We performed a retrospective study of the prospectively maintained National Cancer Database21 (NCDB) to identify colorectal and appendiceal cancer patients who received CRS+HIPEC for the treatment of peritoneal metastases between 2010 and 2019 (n = 1,686,787). The NCDB is a joint project of the Commission on Cancer of the American College of Surgeons and the American Cancer Society. The data used in the study are derived from a de-identified NCDB file. The American College of Surgeons and Commission on Cancer have not verified and are not responsible for the analytic or statistical methodology employed or the conclusions drawn from these data by the investigator.21We selected surgically treated colorectal and appendiceal cancer patients with the following surgical codes: colon: 30, 31, 40, 50, 51, 55, 56, 57, 60, 65, 66, 70, 80; rectosigmoid: 60, 70, 80; and rectum: 30, 40, 50. To identify the study population of colorectal and appendiceal cancer patients with peritoneal metastasis who had undergone CRS+HIPEC during the relevant time period, we used the following systemic/surgery sequence codes to identify patients who had undergone CRS+HIPEC: intraoperative systemic therapy: 5; and intraoperative systemic therapy with other systemic therapy administered before or after surgery: 6. Patients who did not have details on insurance status. Follow-up, income, education, race and ethnicity, and grade were excluded. Patients under 40 years of age were, by default, excluded from the study as Medicaid expansion status is only determined in the NCDB for patients over 40 years of age. Medicaid expansion status of states in which patients were diagnosed were classified into the following categories by the NCDB: Medicaid non-expansion states, i.e. states that did expand Medicaid past 2014: TN, NC, ID, GA, FL, MO, AL, MS, KS, TX, WI, UT, SC, SD, VA, OK, NE, WY, and ME; early-expansion states, i.e. states that expanded Medicaid between 2010 and 2013: WA, CA, NJ, MN, DC, and CT; January 2014 expansion states: KY, NV, CO, OR, NM, WV, AR, RI, AZ, MD, MA, ND, OH, IA, IL, VT, HI, NY, and DE; and late-expansion states (after January 2014): NH, IN, MI, PA, AK, MT, LA.22 The case selection schema is shown in Fig. 1.Fig. 1Patient selection process. CRS cytoreductive surgery, HIPEC heated intraperitoneal chemotherapyPatient selection process. CRS cytoreductive surgery, HIPEC heated intraperitoneal chemotherapyStatistical analyses were conducted using SAS version 9.4 (SAS Institute, Inc., Cary, NC, USA). The primary exposure variable studied was the Medicaid expansion status of states in which patients were diagnosed, while the outcome variable studied was the OS, which was defined by the length of time between the date of diagnosis and the date of death or the last documented follow-up in the NCDB. Descriptive statistics were used to investigate the sociodemographic and clinical characteristics of the study population, and the Chi-square test was used to evaluate differences between Medicaid expansion and Medicaid non-expansion groups. A multivariable Cox proportional hazard regression model was performed using the a priori selection method to study the association between the Medicaid expansion status of states in which patients were diagnosed (non-expansion, early expansion, January 2014 expansion, and late expansion) and all-cause mortality after adjusting for age at diagnosis (40–65 years and >65 years), sex (male, female), race and ethnicity (non-Hispanic Black, non-Hispanic White, Hispanic, Asian/other), insurance status (uninsured, Medicaid, Medicare, privately insured, other government), program type (community, comprehensive community, academic/research, integrated network), year of diagnosis, grade (well-differentiated, moderately differentiated, and poorly differentiated), and Charlson–Deyo comorbidity score (0,1,2,3). Income and education variables were removed from the multivariable model since the proportion of missing values was >20% of the final study cohort (21.9% and 21.7%, respectively). Univariate Cox proportional hazard regression analysis was performed to assess the annual survival trend (2010–2017) in Medicaid expansion and non-expansion states. We chose 2017 as the upper limit for survival assessment, given that the median survival in this study population was 22–41 months.Of the 1001 patients, 671 (67%) were from the Medicaid expansion states and 330 (33%) were from the Medicaid non-expansion states. The majority (75.8%, n = 759) of study patients were younger than 65 years of age and had appendiceal primary (77%, n = 771). The majority of patients were of non-Hispanic White race and ethnicity (81.7%, n = 818) and had well-to-moderately differentiated tumors (75.3%, n = 754). Insurance status included 628 (62.7%) privately insured and 265 (26.5%) Medicare, 65 (6.5%) Medicaid, 24 (2.4%) other government insurance, and 19 (1.9%) uninsured. Medicaid expansion had a significantly higher proportion of Hispanic and Asian/other patients (10.3 vs. 5.2%; Chi-square test, p = 0.006), patients from residential areas with a higher median income (49.6 vs. 33%; Chi-square test, p ≤ 0.001), patients with Medicaid insurance (8.1% vs. 3.3%; Chi-square test, p < 0.001), and lower proportion of uninsured (1.6% vs. 2.4%; Chi-square test, p < 0.001) when compared with the Medicaid non-expansion group (Table 1). The median follow-up time was 53 and 48.5 months for the Medicaid non-expansion and expansion groups, respectively, and the median survival time for the Medicaid non-expansion and expansion groups was 93.8 and 107 months, respectively.Table 1Sociodemographic and clinical characteristics of the study population by Medicaid expansion statusOverall [N = 1001]No Medicaid expansion [n= 330 (33%)]Medicaid expansion [n= 671 (67%)]p valueaN%n%n%Age, years40–6575975.8254.077.0505.075.30.553Age >6524224.276.023.0166.024.7SexMale48048.0156.047.3324.048.30.763Female52152.1174.052.7347.051.7Race and ethnicityNon-Hispanic White81881.7273.082.7545.081.20.006Non-Hispanic Black979.740.012.157.08.5Hispanic484.813.03.935.05.2Asian/other383.84.01.234.05.1Income (residence area median income), missing = 219< $40,22711915.263.022.856.011.1<0.001$40,227–50,35314118.062.022.579.015.6$50,354–63,33218023.060.021.7120.023.7> $63,33234243.791.033.0251.049.6Education (residence area high-school graduation failure proportion), missing = 217≥ 17.6%12015.346.016.774.014.60.14810.9–17.5%17922.873.026.5106.020.96.3–10.8%24631.484.030.4162.031.9< 6.3%23930.573.026.5166.032.7Insurance statusMedicaid656.511.03.354.08.1<0.001Medicare26526.585.025.8180.026.8Privately insured62862.7210.063.6418.062.3Uninsured191.98.02.411.01.6Other government insurance242.416.04.98.01.2ProgramCommunity Cancer Program20020.054.016.4146.021.80.121Academic/Research Program66466.3231.070.0433.064.5Integrated Network Cancer Program13713.745.013.692.013.7Charlson–Deyo comorbidity082382.2268.081.2555.082.70.947114014.049.014.991.013.62212.17.02.114.02.13171.76.01.811.01.6GradeWell-differentiated37437.4138.041.8236.035.20.043Moderately differentiated38038.0108.032.7272.040.5Poorly differentiated24724.784.025.5163.024.3aChi-square test valueSociodemographic and clinical characteristics of the study population by Medicaid expansion statusaChi-square test p valueIn the multivariable Cox regression model, the grade of the disease and being diagnosed in Medicaid early-expansion states were found to be independently associated with decreased hazard of death, i.e., all-cause mortality (increased OS). Hazard of death (all-cause mortality) in patients with moderately and poorly differentiated tumors were 2.1 and 4.7 times higher, respectively, than in those with well-differentiated tumors (p < 0.001). It is important to note that the hazard of death was significantly lower in patients from the early-expansion states compared with those from the non-expansion states (hazard ratio (HR) 0.71, 95% confidence interval (CI) 0.52–0.97, p = 0.029). Age, sex, race and ethnicity, insurance type, program type, year of diagnosis, and comorbidity index were not independently associated with all-cause mortality (Table 2).Table 2Multivariable Cox regression analysis model for all-cause mortality (n = 939)HR95% CIp valueLower limitUpper limitAge, years40–641.060.761.460.74965+RefSexFemale0.880.721.090.245MaleRefRace/ethnicityAsian/other0.780.441.380.395Non-Hispanic Black1.140.811.600.447Hispanic1.060.631.800.817Non-Hispanic WhiteRefGradeWell-differentiatedRefModerately differentiated2.121.612.78<0.001Poorly/undifferentiated4.713.596.18<0.001InsuranceMedicaid0.760.481.200.234Medicare1.240.911.680.174Other Government1.040.581.860.897Not insured0.760.331.720.504Private insuranceRefMedicaid expansion statusEarly-expansion states0.710.520.970.029January 2014 expansion states0.900.701.150.399Late-expansion states1.030.761.390.845Non-expansion statesRefCharlson–Deyo score0Ref11.190.891.580.23620.890.411.920.76631.800.923.530.086ProgramAcademic/Research0.940.731.210.630Integrated Network1.070.761.510.700Community Cancer ProgramRefYear of diagnosisAnnual0.970.931.020.184HR hazard ratio, CI confidence interval, Ref referenceMultivariable Cox regression analysis model for all-cause mortality (n = 939)HR hazard ratio, CI confidence interval, Ref referenceWhen analyzing the trend of OS based on the year of diagnosis (2010–2019) using a univariate Cox regression model, the Medicaid non-expansion states had a 1% annual OS decrement (HR 1.011, 95% CI 0.94–1.09, p = 0.773), whereas Medicaid expansion states had a 4% annual survival improvement (HR 0.96, 95% CI 0.91–1.01, p = 0.123). The above improvement in the annual trend of OS is shown in Fig. 2, which compares the survival curves of patients diagnosed in 2010 and 2017 between the Medicaid expansion and non-expansion groups.Fig. 2Trend of overall survival in (A) Medicaid expansion states versus (B) non-expansion statesTrend of overall survival in (A) Medicaid expansion states versus (B) non-expansion statesThis study investigated the association between state Medicaid expansion status and survival outcomes in patients undergoing CRS+HIPEC for colorectal and appendiceal malignancies. We found that Medicaid early-expansion status and grade are independent predictors of survival. Early adopters of Medicaid expansion often face unique challenges and opportunities, leading to innovative practices or policy adjustments that may not be immediately replicated by late adopters. Changes in the Medicaid expansion program over time may also influence the outcomes observed in late-expansion states. Furthermore, compared with those without Medicaid expansion, year-over-year survival improvements were observed in patients with colorectal and appendiceal malignancies treated with CRS+ HIPEC in states with Medicaid expansion. These findings suggest that Medicaid expansion has improved survival outcomes likely by improving access to healthcare and narrowing the equity gap in cancer care and outcomes. Our data on travel distances provide compelling evidence supporting the improved access to healthcare in Medicaid expansion states. Patients in Medicaid expansion states had a significantly lower median travel distance to the hospital {19.3 miles (interquartile range [IQR] 51.7 miles) vs. 39.95 miles (IQR 114.6 miles), p < 0.0001}.Compared with prior studies, the main strength of this work is its large sample size for a rare patient population. NCDB captures approximately 72% of newly diagnosed cancer cases in the US.21 Previous studies examining the relationship between insurance status and survival in this patient population were limited by small sample sizes, did not specifically focus on gastrointestinal malignancies, and yielded conflicting results.15,16One single-institution study with 31 patients found no impact of insurance on the survival of patients who underwent CRS+HIPEC for various primary and secondary peritoneal carcinomatoses.15 Another retrospective cohort study, also from a single institution, showed lower OS in the underinsured group (including Medicaid and uninsured patients) compared with the privately insured group in patients who received CRS+HIPEC for peritoneal carcinomatosis.16While prior studies have demonstrated disparities in outcomes for CRS+HIPEC in various patient populations, this study found no association between age, sex, race and ethnicity, program type, comorbidities, and survival in patients undergoing CRS+HIPEC. It is worth mentioning that we did not find an association between the female sex and improved survival as compared with a study focused solely on appendiceal malignancy patients who underwent CRS+HIPEC.23 This is likely due to the difference in the study cohort, i.e., our study included both colorectal and appendiceal malignancies, unlike the other study that included only appendiceal malignancies.Several limitations of this study should be acknowledged. The need for specific procedural codes for CRS in the NCDB could impact the accuracy of the analysis. However, our method has been standardized with concurrent treatment regimens only particular to CRS+HIPEC. Disease-specific survival data are not available in the NCDB; therefore, our analysis relies on OS as the primary outcome measure. Incomplete data and loss of follow-up are inherent limitations of retrospective analyses and could introduce bias. The NCDB collects data from the Commission on Cancer-accredited programs, potentially limiting generalizability to patients treated outside these centers.The implications of these findings are significant for public health and clinical medicine. The study provides evidence that Medicaid expansion is associated with improved survival outcomes in patients undergoing CRS+HIPEC for colorectal and appendiceal malignancies.Additional studies should address the limitations identified in this study. Institutional or registry studies may allow for the collection of additional data, such as the Peritoneal Carcinomatosis Index (PCI), which is considered an essential prognostic factor in patients undergoing CRS+HIPEC. Additionally, specific procedural codes for CRS and HIPEC, as well as disease-specific survival, would enable more accurate analysis and interpretation.24,25 Further research is also warranted to explore the impact of insurance status on survival outcomes in other cancer types and patient populations.Our study demonstrates a significant association between residing in a state with Medicaid expansion and improved OS for patients undergoing CRS+HIPEC for lower gastrointestinal cancer with peritoneal metastases. This finding suggests that increased access to healthcare, potentially facilitated by Medicaid expansion, may be a contributing factor to improved patient outcomes.