Xia Wang, Maxine D. Chang, Marie Catherine Lee, and Bethany L. Niell
Alexander P. Cole, Chang Lu, Marieke J. Krimphove, Julie Szymaniak, Maxine Sun, Sean A. Fletcher, Stuart R. Lipsitz, Brandon A. Mahal, Paul L. Nguyen, Toni K. Choueiri, Adam S. Kibel, Adil H. Haider, and Quoc-Dien Trinh
Background: Insurance coverage is associated with better cancer outcomes; however, the relative importance of insurance coverage may differ between cancers. This study compared the association between insurance coverage at diagnosis and cancer-specific mortality (CSM; insurance sensitivity) in 6 cancers. Patients and Methods: Using the SEER cancer registry, data were abstracted for individuals diagnosed with ovarian, pancreatic, lung, colorectal, prostate, or breast cancer in 2007 through 2010. The association between insurance coverage at diagnosis and CSM was modeled using a Fine and Gray competing-risks regression adjusted for demographics. An interaction term combining insurance status and cancer type was used to test whether insurance sensitivity differed between cancers. Separate models were fit for each cancer. To control for lead-time bias and to assess whether insurance sensitivity may be mediated by earlier diagnosis and treatment, additional models were fit adjusting for disease stage and treatment. Results: Lack of insurance was associated with an increased hazard of CSM in all cancers (P<.01). The magnitude of the effect differed significantly between cancers (P interaction=.04), ranging from an adjusted hazard ratio of 1.13 (95% CI, 1.01–1.28) in ovarian and 1.19 (95% CI, 1.11–1.29) in pancreatic cancer to 2.19 (95% CI, 2.02–2.37) in breast and 2.98 (95% CI, 2.54–3.49) in prostate cancer. The benefit of insurance was attenuated after adjusting for stage and treatment (eg, screening/early treatment effect), with the largest reductions in prostate, breast, and colorectal cancers. Conclusions: Greater insurance sensitivity was seen in screening-detected malignancies with effective treatments for early-stage disease (eg, prostate, breast, and colorectal cancers). Given that this differential is significantly reduced after adjusting for stage and treatment, our results suggest that a significant portion (but not all) of the benefit of insurance coverage is due to detection and treatment of certain curable early-stage cancers.