The impact of high-risk zip code on receiving standard of care therapy and survival in esophageal adenocarcinoma

Introduction

Disparities in oncologic outcomes driven by socioeconomic status (SES) have been detailed in a variety of cancers, including esophageal malignancies (1,2). Current literature offers that patients of lower SES experience greater delays in care and a reduction in likelihood of receiving guideline-concordant therapy (3,4). As esophageal adenocarcinoma (EAC) remains a prevalent and persistently morbid disease in the U.S., conversation to mitigate the disparities conferred by race, ethnicity, income, and other factors is of utmost importance (5,6).

Several studies have detailed the impact of individual disparaging factors such as race, type or lack of insurance, education level, and income on therapy received and survival in EAC (7-12). Despite knowledge of these trends, there remains a paucity of data in the post-CROSS era and much of the current literature focuses on isolated elements of SES without identification of a more intersectional variable. A similar effort to identify an adequate surrogate for large-scale SES analysis has been undertaken in lung cancer research, utilizing zip code of residence as an easily identifiable metric (13). As such, we sought to evaluate the impact of zip code as a surrogate for SES and associated negative social determinants of health (SDOH) on the likelihood of receiving standard-of-care therapy and associated overall and median survival in EAC.

While quantifying the effect of disparities is essential, establishment of actionable interventions in such groups offers an avenue for impactful change and remains the ultimate goal for such research. This undertaking first requires a methodology for identifying at-risk patients. As an objective and easily identifiable metric, zip code offers a simple form of risk stratification. We are hopeful that the determination of high-risk demographic groups will translate to larger team-based interventions with the goal of mitigating disparities in EAC. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-637/rc).

Methods

The National Cancer Database (NCDB) is one of the largest cancer registries, capturing approximately 70% of all new cancer diagnoses in the United States annually (14). We queried the NCDB from 2012–2021 to identify patients with non-metastatic EAC. Those with clinical stage (cSTAGE) 1 disease were excluded as the optimal treatment approach in such patients remains highly debated. The NCDB crossmatches patient zip code of residence with the American Community Survey census data to determine median household income. High school education achievement status is determined in a similar fashion. These two variables were utilized to create a “high-risk” zip code group. Those patients in the lower two quartiles of median income (<$50,353) or education (>10.9% without a high school diploma), or both, were classified as a high-risk zip code of residence. The remaining patients were classified as a low-risk zip code group. Patients without income or high school education quartile data were excluded from the study. As this study utilized publicly available de-identified data, ethical and institutional review board approval was not necessary. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Statistical analysis

Comparisons between the high-risk and low-risk zip code groups were made using Chi-squared and equal-variance t-test where appropriate. Multivariable Cox proportional hazard models were employed to evaluate the association of zip code with overall survival (OS). Kaplan-Meier survival analyses were completed. Analyses were performed in SAS v9.4 (SAS Institute; Cary, NC, USA) in concert with a statistician at the Medical College of Wisconsin.

Results

Of the 10,520 patients identified, 25.01% resided in a high-risk zip code (Figure 1). Those in the high-risk group were slightly younger (62.7 vs. 64.1 years, P<0.001), more likely to be of Black race (3.53% vs. 1.13%, P<0.001) or Hispanic ethnicity (5.20% vs. 1.40%, P<0.001) and lacking insurance (2.66% vs. 1.15%, P<0.001) or on government insurance (57.85% vs. 52.77%, P<0.001). There were no differences between the two groups in clinical stage at diagnosis (P=0.18), pathologic stage at resection (P=0.31), or rate of lymphovascular invasion (LVI) (P=0.20). There was also no difference in the type of center where care was received (academic vs. community) between the two groups. In an extended model, clinical stage at diagnosis was evaluated, including patients with cSTAGE I–IV disease. Again, there was no difference in distribution amongst the high and low-risk zip code groups (P=0.18), even among those with earlier stage or metastatic disease (Table S1).

Figure 1 CONSORT diagram.

Irrespective of treatment regimen, high-risk zip code served as a negative independent prognostic factor for OS [hazard ratio (HR) 1.196, 95% confidence interval (CI): 1.10–1.29, P<0.001] (Table S2). Associated reductions in median survival (high-risk 36.90 months, 95% CI: 34.40–40.05, low-risk 43.70 months, 95% CI: 41.69–46.19) and 5-year OS (high-risk 37.88%, low-risk 42.97%, P<0.001) were observed.

In this cohort (n=10,520), 89.33% (n=9,398) received standard-of-care trimodal therapy. Of those that did receive trimodal therapy, 7,077 (75.30%) resided in a low-risk zip code. Similar to the cohort of allcomers, those in the high-risk group of patients receiving trimodal therapy were older, more likely to be of Black race or Hispanic ethnicity, lacking insurance or on government insurance, and to have positive margins on resection. There remained no differences between the two groups in clinical stage, pathologic stage, or rate of LVI (Table 1). In a multivariable model, high-risk zip code was independently associated with reduced odds of receiving guideline-concordant trimodal therapy [odds ratio (OR) 0.764, 95% CI: 0.59–0.99, P=0.047] (Table 2). Despite receiving CROSS-concordant therapy, high-risk zip code persisted as an independent negative prognostic factor for survival (HR 1.189, 95% CI: 1.09–1.29, P=0.001) (Table 3). A reduction in median survival was observed in the high-risk zip code group (38.41 months, 95% CI: 35.98–43.24) compared to the low-risk zip code group (44.65 months, 95% CI: 42.15–47.15). Similarly, 5-year OS was reduced in the high-risk zip code group at 39.12% versus 43.31% in the low-risk group (P=0.001) (Figure 2).

Figure 2 Overall survival in patients with EAC receiving trimodal therapy by zip code group. EAC, esophageal adenocarcinoma.

A significant delay between diagnosis and esophagectomy was observed in the high-risk group (152.5±47.08 days) when compared to the low-risk group (143.5±41.98 days, P<0.001), likely driven by a prolongation between diagnosis and initiation of chemotherapy (high-risk 45.40±25.54 days, low-risk 40.70±26.40 days, P<0.001). The low-risk zip code group also demonstrated a statistically higher mean number of lymph nodes harvested (17.65±9.24) than the high-risk group (16.82±9.30, P=0.001), though absolute difference was less than one node.

In an effort to understand the observed survival benefit of a low-risk zip code, subsequent analyses were conducted evaluating specific pathologic data. On univariate analysis, the high-risk zip code group demonstrated a higher incidence of positive margins compared to the low-risk group (8.99% versus 6.95%, P=0.009). Subsequent multivariate analysis revealed that zip code of residence was not independently associated with margin positivity (OR 1.280, 95% CI: 0.93–1.76, P=0.13). In additional multivariate models, zip code of residence did not reveal an association with likelihood of upstaging (OR 0.958, 95% CI: 0.79–1.17, P=0.67) or downstaging (OR 0.985, 95% CI: 0.85–1.15, P=0.85) at resection.

Subset analyses evaluating the more immediate post-operative course were also completed. The high-risk group demonstrated a slightly prolonged postoperative length of stay (LOS) (12.97±9.5 versus 11.65±9.54 days, P<0.001). 30-day mortality rates were comparable between groups (high risk 2.67%, low-risk 2.13%, P=0.26). Similarly, there was no statistical difference in 90-day mortality rates between groups (high-risk 5.99%, low-risk 5.12%, P=0.14). In multivariate modeling, zip code of residence was not associated with early postoperative mortality (30-day: OR 0.787, 95% CI: 0.54–1.15, P=0.23; 90-day: OR 1.146, 95% CI: 0.86–1.53, P=0.35).

Discussion

It has been well established that lower SES is associated with poor outcomes in a multitude of cancers. While various sociodemographic factors have been investigated in EAC, there remains no single, easily discernible surrogate for SES which offers tangible opportunity for intervention (7,8,15). While understanding the implications of individual patient factors is certainly important, identification of a targetable metric with the goal of team-based equitable care offers an avenue for impactful change where disparities persist. Erhunmwunsee et al. demonstrated the utility of income and education derived from zip code as surrogates for SES over race in the pre-CROSS era, however, limited analyses exist in the more recent era of standardized trimodal therapy (16). As such, our aim was to evaluate the potential utility of zip code of residence when identifying patients at an increased risk for poor outcomes in EAC.

The NCDB currently serves as the largest and most representative cancer registry in the United States. Collecting data from over 1,500 Commission on Cancer-accredited facilities, it offers insight to socioeconomic trends that cannot be captured in single-institution studies (14). The NCDB uses patient zip code cross-referenced with U.S. Census data to infer median household income and percent of residents without a high-school diploma. We utilized these variables based on zip code of residence to create a “high-risk” demographic group comprised of patients in the lower two quartiles of median household income (<$50,353) and/or rate of individuals without a high school degree (>10.9%). While income and education level are often used interchangeably, their inherent impact as independent SDOH is quite different. Education level is often a representation of SES earlier in life and thus is thought to be reflected in disease onset, while income is more strongly associated with care compliance and disease progression (17). As such, we believe the combination of the two variables offers a more comprehensive surrogate for SES.

When evaluating nearly 11,000 patients with EAC, irrespective of treatment modality, almost a quarter of the population resided in a high-risk zip code. This datapoint alone serves to emphasize the gravity of additional conversation in this arena, with the understanding that one in every four patients seeking treatment for EAC is at risk of worse outcomes based solely on where they live. Interestingly, despite the study being limited to the post-CROSS era in patients with non-metastatic cSTAGE II and III EAC, we found that approximately 10% of patients did not receive standard-of-care trimodal therapy. Those residing in a high-risk zip code demonstrated reduced odds of receiving trimodal therapy (OR 0.764, P=0.047), again despite comparable disease-related factors and comorbidities between groups. The NCDB does not offer insight as to why patients did not receive neoadjuvant chemoradiation followed by esophagectomy; however, less than 0.5% of all patients were identified as refusing one or more elements of therapy.

It is often presumed that those at a sociodemographic disadvantage present for evaluation later with more advanced disease, given the inherent barriers to healthcare access alongside reduced health literacy. Interestingly, there was no difference in the distribution of clinical stage at diagnosis in our study cohort comprised of those with cSTAGE II–IIIb disease (P=0.06). To evaluate for potential selection bias, we completed an extended analysis of patients presenting with cSTAGE I–IV EAC and noted a similar trend, with no significant difference in distribution of stage at diagnosis between the high and low-risk zip code groups (P=0.18). We further compared the high and low-risk groups and noted differences only in age (64.0 vs. 62.7 years, P<0.001), prevalence of Black race (3.53% vs. 1.13%, P<0.001), Hispanic ethnicity (5.20% vs. 1.40%, P<0.001), and non-private payor rates (61.78% vs. 55.23%, P<0.001). There was no observed difference in type of treatment facility (academic vs. non-academic, P=0.97) or pathologic stage (P=0.63). We find this similarity in disease distribution to be of marked importance, given that much of the disparity observed across much wider oncologic outcomes is often attributed to more advanced stage disease and other clinical characteristics. Additionally, the differences noted in race, ethnicity, and insurance status further support the utility of zip code as a surrogate for larger SES, given that those negative SDOH examined in previous literature (Black race, Hispanic ethnicity, government insurance) remained more prevalent in the high-risk zip code group. We also find it important to highlight the lack of difference in treatment center type between the two groups, given that this is a frequently cited cause for disparities in outcomes.

Among the nearly 10,000 patients who did receive trimodal therapy, high-risk zip code persisted as an independent negative prognostic factor for survival, with a greater than 6-month discrepancy in median survival between high and low-risk groups. The 5-year OS was also significantly impacted with a nearly 5% reduction observed in those living in a high-risk zip code. Similar rates of upstaging and downstaging at resection were seen between groups, with no observable impact of zip code when considering treatment response. Additionally, zip code was not associated with complete pathologic response (18). These trends are notable—patients at opposite ends of the socioeconomic spectrum present with equivalent clinical disease and appear to respond similarly to neoadjuvant therapy, with comparable pathologic stages at resection (19). Despite this, there persists a discrepancy in survival between the two groups, necessitating the consideration of socioeconomic factors and SDOH in the high-risk group as an opportunity for intervention.

It has been proposed that treatment center volume has an impact on outcomes in esophageal cancer. Center volume is not explicitly reported in the NCDB but rather is stratified by treatment center type as derived from annual cancer diagnoses. While center type can be utilized to loosely infer nonspecific cancer case volume, this is not a reliable metric for operative expertise when examining survival outcomes (20). To this extent, Rhodin and colleagues recently published a model identifying an inflection point of improved survival at >10 esophagectomies performed annually (21). In their analysis, patients residing in zip codes with a lower education level (a metric used to create our high-risk zip code group) were more likely to receive treatment at a lower volume center, though specific subset analyses looking at this group were not completed. As such, the role of treatment center volume on outcomes in patients at a sociodemographic disadvantage remains unclear and is of vital importance for future investigations. The aim of our study was to evaluate the utility of zip code as a surrogate for larger sociodemographic status in the current treatment paradigm. Given that it is not standard practice to refer high-risk sociodemographic patients to high-volume centers, volume was not a component of our analysis. We intend to further delineate the role of center volume as a potentially mitigating factor for disparities conferred by sociodemographic status in future studies.

We completed several subset analyses to better evaluate the independent impact of SES on survival, given known confounding factors, including lymph node harvest and margin positivity, that serve as objective surrogates for expertise. Ongoing debate regarding lymph node harvest in this group suggests a survival benefit with a more extensive lymphadenectomy. In our cohort, the low-risk zip code group demonstrated a slightly increased mean number of lymph nodes harvested; however, this difference was less than one node (17.65 vs. 18.92 nodes, P=0.003) and likely has quite limited clinical significance. When evaluating margins, the high-risk zip code group demonstrated an increased rate of R1 resection (9% vs. 7%, P=0.009), however, on multivariate analysis there was no association between zip code and margin positivity (P=0.13).

We also sought to evaluate short-term outcomes to determine if variation in postoperative course was driving the observed disparity in survival. Despite the lack of postoperative complication data in the NCDB, we conducted analyses evaluating 30-day and 90-day mortality and LOS as a surrogate for postoperative course. A difference was observed in LOS (13 days in the high-risk group, 9.5 days in the low-risk group, P<0.001), but 30- and 90-day mortality rates were comparable between groups (P=0.26, P=0.14). The lack of notable difference in short-term survival implies comparable immediate postoperative outcomes in our two patient groups, irrespective of potential differences in treatment center type or expertise. LOS data is historically much more challenging to interpret given disposition and discharge barriers in those of lower SES that may prolong inpatient stay.

We believe these findings to be consequential. Despite wide acceptance of poor outcomes in those at a sociodemographic disadvantage, a deeper comprehension of the underlying etiology for such disparities remains evasive in oncologic care. Historically, it was believed that a lack of access and subsequent delays in care were the root cause of reduced survival for those patients belonging to at-risk groups. Our study highlights the multifactorial nature of healthcare disparities extending beyond access to care, given comparable clinical features, disease stage, and high-risk characteristics across both high- and low-risk groups. Our high-risk zip code cohort did demonstrate a negative association with receipt of guideline-concordant trimodal therapy, bringing to question both patient education and cultural barriers to care.

In subsequent analyses of those patients receiving trimodal therapy, the trend of reduced survival in those of a high-risk zip code clearly highlights the multifactorial nature of disparities in esophageal cancer treatment. As discussed previously, it is feasible to consider that treatment center volume rather than center type may play a role, though examination of this remains challenging. Additionally, factors such as follow-up care, patient-specific patterns (smoking, alcohol consumption), and variation in compliance with adjuvant therapy may be at play. These factors are exceedingly difficult to examine in a retrospective database study and remain a limitation to a comprehensive understanding of the etiology of survival disadvantages in this group.

While identification of these trends is important, actionable intervention is essential to mitigate these disproportionate findings. Those at a socioeconomic disadvantage are also known to have additional burdens to accessing care, including transportation expenses, healthcare navigation skills, health literacy, social support, childcare, and more (4,22,23). Our finding of delayed time from diagnosis to systemic therapy and surgery in the high-risk zip code group underscores the reduced access to care seen in patients with lower SES. Fortunately, many of these barriers to care are intervenable with a multidisciplinary, team-based approach.

It is our hope that these data encourage the identification of high-risk patients with EAC and that those identified as residing in a high-risk zip code could be offered additional support to ensure equitable outcomes. When considering avenues for impactful change in this arena, we believe our data reveals two nadirs of equitable care—at diagnosis, as well as following resection. Further work is necessary to evaluate underlying obstacles to guideline-concordant therapy that exist at diagnosis, however improved care coordination for patient to minimize transportation and communication barriers when establishing complex neoadjuvant regimens may be beneficial. A similar focus throughout the neoadjuvant and postoperative period may serve to further mitigate the discrepancy in outcomes. We would like to offer the idea of multidisciplinary care coordination, not unlike that which currently exists in many thoracic oncology clinics. We propose a team comprised of the thoracic surgeon, medical oncologist, radiation oncologist, nutritionist, case manager, and social worker. Following diagnosis, the case manager would work to coordinate appointments so as to minimize number of visits to the clinic in the earliest stages of diagnosis. Additionally, an early emphasis on education and support group integration would be utilized at the first appointment rather than closer to resection. Resources such as parking and meal vouchers should be arranged and provided for both patient and family members at each appointment, as well as easy-to-access education resources and anticipated treatment regimen schedule to set reasonable expectations and target those patients most at risk for loss-to-follow-up. This model is in the early pilot stages at our institutions and we are eager to continue developing an approach best suited to our high-risk patients. We are hopeful that through a variety of targeted interventions such as these, we can work to mitigate the disparities we now know to exist.

Our study is not without limitations, the most notable of which are consistent with all large database endeavors. The NCDB lacks data regarding post-resection surveillance, disease recurrence, or disease-specific survival. These trends would be of particular use when discussing targeted interventions to improve survival in high-risk groups. Additionally, there is a lack of granularity regarding specific systemic treatment regimens as well as compliance with the prescribed therapy. We also acknowledge that utilizing zip code does not capture all SDOH experienced by individual patients. While these limitations are notable, we are hopeful that our population of nearly 11,000 patients with EAC in the post-CROSS era offers value as a nationally representative population with a variety of sociodemographic datapoints for evaluation as a surrogate for larger SES.

Conclusions

These data identify notable inequality in EAC clinical outcomes based solely on patient’s zip code of residence. Despite remarkably similar disease profiles at diagnosis, those living in a less educated or less affluent area have decreased odds of receiving guideline-concordant treatment as well as significantly worse median and OS. Moreover, for patients residing in a high-risk zip code that did receive guideline-concordant treatment, the reduction in survival persists compared to their low-risk counterparts. While the goal of our study was to quantify these anticipated trends, we are hopeful that this data will not only encourage the essential conversations surrounding these issues but also translate into actionable interventions that improve care for this at-risk group. In the era of multidisciplinary team-based care and rapid technological advancement, it is our responsibility and our charge to mitigate disparities.

Acknowledgments

This work has been presented on Southern Thoracic Surgical Association, Austin, TX (November 7-10, 2024) by Song Kim on behalf of the research team.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-637/rc

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-637/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-637/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Wasif N, Etzioni D, Habermann EB, et al. Racial and Socioeconomic Differences in the Use of High-Volume Commission on Cancer-Accredited Hospitals for Cancer Surgery in the United States. Ann Surg Oncol 2018;25:1116-25. [Crossref] [PubMed]
2. Morris AM, Rhoads KF, Stain SC, et al. Understanding racial disparities in cancer treatment and outcomes. J Am Coll Surg 2010;211:105-13. [Crossref] [PubMed]
3. Macdonald S, Macleod U, Campbell NC, et al. Systematic review of factors influencing patient and practitioner delay in diagnosis of upper gastrointestinal cancer. Br J Cancer 2006;94:1272-80. [Crossref] [PubMed]
4. Wang N, Cao F, Liu F, et al. The effect of socioeconomic status on health-care delay and treatment of esophageal cancer. J Transl Med 2015;13:241. [Crossref] [PubMed]
5. Liu CQ, Ma YL, Qin Q, et al. Epidemiology of esophageal cancer in 2020 and projections to 2030 and 2040. Thorac Cancer 2023;14:3-11. [Crossref] [PubMed]
6. Siegel RL, Miller KD, Fuchs HE, Jemal A. Cancer statistics, 2022. CA Cancer J Clin 2022; [Crossref]
7. Geng CX, Gudur AR, Radlinski M, et al. Socioeconomic Disparities Affect Outcomes in Early-Stage Esophageal Adenocarcinoma: A SEER Analysis. Clin Gastroenterol Hepatol 2023;21:2797-2806.e6. [Crossref] [PubMed]
8. Sridhar P, Bhatt M, Qureshi MM, et al. Esophageal Cancer Presentation, Treatment, and Outcomes Vary With Hospital Safety-Net Burden. Ann Thorac Surg 2019;107:1472-9. [Crossref] [PubMed]
9. Paulson EC, Ra J, Armstrong K, et al. Underuse of esophagectomy as treatment for resectable esophageal cancer. Arch Surg 2008;143:1198-203; discussion 1203. [Crossref] [PubMed]
10. Thein HH, Anyiwe K, Jembere N, et al. Effects of socioeconomic status on esophageal adenocarcinoma stage at diagnosis, receipt of treatment, and survival: A population-based cohort study. PLoS One 2017;12:e0186350. [Crossref] [PubMed]
11. Chen KA, Strassle PD, Meyers MO. Socioeconomic factors in timing of esophagectomy and association with outcomes. J Surg Oncol 2021;124:1014-21. [Crossref] [PubMed]
12. Eng OS, Nelson RA, Konstantinidis I, et al. Disparities in survival after trimodality therapy for esophageal adenocarcinoma. Dis Esophagus 2018;31: [Crossref] [PubMed]
13. Gupta A, Omeogu CH, Islam JY, et al. Association of area-level socioeconomic status and non-small cell lung cancer stage by race/ethnicity and health care-level factors: Analysis of the National Cancer Database. Cancer 2022;128:3099-108. [Crossref] [PubMed]
14. Boffa DJ, Rosen JE, Mallin K, et al. Using the National Cancer Database for Outcomes Research: A Review. JAMA Oncol 2017;3:1722-8. [Crossref] [PubMed]
15. Molena D, Mungo B, Stem M, et al. Does Quality of Care Matter? A Study of Adherence to National Comprehensive Cancer Network Guidelines for Patients with Locally Advanced Esophageal Cancer. J Gastrointest Surg 2015;19:1739-47. [Crossref] [PubMed]
16. Erhunmwunsee L, Gulack BC, Rushing C, et al. Socioeconomic Status, Not Race, Is Associated With Reduced Survival in Esophagectomy Patients. Ann Thorac Surg 2017;104:234-44. [Crossref] [PubMed]
17. Herd P, Goesling B, House JS. Socioeconomic position and health: the differential effects of education versus income on the onset versus progression of health problems. J Health Soc Behav 2007;48:223-38. [Crossref] [PubMed]
18. Donato BB, Campany ME, Brady JT, et al. Complete pathologic response in esophageal adenocarcinoma: does it make a difference? Dis Esophagus 2024;37:doae068. [Crossref] [PubMed]
19. Donato BB, Campany ME, Brady JT, et al. Lymphovascular Invasion is an Independent Negative Prognostic Factor in Esophageal Adenocarcinoma. Ann Surg Oncol 2024;31:6680-90. [Crossref] [PubMed]
20. Ju MR, Karalis JD, Leonard G, et al. Defining Volume Targets for Regionalization of Pancreaticoduodenectomy, Esophagectomy, and Major Lung Resection Centers to Improve Oncologic and Surgical Outcomes. Ann Surg Oncol 2024;31:499-513. [Crossref] [PubMed]
21. Rhodin KE, Raman V, Jensen CW, et al. The Effect of Center Esophagectomy Volume on Outcomes in Clinical Stage I to III Esophageal Cancer. Ann Surg 2023;278:79-86. [Crossref] [PubMed]
22. McMaughan DJ, Oloruntoba O, Smith ML. Socioeconomic Status and Access to Healthcare: Interrelated Drivers for Healthy Aging. Front Public Health 2020;8:231. [Crossref] [PubMed]
23. Lazar M, Davenport L. Barriers to Health Care Access for Low Income Families: A Review of Literature. J Community Health Nurs 2018;35:28-37. [Crossref] [PubMed]