Eye drops for dry eye disease during pregnancy and adverse neonatal outcomes: high-dimensional propensity score analyses

ABSTRACT

We investigated the association between exposure to eye drops prescribed for dry eye disease (DED) during pregnancy and adverse neonatal outcomes. Pregnant women with DED in the JMDC Claims Database (JMDC Inc., Tokyo, Japan) between 2005 and 2020 were included. According to their prescriptions during the first trimester, the women were classified into three exposed groups (hyaluronate 0.1% alone, hyaluronate 0.3% alone, and diquafosol alone) and an unexposed group (no eye drops for DED). We compared adverse neonatal outcomes (congenital anomalies, preterm birth, low birthweight, and composite outcome of these three) between the exposed and unexposed groups. We conducted a high-dimensional propensity score analysis using many variables in the database together with known potential confounders such as chronic comorbidities. We identified 4,808 eligible women, of whom 936 (19.5%) received eye drops for DED. Congenital anomalies occurred in 7.2% and 6.0%, preterm birth in 3.1% and 4.0%, low birthweight in 7.9% and 6.0%, and composite outcome in 14.9% and 12.3% of women in the hyaluronate 0.1% alone and unexposed groups, respectively. The high-dimensional propensity score analysis showed that hyaluronate 0.1% alone was not significantly associated with increases in congenital anomalies (risk difference, 0.4% [95% confidence interval, −1.9% to 2.7%]), preterm birth (−0.6% [−2.2% to 0.9%]), low birthweight (1.8% [−0.6% to 4.1%]), or composite outcome (1.9% [−1.3% to 5.1%]). Similar results were obtained in the hyaluronate 0.3% alone and diquafosol alone groups. Use of eye drops for DED in pregnant women was not associated with adverse neonatal outcomes.

KEYWORDS:

• Adverse neonatal outcome
• dry eye disease
• eye drop
• high-dimensional propensity score
• pregnancy

Acknowledgments

None.

Conflict of interest

None of the following authors have any proprietary interests or conflicts of interest related to this submission.

Data availability statement

Data not available due to commercial restrictions. This submission has not been published anywhere previously and that it is not simultaneously being considered for any other publication. We have previously submitted this paper to Ocular Surface. However, it was desk-rejected because of the difference of the scope.

Additional information

Funding

This work was supported by the Ministry of Education, Culture, Sports, Science and Technology, Japan [20H03907 and 17H05077]; Ministry of Health, Labour and Welfare, Japan [19AA2007].