Health Services Research
Purpose/Objective(s): Between 2010 and 2013, nearly 100 women’s health clinics (WHC) closed throughout the United States following restrictions placed by state legislatures related to reproductive services. In this study, we assessed whether changes in the number of WHCs providing these services during this period were associated with changes in screening for, stage at diagnosis, and mortality due to cervical cancer. We hypothesized that clinic closures would be associated with declines in screening and, consequently, worse outcomes.
Materials/Methods: The Behavioral Risk Factors Surveillance Study (BRFSS) was used to evaluate screening utilization, and data regarding stage at diagnosis and mortality were obtained using the Surveillance, Epidemiology, and End Results 18 Registry (SEER). States were divided into two cohorts—those with a decrease in the number of WHC providing abortion per capita (DIC) and those with no decrease (NDIC) between 2010 and 2013. Given this period of interest, we limited analyses to women included in the BRFSS and SEER in 2008 – 2009 and compared outcomes to women included in 2014-2015 in DIC versus NDIC states. Descriptive statistics were used to evaluate cohorts by time period, and difference-in-difference (DID) analyses with multiple imputation and Kaplan-Meier survival analysis were used to compare outcomes.
Results: A total of 197,143 cases from the BRFSS database and 10,652 patients from the SEER database were eligible for inclusion in this study. At baseline, population characteristics differed by race/ethnicity, metropolitan residence, county income and education, and insurance status between DIC and NDIC states (p<0.01 for each subgroup) in both BRFSS and SEER patients in both time periods. After DID analysis with adjustment, there was a decrease in screening in DIC states relative to NDIC states overall (-1.63 percentage points [PP], p=0.01) and among subgroups with the greatest magnitude changes noted in Hispanic women (-5.82 PP, p=0.01), women ages 21-34 (-5.19 PP, p<0.01), unmarried (-4.10 PP, p<0.003), and uninsured patients (-6.88 PP, p=0.01). In evaluating stage at diagnosis, we observed a decrease in early-stage diagnoses among 18-34-year-olds in DIC relative to NDIC states (-13.2 percentage points [PP], 95% CI = -25.17 to -1.23, p=0.031). Additionally, a significant increase in the hazard ratio (HR) of death in DIC relative to NDIC states was noted overall (HR: 1.36, 95% CI = 1.02 to 1.83, p=0.037) and among metro residents (HR: 1.40, 95% CI = 1.04 to 1.9, p=0.027).
Conclusion: The passage of laws resulting in the closure of WHCs between 2010 and 2013 was associated with adverse effects on women’s health including decreased screening, increased stage at presentation, and also higher mortality from cervical cancer, even in the short term. Though causality cannot be confirmed, these findings warrant further evaluation, particularly in the context of ongoing legal battles related to these laws.