Statewide trends and factors associated with genetic testing for hereditary cancer risk in Arkansas 2013-2018

Hered Cancer Clin Pract. 2022 May 23;20(1):19. doi: 10.1186/s13053-022-00226-0.

Abstract

Background: Early identification of hereditary cancer risk would save lives, but genetic testing (GT) has been inadequate. We assessed i) trends for hereditary breast and ovarian cancer (HBOC), Lynch syndrome, and other GT and ii) factors associated with receipt of GT.

Methods: We used data from the Arkansas All-Payer Claims Database from January 2013 through June 2018 (commercial, Medicaid), December 2017 (state employee), or December 2016 (Medicare) and identified enrollees with ≥1 month of enrollment. Using Current Procedural Terminology (CPT-4) codes, rates for GT were calculated per 100,000 person-quarters and time series regressions estimated. Second, GT and covariate information for enrollees with 24 months of continuous enrollment were used to estimate separate logistic regression models for each GT category.

Results: Among 2,520,575 unique enrollees, HBOC testing rates were 2.2 (Medicaid), 22.0 (commercial), 40.4 (state employee), and 13.1(Medicare) per 100,000 person-quarters and increased linearly across all plans. Older age (OR=1.24; 95%CI 1.20 - 1.28), female sex (OR=18.91; 95%CI 13.01 - 28.86), higher comorbidity burden (OR=1.08; 95%CI 1.05 - 1.12), mental disorders (OR=1.53; 95%CI 1.15 - 2.00), and state employee coverage (OR=1.65; 95%CI 1.37 - 1.97) were positively associated with HBOC testing. Less than 1 of 10,000 enrollees received Lynch syndrome testing, while < 5 of 10,000 received HBOC testing.

Conclusion: GT rates for hereditary cancer syndromes have increased in Arkansas but remain low. Receipt of GT was explained with high discrimination by sex and plan type.

Impact: Expansion of GT for hereditary cancer risk in Arkansas is needed to identify high-risk individuals who could benefit from risk-reduction strategies.

Keywords: All Payer Claims Data; Arkansas; Genetic testing; Hereditary cancers; Predictive model.