Impact of Medicaid expansion on postoperative mortality in women with gynecologic cancer: a difference-in-difference analysis (020)

Abstract

Objectives: To compare 30 and 90-day postoperative mortality after surgery for gynecologic malignancy before and after the advent of Medicaid expansion in 2014 in states which did and did not expand Medicaid using a difference-in-difference (DID) analysis. Methods: We included women aged 40-64 years diagnosed with a primary endometrial, ovarian, or cervical malignancy in the National Cancer Database between 2010 and 2016 who underwent surgery. The outcome variables were 30-day and 90-day postoperative mortality. We used logistic regression to evaluate mortality between pre-2014 (2010-2013) and post-2014 (2014-2016) time periods for states which did expand (“expansion states”) and did not expand (“non-expansion states”) Medicaid in January 2014. We further used a quasi-experimental DID logistic regression to compare postoperative mortality after 2014 between expansion and non-expansion states. Our main regression models adjusted for cancer stage, age, race, ethnicity, comorbidity score, health insurance, geographic location, income, education, and urban-rural location. Subgroup analyses were unadjusted, odds ratios (ORs) were reported for logistic regression models, and DID ORs were reported for difference-in-difference analyses. Results: Among 169,731 women, 30-day postoperative mortality improved for endometrial cancer in expansion compared to nonexpansion states after 2014 (DID OR: 0.54, 95% CI: 0.31-0.96) on adjusted regression. There was no significant DID change in 30-day postoperative mortality for ovarian and cervical cancer between expansion and non-expansion states. However, the 30-day postoperative mortality in expansion states after 2014 significantly decreased for endometrial cancer (OR: 0.42, 95% CI: 0.26-0.67) and ovarian cancer (OR: 0.67, 95% CI: 0.46-0.99) and significantly increased for cervical cancer (OR: 3.82, 95% CI: 1.12-13.01). Univariable subgroup analysis demonstrated improved 30-day postoperative mortality for Black women with endometrial cancer in expansion compared to non-expansion states after 2014 (DID OR: 0.22, 95% CI: 0.05-0.95). There were no significant DID changes in 90-day postoperative mortality after 2014 based on expansion status for any primary cancer site. However, 90-day postoperative mortality decreased significantly for endometrial (OR: 0.66, 95% CI: 0.50-0.85) cancer in expansion states after 2014. There was improved 90-day postoperative mortality for women with ovarian cancer in the Midwest in expansion compared to non-expansion states after 2014 on univariable subgroup analysis (DID OR: 0.48, 95% CI: 0.26-0.91). Objectives: To compare 30 and 90-day postoperative mortality after surgery for gynecologic malignancy before and after the advent of Medicaid expansion in 2014 in states which did and did not expand Medicaid using a difference-in-difference (DID) analysis. Methods: We included women aged 40-64 years diagnosed with a primary endometrial, ovarian, or cervical malignancy in the National Cancer Database between 2010 and 2016 who underwent surgery. The outcome variables were 30-day and 90-day postoperative mortality. We used logistic regression to evaluate mortality between pre-2014 (2010-2013) and post-2014 (2014-2016) time periods for states which did expand (“expansion states”) and did not expand (“non-expansion states”) Medicaid in January 2014. We further used a quasi-experimental DID logistic regression to compare postoperative mortality after 2014 between expansion and non-expansion states. Our main regression models adjusted for cancer stage, age, race, ethnicity, comorbidity score, health insurance, geographic location, income, education, and urban-rural location. Subgroup analyses were unadjusted, odds ratios (ORs) were reported for logistic regression models, and DID ORs were reported for difference-in-difference analyses. Results: Among 169,731 women, 30-day postoperative mortality improved for endometrial cancer in expansion compared to nonexpansion states after 2014 (DID OR: 0.54, 95% CI: 0.31-0.96) on adjusted regression. There was no significant DID change in 30-day postoperative mortality for ovarian and cervical cancer between expansion and non-expansion states. However, the 30-day postoperative mortality in expansion states after 2014 significantly decreased for endometrial cancer (OR: 0.42, 95% CI: 0.26-0.67) and ovarian cancer (OR: 0.67, 95% CI: 0.46-0.99) and significantly increased for cervical cancer (OR: 3.82, 95% CI: 1.12-13.01). Univariable subgroup analysis demonstrated improved 30-day postoperative mortality for Black women with endometrial cancer in expansion compared to non-expansion states after 2014 (DID OR: 0.22, 95% CI: 0.05-0.95). There were no significant DID changes in 90-day postoperative mortality after 2014 based on expansion status for any primary cancer site. However, 90-day postoperative mortality decreased significantly for endometrial (OR: 0.66, 95% CI: 0.50-0.85) cancer in expansion states after 2014. There was improved 90-day postoperative mortality for women with ovarian cancer in the Midwest in expansion compared to non-expansion states after 2014 on univariable subgroup analysis (DID OR: 0.48, 95% CI: 0.26-0.91).