The Evidence is Mounting: Insurance Coverage Disruptions Lead to Worse Outcomes among Children Diagnosed with Cancer

Highlight

In this issue of Pediatric Blood & Cancer, Johnson et al. leveraged Surveillance, Epidemiology, and End Results (SEER)-Medicaid linked data to examine 1) stage at diagnosis and 2) survival disparities in children diagnosed with cancer who were enrolled in Medicaid.¹ An additional and novel component to this study was the assessment of Medicaid coverage disruptions on delays in diagnosis and inferior survival outcomes. Not surprisingly, the authors demonstrated that children and adolescents enrolled in Medicaid are more likely to have a distant stage diagnosis and worse survival outcomes compared to children on private insurance. This is consistent with previous studies evaluating the impact of insurance coverage on outcomes among children with acute lymphoblastic leukemia and other cancers.^2–4 In fact, this observation supports the persistent disparities related to health equity among all individuals diagnosed with cancer. That said, a key limitation to these studies is that health insurance was measured at a single time point (e.g., diagnosis, initiation of therapy). By using a monthly Medicaid Flag variable in the SEER-Medicaid linked data, Johnson et al. could ascertain if a child had continuous or discontinuous coverage. Notably, significantly worse outcomes were seen among those who had disruptions in coverage.

While this may be one of the first assessments exploring Medicaid enrollment discontinuity (i.e., “churning”) among children diagnosed with cancer, as the authors indicate, there are other studies evaluating the impact of this phenomenon on other pediatric outcomes. For example, using 1996–2005 Medical Expenditure Panel Survey data, Yu et al. reported that hat children who lost Medicaid coverage and became uninsured experienced significant reductions in health care utilization, including well-child visits, physician visits, and prescription drug use, relative to their counterparts who stayed in Medicaid.⁵ These findings provide evidence that policies must be implemented that protect children from such lapses in coverage. Furthermore, this is a public health issue that demands attention from those entrusted with formulating these policies.

While the mechanisms underlying these disparities are likely to be multifactorial, it is difficult to deny the structural underpinnings of these disparities.⁶ For example, a later diagnostic stage could be due to delays in provider contact due to limited access to care. The results from this study support this in that children with enrollment discontinuity were more likely to have a distant stage diagnosis compared to children who had continuous enrollment. For inferior survival outcomes, one potential mechanism that has been proposed is lack of medication adherence associated with financial difficulties. In a systematic review by Goh et al., the authors concluded that 14 out of 15 studies evaluated supported the hypothesis that financial difficulties influenced medication adherence.⁷ It is possible that disruptions in insurance coverage could impact these financial difficulties.

Overall, there is a growing awareness of the importance of social determinants of health—the conditions in the places where people live, learn, work, and play—on outcomes among children with cancer.⁸ A recent review suggested that poverty has emerged as the primary social determinant of inferior outcomes among children diagnosed with cancer.⁶ In fact, poverty, healthcare access, and race/ethnicity interact to contribute to inferior outcomes. This is true when evaluating survival among children with acute leukemia in Texas, where those who live along the Texas-Mexico border have a 30% greater hazard of death compared to those who do not live along the border.⁹ Evaluating the impact of insurance coverage among this population would likely yield novel insights into these disparities. More importantly, there are calls to address these issues in clinical trials. For example, Bona and Keating correctly pointed out that if these factors were tumor mutations, the race to intervene would be well underway.⁸

Aside from the importance of the specific findings from this study, it also points to the utility of linking multiple data platforms to inform our understanding of cancer in children. In fact, that National Cancer Institute has invested in the Childhood Cancer Data Initiative (CCDI) with three foundational goals 1) gather data from every child, adolescent, and young adult diagnosed with a childhood cancer, regardless of where they receive their care, 2) create a national strategy of appropriate clinical and molecular characterization to speed diagnosis and inform treatment for all types of childhood cancers, and 3) develop a platform and tools to bring together clinical care and research data that will improve preventive measures, treatment, quality of life, and survivorship for childhood cancers.¹⁰ The potential power for exploring the etiologies of pediatric cancer, as well as outcomes among children diagnosed with cancer, will become more evident in the coming years, but this assessment points to the strength of these data infrastructures in understanding the epidemiology of childhood cancer.

While the evidence is mounting that insurance coverage disruptions lead to worse outcomes among children diagnosed with cancer, the challenge remains in terms of how the disparities that Johnson et al. have demonstrated will be addressed. Aside from additional research in this area, now is clearly the time to address this through novel clinical trials and policy changes.