Fluconazole in early pregnancy and birth defects: a cohort study

In this large US Medicaid cohort, first-trimester fluconazole prescription fills were associated with a small absolute increase in claims-identified musculoskeletal malformations among live births compared with topical-azole fills. The study did not find statistically significant increases in its conotruncal or oral-cleft estimates. It does not estimate miscarriage, stillbirth, neonatal death, or an individual pregnancy's risk, and a filled prescription does not show that a medicine was taken.

The cohort adds a large, adjusted comparison of malformation outcomes after first-trimester oral fluconazole and topical-azole prescription fills. Its outcome is narrower than “all pregnancy outcomes”: the analysis concerns claims-identified malformations among live births, so it cannot answer questions about pregnancy loss or neonatal death.

This retrospective population-based cohort used US Medicaid claims from 2000 through 2014. It included 1,969,954 pregnancies overall, including 37,650 with a first-trimester fluconazole fill and 82,090 with a topical-azole fill. The malformation analysis covered live births and identified outcomes in claims through 90 days after delivery.

The study compared prescription fills for oral fluconazole with fills for topical azoles during the first trimester. Its outcomes were claims-identified major congenital malformations, including musculoskeletal, conotruncal, and oral-cleft categories. A prescription fill is a measure of dispensed medicine, not confirmation of ingestion. The study did not assess treatment effectiveness or compare oral and topical treatment for curing vaginal yeast infection in pregnancy.

Among pregnancies resulting in live births, claims-identified musculoskeletal major malformations occurred at 52.1 per 10,000 fluconazole-exposed pregnancies and 37.3 per 10,000 topical-azole-exposed pregnancies. The adjusted relative risk was 1.30 (95% CI 1.09–1.56); the adjusted risk difference was 12.15 per 10,000 pregnancies (95% CI 3.69–20.60).

For conotruncal malformations, the adjusted relative risk was 1.04 (95% CI 0.70–1.55); for oral clefts it was 0.91 (95% CI 0.61–1.35). These estimates did not show statistically significant increases. In a smaller group with more than 450 mg of fluconazole exposure, the adjusted relative risk for musculoskeletal malformations was 1.98 (95% CI 1.23–3.17). That dose-stratum estimate is less precise and applies to this outcome in that group.

These results describe a claims-based US Medicaid comparison of first-trimester prescription fills and selected malformations diagnosed among live births. They do not measure medicine ingestion, the chance of a pregnancy loss, stillbirth, or neonatal death, or a person's own baseline risk. The topical-azole group is a comparator for malformation outcomes; this study does not compare the effectiveness of oral and topical treatment for vaginal yeast infection during pregnancy.

Prescription claims record dispensing, not whether or how a medicine was taken. The analysis identifies malformations through 90 days after delivery and is restricted to live births; it cannot be used to estimate miscarriage or other non-live-birth outcomes. Like other observational comparisons, the estimates may be affected by differences in the reason for treatment and other factors not fully captured in claims. The higher-dose analysis involved a smaller group. A relative or absolute group estimate from this study cannot be converted into an individual risk without relevant baseline and clinical information.

Supporting for one specific outcome: this large cohort found an adjusted association between first-trimester fluconazole fills and musculoskeletal malformations among live births, with a small adjusted absolute difference. It did not find statistically significant increases for conotruncal malformations or oral clefts. The study is one contribution to a larger observational evidence base and does not establish a causal effect or resolve other pregnancy outcomes.

No practical change. The cohort adds context about selected malformations but does not set pregnancy treatment guidance. Use the linked local treatment guidance and its authority records for the applicable action; those recommendations are not inferred from this study's estimate.

The study was funded by the Division of Pharmacoepidemiology and Pharmacoeconomics at Brigham and Women's Hospital; KJG received a K12 BIRCWH career development award from the Office of Research on Women's Health to Harvard Medical School. The funders had no role in the study design, conduct, interpretation, writing, or submission decision.

All authors disclosed support for the submitted work from the Brigham and Women's Hospital division. KFH reported NIH grants during the study and institutional grants from Eli Lilly, GlaxoSmithKline, and Boehringer Ingelheim outside it. BTB reported NIH grants during the study, institutional grants from Eli Lilly, GlaxoSmithKline, Baxalta, Pacira, and Pfizer outside it, and consulting fees from Merck for Mothers, Aetion, and the Alosa Foundation. KJG reported consulting and educational-speaking work on expanded carrier screening and preterm birth for Quest Diagnostics, service on an Illumina advisory panel about predicting adverse pregnancy outcomes, and personal fees from BillionToOne. SH-D reported NIH grants during the study, institutional grants from Lilly, Takeda, Pfizer, and GlaxoSmithKline, personal fees from UCB outside it, and service as an advisor to the Antipsychotics Pregnancy Registry and an epidemiologist for the North American Antiepileptics Pregnancy Registry, both at Massachusetts General Hospital. The remaining authors declared no conflicts.

Read the original publication (published 20 May 2020).