UCSF

Aims

To determine how a diabetes diagnosis affects contraception use.

Methods

This retrospective cohort study used private insurance data from non-pregnant women aged 15-49 years, 2000-2014. We identified women with a new diabetes diagnosis and a control group without diabetes, matched on important potential confounders. We compared rates of prescription or procedural contraception use in the two groups before and after an index date (diabetes diagnosis and outpatient visit, respectively), yielding difference-in-differences estimates of the effect of a diabetes diagnosis on contraception use.

Results

We identified 75,355 women with a new diabetes diagnosis and 7.5 million women without a diabetes diagnosis. Overall rates of contraception use did not increase in the year after diagnosis (absolute difference-in-difference: 0.4% [99.9% CI, -2.1% to 2.9%]; p < 0.001). In method-specific analyses, there was a decline in estrogen-containing and injectable contraceptives in the year after diagnosis (absolute difference-in-difference: -2.2% [-4.0% to -0.4%] and -0.8% [-1.5% to -0.1%], respectively; p < 0.001); no corresponding increase was noted for intrauterine contraception or subdermal implants.

Conclusions

Women with diabetes are less likely to use contraception after their diabetes diagnosis. Efforts are needed to ensure that women with diabetes receive the counseling and clinical services needed to carefully plan their pregnancies.