Food Deserts Increase Readmission Following Tri-Modality Therapy for Esophageal Cancer: A Multi-Institutional Study

Purpose: Food deserts (FD) are low-income census tracts with poor access to supermarkets and have been associated with worse outcomes in breast, colon, and a small number of esophageal cancer (EC) patients.[1,2] This study investigated FD status on readmission rates in a multi-institutional cohort of EC patients undergoing tri-modality therapy.

Methods: A retrospective review of patients who underwent neo-adjuvant chemoradiation followed by esophagectomy (any type) at six high-volume academic institutions from Jan 2015 to July 2019 was performed. Demographics, nutrition parameters (including sarcopenia based on pre/post neoadjuvant psoas muscle index[2]), treatment characteristics, and 30-day complications were reviewed. Exclusion criteria included no neoadjuvant treatment and missing psoas index. FD status was defined by the USDA Food Access Research Atlas[3] and cross-referenced with patients’ home zip codes. Readmission was defined as readmission to any hospital for any reason within 30-days of esophagectomy. Univariate analysis was conducted using Student’s t-test for continuous variables and Chi square test for categorical variables. Multivariable logistic regression was then used to model readmission status on FD status adjusted for measures statistically associated with readmission with P< 0.10 in univariate analysis. The final model included FD status, any complication, and grade III/IV complications.

Results: In the initial analysis, 379 patients were included, with 67 (17.7%) residing in a FD. Pre-treatment prevalence of sarcopenia was 9.5 % (36/379) and did not vary by FD status. Univariate analysis demonstrated FD patients had significantly lower median household income with higher rates of pathologic stage I disease and 30-day readmissions (p < 0.05). No differences were seen in length of stay, complications, or mortality between FD and non-FD patients. Patients with a 30-day readmission were compared to those who were not readmitted (Table 1). Three hundred and seventy-two patients were included in this analysis after exclusion for in-hospital mortality (n=6) and missing readmission status (n=1). Univariate analysis found higher rates of FD status (p=0.006), higher initial BMI (p=0.04), presence of any complication (p < 0.001) and presence of grade III/IV complications (p < 0.001) in patients who were readmitted. Multivariable analysis for readmission showed patients who lived in a FD were 2.4 times more likely to be readmitted (OR 2.37, 95% CI 1.26-4.45) compared to those who did not. Residing in a FD remained a significant risk factor after adjusting for any complication and grade III/IV complications (AOR 2.04, 95% CI 1.03-4.06). The final model had a c-statistic of 0.77.

Conclusion: FD status is associated with 30-day readmission after esophagectomy in patients undergoing tri-modality treatment for EC in a diverse, multi-institution population. This easily identifiable risk factor may serve as a prospective risk factor for readmission and allow surgeons to identify these patients for preventative interventions prior to definitive surgical care.

Identify the source of the funding for this research project: No direct funding for this project was required.