Background: There will be 97,220 new cases and 50,630 colon cancer deaths in the United States with over 1,000,000 survivors in 2018. Factors that could affect colon cancer survival include age, stage, treatment, and others. Previous studies have shown a statistically significant relationship between payer status and cancer patient survival in retrospective studies. However, in retrospective studies, patient baseline characteristics between payer status are not comparable. Few studies have addressed payer status's effect on the overall survival of patients using propensity score matching (PSM).

Method: About 66,493 stage II/III colon cancer patients obtained from the de-identified National Cancer Database, aged 40-64 years old, diagnosed between 2004 and 2014, were analyzed. All patients received surgery and patients who received radiation therapy, hormone therapy, immunotherapy, palliative care, or other therapies other than chemotherapy were excluded. Only private or Medicaid Payer status was included. Propensity score was calculated by computing probability of patients being in the Medicaid group using logistic regression. The PSMatch procedure from SAS was used to perform propensity score matching on patients with Medicaid and private insurance. Greedy nearest neighbor matching method was used to match one Medicaid to one privately insured patient with a caliper of 0.2. At the same time, exact match was done for gender, age group, race, and stage at diagnosis. Multivariate Cox regression was used to estimate the effect payer state effect on survival before and after PSM.

Results: Among the 66,493 patients, 90.3% were privately insured and 9.7% had Medicaid. In univariate analysis payer status was a significant predictor of overall survival. Before PSM, the median overall survival for patients with private insurance was 12.75 years while Medicaid had an MOS of 9.02 years. After PSM, 6,166 patients with private insurance had a MOS of >12.82 years and 6,167 Medicaid patients had an MOS of 8.88 years. After PSM, patients with Medicaid had a hazard ratio of 1.5, a 50% increased risk of death. With the PSM, payer status proved to be statistically significant predictor of overall survival of colon cancer.

Conclusion: This study has indicated that, by using PSM method, payer status can be shown as a significant predictor of survival for colon cancer patients. In addition, chemotherapy, race, age, and other socioeconomic factors were also significant predictors of the overall survival. Other covariates not studied or mediation effect of payer on survival of cancer patients need to be investigated in future research.

Citation Format: Runhua Shi, Lawrence Shi, Glenn Mills. Propensity score matching analysis of payer status's effect on the survival of colon cancer patients [abstract]. In: Proceedings of the Eleventh AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2018 Nov 2-5; New Orleans, LA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2020;29(6 Suppl):Abstract nr A097.