Limitations
Limitations of this study include those inherent to the SEER database. Data on patients’ overall health status and comorbidity burden were unavailable, which may impact surgical candidacy and the decision to pursue definitive treatment. We were not able to control for potentially confounding behavioral risk factors (e.g. smoking and alcohol consumption) and socioeconomic variables (e.g. education level, median household income, and metropolitan status), which may be able to explain at least some of the observed differences amongst treatment type across insurance categories, especially pre-ACA. Finally, this study is also subject to limitations inherent to the use of any large databases, such as the potential for misclassification and coding errors. In most states, individuals who are diagnosed with cancer can qualify for Medicaid, with the eligibility date assigned as the date of diagnosis. Thus, patients can move from the uninsured group to the Medicaid-insured group, confounding the classification of insurance status. In addition, it is also possible that individuals’ insurance coverage may change over the course of their treatment, which may not be captured by SEER.