Cervical Cancer Screening Guideline Adherence Before and After Guideline Changes in Pennsylvania Medicaid

To assess changes in cervical cancer screening after the 2009 American College of Obstetricians and Gynecologists' guideline change and to determine predictors associated with underscreening and overscreening among Medicaid-enrolled women.

We performed an observational cohort study of Pennsylvania Medicaid claims from 2007 to 2013. We evaluated guideline adherence of 18- to 64-year-old continuously enrolled women before and after the 2009 guideline change. To define adherence, we categorized intervals between Pap tests as longer than (underscreening), within (appropriate screening), or shorter than (overscreening) guideline-recommended intervals (±6-month). We stratified results by age and assessed predictors of underscreening and overscreening through logistic regression.

Among 29,650 women, appropriate cervical cancer screening significantly decreased after the guideline change (from 45% [95% confidence interval (CI) 44-46%] to 11% [95% CI 11-12%] among 17,360 younger than 30 year olds and from 27% [95% CI 26-28%] to 6% [95% CI 6-7%] among 12,290 women 30 years old or older). Overscreening significantly increased (from 6% [95% CI 5-6%] to 67% [95% CI 66-68%] in those younger than 30 years old and from 54% [95% CI 52-55%] to 65% [95% CI 64-67%] in those 30 years old or older), whereas underscreening significantly increased only in those 30 years old or older (from 20% [95% CI 19-21%] to 29% [95% CI 27-30%]). Pap tests after guideline change, pregnancy, Managed Care enrollment (in those younger than 30 years old), and black race (in those younger than 30 years old) were associated with underscreening. Pap tests after guideline change, more visits, more sexually transmitted infection testing, and white race (in those 30 years old or older) were associated with overscreening.

We observed high rates of cervical cancer overscreening and underscreening and low rates of appropriate screening after the guideline change. Interventions should target both underscreening and overscreening to address these separate yet significant issues.