Using National IVF Registries to Validate Clinical Outcomes Following IVF Covered by Health Insurance

David Suh; Phyllis Yan; Rodney L Dunn; Edward C Norton; Vanessa K Dalton; Erica E Marsh; Marissa S Weiss; James M Dupree

doi:10.1016/j.fertnstert.2024.12.015

Using National IVF Registries to Validate Clinical Outcomes Following IVF Covered by Health Insurance

Fertil Steril. 2024 Dec 12:S0015-0282(24)02445-2. doi: 10.1016/j.fertnstert.2024.12.015. Online ahead of print.

Authors

David Suh¹, Phyllis Yan², Rodney L Dunn², Edward C Norton³, Vanessa K Dalton⁴, Erica E Marsh⁴, Marissa S Weiss⁵, James M Dupree⁶

Affiliations

¹ Department of Health Management and Policy, University of Michigan, Ann Arbor, MI. Electronic address: [email protected].
² Department of Urology, University of Michigan, Ann Arbor, MI.
³ Department of Health Management and Policy, University of Michigan, Ann Arbor, MI.
⁴ Department of Obstetrics and Gynecology, University of Michigan, Ann Arbor, MI.
⁵ Division of Reproductive Endocrinology and Infertility, Department of Obstetrics and Gynecology, University of Pennsylvania, Philadelphia, PA.
⁶ Department of Urology, University of Michigan, Ann Arbor, MI; Department of Obstetrics and Gynecology, University of Michigan, Ann Arbor, MI.

PMID: 39674336
DOI: 10.1016/j.fertnstert.2024.12.015

Abstract

Objective: To evaluate in vitro fertilization (IVF) cycles covered by health insurance using a national commercial claims database and to validate key clinical events against national IVF registries.

Design: Retrospective cohort study SUBJECTS: US women aged 20-44 who underwent IVF from 2005 to 2020 in Optum's de-identified Clinformatics® Data Mart Database (CDM).

Exposure: Undergoing IVF MAIN OUTCOME MEASURES: IVF cycles and rates of pregnancies (inclusive of losses and terminations), live births, and live birth types (e.g., singleton, twin, and triplet or higher order).

Results: We identified more than 3,000 IVF cycles in each year from 2005 to 2020 within CDM. When comparing our rates of clinical outcomes to external benchmark data, the results were similar across all the years of our study. For example, in 2020 the percentage of pregnancies after first embryo transfer was 62.03% (95% CI: 59.48-64.47) in CDM and 64.96% in data from the Society for Assisted Reproductive Technology (SART). The rate of live birth after first embryo transfer was 44.58% (95% CI: 41.90-47.21) in CDM in 2020 and 46.95% in SART in 2020. The rate of singleton births was 94.17% (95% CI: 92.24-96.10) in CDM in 2020; and this rate was 94.37% in SART in 2020. For twin births, the rate was 5.48% (95% CI: 3.60-7.35) in CDM in 2020 and 5.46% in SART in 2020. The rate of triplet or higher order births was 0.35% (95% CI: 0.00-0.84) in CDM in 2020 and 0.17% in SART in 2020.

Conclusion: We found CDM can be used to accurately identify IVF cycles covered by insurance and key clinical outcomes such as rates of pregnancies, live births, and live birth types, and our reported rates were similar to national IVF clinical registry data. Our findings support that CDM is a robust data source to conduct research about IVF insurance coverage and can accurately evaluate clinical outcomes resulting from IVF. Policymakers who are considering insurance coverage for IVF can use CDM to model and measure the impact of new or existing policies for IVF insurance coverage.

Keywords: In vitro fertilization; health insurance; infertility; live birth; pregnancy.