including treatment-specific and demographic variables, are not subject to validation. , This study included all SART-reported ART cycles between January 1, 2014 and December 31, 2022. Gamete intrafallopian transfers, zygote intrafallopian transfers, and donor embryo cycles were excluded when data were requested from SART. Initially, only cycles where the patient was banking oocytes for donation were excluded (n=11,558) prior to analysis, leaving a sample size of 2,278,190 cycles. The study was approved for exemption by the Emory University Institutional Review Board as the study only involved the secondary analysis of deidentified data.
Residency status
SART-CORS includes a variable for US residency as well as a separate variable for patient country. For the 3266 cycles (0.1%) in which US residency status was marked as “Refused” (eg, missing), we used the country of residence variable to classify residency status, when possible. After this process, 3023 cycles (0.1%) remained unclassified and were excluded from residency-based analyses.
Assisted reproductive technology procedures
SART-CORS includes data on specific ART procedures, including use of donor oocytes, use of a GC, preimplantation genetic testing (PGT), and intracytoplasmic sperm injection (ICSI).
The primary treatment, pregnancy, and birth outcomes of interest included clinical pregnancy, live birth, miscarriage, multiple birth, and preterm delivery. For both clinical pregnancy and live birth, the denominator was all cycles with at least one embryo transferred. For miscarriage, we only included cycles with confirmed clinical pregnancy. Multiple birth (singleton vs twins, triplets, or higher-order multiple) was only assessed among live births. Gestational age at birth was calculated as delivery date minus transfer date plus day of embryo transfer plus 14, consistent with guidelines from the American College of Obstetricians and Gynecologists. Any live birth with a gestational age prior to 37 weeks was considered preterm. All preterm birth analyses were stratified by plurality of birth.
Statistical analyses
We quantified the utilization of ART by non-US residents by comparing the absolute number and percentage of ART cycles attributable to non-US residents each year over the study period. We also quantified autologous oocyte retrievals, oocyte banking, donor oocyte cycles, and embryo transfers among non-US residents. For each patient’s first ART cycle in our database, we compared demographic, reproductive, and clinical characteristics at the time of their first cycle between US residents and nonresidents.
To assess differences in ART treatment utilization, we compared the use of donor sperm, PGT, and ICSI in all fertilization cycles by US residency status as well as separately among autologous and donor oocyte fertilization cycles. Among embryo transfer cycles, we compared the use of donor oocytes, GCs, and frozen embryo transfers; the number and stage of embryos transferred; and the treatment, pregnancy, and birth outcomes among US and non-US residents. We used ArcGIS Online, a geographic information system software platform, to map the geographic location of non-US residents according to the volume of ART use in the United States. For the 15 countries with the highest volume of cycles, we assessed differences in the use of ART techniques by calculating the percentage of fertilization cycles using donor sperm, ICSI, and PGT and the percentage of embryo transfer cycles using frozen embryo transfer, donor oocytes, and GCs by non-US residents from each country. For the top 5 countries, we examined the number of cycles undertaken each year to determine if there were any notable time trends.
To determine whether ART outcomes differed by residency status, we compared the average number of embryos transferred, the percentage of embryo transfer resulting in live birth, and the percentage of live births that were multiple births by residency status, stratified by the use of donor oocytes and further stratified by age group among the autologous oocyte embryo transfer cycles. To account for the differences in demographic and treatment characteristics between groups, we used multivariable log-binomial regression models with generalized estimating equations to evaluate the association between US-residency status and ART outcomes, stratified by oocyte source. Effect estimates are presented as crude and adjusted risk ratios (aRRs) with 95% confidence intervals (CIs). The autologous oocyte models adjusted for age and body mass index of the intended parent, fresh or frozen embryo transfer, use of PGT, and number of embryos transferred. Donor oocyte models adjusted for donor age, fresh or frozen embryo transfer, use of PGT, and number of embryos transferred. Covariates were selected for inclusion in the multivariable models if they varied by US residency status and were associated with clinically relevant outcomes following embryo transfer. A missing indicator variable was included, where necessary, to account for covariates with missing data (eg, patient and donor age, patient body mass index, and number of embryos transferred).
Results
From January 2014 through December 2022, our database included 2,275,167 ART cycles, of which 59,246 (2.6%) were reported from non-US residents. There was increasing utilization of ART by non-US residents, before declining in 2020 and 2021, coinciding with the COVID-19 pandemic ( Figure 1 ). In 2022, the number of cycles approached prepandemic levels. Trends were similar for the specific types of ART cycles. Each year, the proportion of cycles from non-US residents ranged between 1.0% and 3.5% ( Supplemental Figure 1 ). Details regarding missing data for study variables are reported in the tables and corresponding footnotes.
Assisted reproductive technology cycles by non-US residents by year, 2014-2022
Demographics
International patients were typically older than US patients, with 35% being 41 years or older at the time of their first cycle compared to 16% of US residents. Non-US residents were less likely to be nulligravid (42% vs 55%) and nulliparous (60% vs 77%) compared to US residents. A diagnosis of diminished ovarian reserve was more common among non-US patients (38%) than among US residents (26%) ( Table 1 ).
Table 1
Intended parent demographic and reproductive history characteristics for assisted reproductive technology (ART) cycles among US residents and non-US residents, 2014–2022
| Intended parent characteristics | US residents | Non-US residents | ||
|---|---|---|---|---|
| N | % | N | % | |
| 901,644 | 97% | 26,580 | 3% | |
| Patient age at first ART cycle (y) | ||||
| <35 | 384,799 | 43% | 7516 | 28% |
| 35–37 | 207,549 | 23% | 5042 | 19% |
| 38–40 | 167,711 | 19% | 4670 | 18% |
| ≥41 years | 141,585 | 16% | 9352 | 35% |
| Body mass index (kg/m 2) | ||||
| <18.5 | 19,749 | 2% | 1390 | 5% |
| 18.5–24.9 | 393,832 | 44% | 12,010 | 45% |
| 25–29.9 | 199,266 | 22% | 2441 | 9% |
| ≥30.0 | 175,646 | 19% | 1131 | 4% |
| Unknown/missing | 113,151 | 13% | 9608 | 36% |
| Patient race/ethnicity | ||||
| Non-Hispanic White | 378,238 | 42% | 3773 | 14% |
| Non-Hispanic Black | 50,949 | 6% | 451 | 2% |
| Asian/Pacific Islander | 99,932 | 11% | 10,142 | 38% |
| Hispanic | 52,230 | 6% | 839 | 3% |
| Other | 1388 | 0% | 9 | 0% |
| Unknown/missing | 318,907 | 35% | 11,366 | 43% |
| Number of prior pregnancies | ||||
| 0 | 497,955 | 55% | 11,273 | 42% |
| 1 | 183,581 | 20% | 3937 | 15% |
| >2 | 212,548 | 24% | 6775 | 25% |
| Unknown/missing | 7560 | 1% | 4595 | 17% |
| Number of prior live births | ||||
| 0 | 691,111 | 77% | 15,963 | 60% |
| 1 | 142,616 | 16% | 3598 | 14% |
| >2 | 60,193 | 7% | 2397 | 9% |
| Unknown/missing | 7724 | 1% | 4622 | 17% |
| Number of prior spontaneous abortions | ||||
| 0 | 694,079 | 77% | 16,882 | 64% |
| 1 | 125,593 | 14% | 2773 | 10% |
| >2 | 74,188 | 8% | 2301 | 9% |
| Unknown/missing | 7784 | 1% | 4624 | 17% |
| Infertility diagnosis | ||||
| Diminished ovarian reserve | 234,726 | 26% | 10,023 | 38% |
| Endometriosis | 56,548 | 6% | 761 | 3% |
| Male factor | 252,101 | 28% | 3159 | 12% |
| Ovulatory dysfunction | 35,038 | 4% | 320 | 1% |
| Tubal factor | 98,792 | 11% | 965 | 4% |
| Uterine factor | 45,701 | 5% | 1389 | 5% |
| Unexplained | 98,178 | 11% | 1649 | 6% |
| Other | 234,255 | 26% | 9878 | 37% |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
