Abstract
Background/Aim: Surgery is the cornerstone of treatment for papillary thyroid cancer (PTC), yet some patients refuse surgery, which may impact their survival outcomes. Understanding factors associated with surgery refusal could inform interventions to improve acceptance of recommended care. Patients and Methods: This retrospective cohort study evaluated the impact of surgery refusal on overall survival (OS) and identified associated demographic and clinical factors using data on patients with PTC from the National Cancer Database from 2004 to 2019. OS was analyzed using the Kaplan–Meier method, with survival curves compared using the log-rank test. Patient characteristics were analyzed using Pearson Chi-square tests or Wilcoxon Rank Sum tests. Results: Of the 201,051 patients with PTC who were advised to undergo surgery, 200,656 (99.8%) underwent surgery, while 395 (0.2%) refused. Patients who refused surgery were older (mean age 55.2 years vs. 48.7 years), more often male (27.8% vs. 22.8%), and represented higher proportions of Black, Asian, Hispanic, lower-income, uninsured, and non-privately insured patients (p<0.001). OS rates were significantly lower in patients who refused surgery, with one-year and five-year survival rates of 87% and 34%, respectively, compared to 96% and 56% for those who underwent surgery. Conclusion: Surgery refusal in patients with PTC was associated with poorer OS outcomes and was more frequent among older adults, socioeconomically disadvantaged populations, and racial and ethnic minorities. Interventions addressing patient concerns and barriers to surgery are critical to improving treatment acceptance and survival among these groups.
- Thyroid cancer
- papillary thyroid cancer (PTC)
- cancer disparities
- cancer survival
- refusal
- thyroid surgery
- thyroidectomy
Thyroid cancer is the most common endocrine malignancy, with an estimated 44,020 new cases and 2,170 deaths in the US projected for 2024 (1). Among these, papillary thyroid cancer (PTC) constitutes approximately 84% of all thyroid cancer cases and generally has a favorable prognosis, with a five-year relative survival rate exceeding 99% (1-3). Thyroid surgery, including total thyroidectomy or lobectomy, is the primary treatment for PTC and is typically curative (4). These surgical interventions are considered safe, with low long-term complication rates (5).
Despite the established safety and efficacy of thyroid surgery, a low proportion of patients with PTC refuse surgery despite clinical recommendations and the associated benefits. Refusal of recommended surgical treatment for thyroid cancer has been linked to poorer overall survival (OS) and disease-specific survival (DSS) (6). Research on surgical refusal in other cancers has identified concerns regarding safety, physical side effects, treatment efficacy, cost, impacts on quality of life, and prior negative family experiences as potential reasons for refusal (7, 8). Sociodemographic factors, including older age, uninsured status, Hispanic ethnicity, Black and Asian race, and lower income, are consistently associated with higher surgery refusal rates across multiple cancers (9-11). Disparities in refusal rates among disadvantaged populations can substantially impair their health outcomes (9, 10).
This study aimed to evaluate the impact of surgery refusal on OS in patients with PTC and identify the demographic and clinical factors associated with refusal. Understanding these disparities can inform strategies that address patient concerns, mitigate barriers, and promote surgery utilization, ultimately enhancing patient outcomes.
Patients and Methods
Design. This retrospective cohort study utilized data from the National Cancer Database (NCDB) from 2004 to 2019. The NCDB is a facility-based clinical surveillance registry developed through a collaboration between the American Cancer Society and the American College of Surgeons’ Commission on Cancer. It covers approximately 70% of new cancer cases in the US, collected from over 1,500 accredited facilities (12). This study did not require Institutional Review Board approval as the NCDB data is de-identified.
Study population. The NCDB was queried for patients diagnosed with PTC. Sociodemographic variables included age at diagnosis, race, ethnicity, sex, patient distance to the treatment facility, income, insurance status by primary insurance carrier at diagnosis, treatment facility type, and urban-rural classification. Patient distance to the treatment facility represented the mileage between each patient’s zip code and the reporting hospital. Income was defined as the median household income for each patient’s zip code based on the American Community Survey, adjusted for inflation. Treatment facility type followed the Commission on Cancer’s classification of reporting facilities by program structure, services provided, and caseload. Rural, urban, or metro classification was determined using rural-urban continuum codes from the United States Department of Agriculture Economic Research Service based on each patient’s county (13). Clinical characteristics included treatment type, stage, and grade. Staging followed the American Joint Committee on Cancer 6th and 7th edition guidelines.
Statistical analysis. Descriptive statistics summarized patient sociodemographic and clinical characteristics. Associations between categorical variables and surgery refusal were analyzed using Pearson Chi-square tests, represented as frequencies and relative frequencies. The Wilcoxon Rank Sum test was used for continuous variables, represented as medians, means, and standard deviations. OS was defined as the time from cancer diagnosis to death. Factors associated with OS were analyzed using both univariable and multivariable Cox proportional hazards models, as shown in Table I. Hazard ratios (HR), 95% confidence intervals (CI), and p-values were reported. OS was analyzed using the Kaplan–Meier (KM) method, with survival curves compared using the log-rank test. The numbers at risk were included in the KM curves. Statistical analysis was performed using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA).
Results
Patient characteristics. The study included 205,389 patients with PTC identified in the NCDB from 2004 to 2019. The demographics of these patients are shown in Table II. The mean age was 48.7 years, and 77.2% of patients were female. 87.1% of patients were White, 7.0% were Black, 0.8% were Asian or Pacific Islander, 0.3% were Native American, and 9.3% were Hispanic. 78.9% of patients had stage I disease. Of the 201,051 patients recommended for surgery, 99.8% underwent surgery, while 0.2% refused (Table II).
Factors associated with surgery refusal. Patients who refused surgery had a mean age of 55.2 years, significantly higher than the overall mean age of 48.7 years (p<0.001). Male patients constituted 22.8% of the overall cohort, compared to 27.8% among those who refused surgery (p<0.001). Refusal rates varied by race and ethnicity: Black patients represented 7.0% of the overall cohort but constituted 12.7% of refusals; Asian or Pacific Islander patients comprised 0.8% of the cohort but 1.6% of refusals; and Hispanic patients accounted for 9.3% of the cohort but 11.8% of refusals (p<0.001) (Table II).
Surgery refusal was associated with economic, geographic, and institutional factors, including lower income, closer proximity to treatment facilities, treatment at academic or research centers, and residence in metro or rural settings (p<0.001). Uninsured status was significantly linked to higher refusal rates, with 2.7% of patients uninsured compared to 6.4% among those who refused surgery (p<0.001). Patients covered by non-private health insurance, such as Medicaid and Medicare, also showed higher refusal rates. 24.7% of patients had non-private health insurance, compared to 43.3% of all surgery refusal patients (p<0.001). Clinical factors associated with surgery refusal included stage II cancer and prior chemotherapy (p<0.001) (Table II).
Overall survival. Patients who refused surgery demonstrated markedly shorter median survival times and lower one-year and five-year OS rates compared to patients who underwent surgery, as shown in Table III. The median survival time was 66.7 months for surgically treated patients versus 43.4 months for those who refused surgery (Table III). The log-rank test revealed statistically significant differences in OS between these groups (p<0.0001) (Figure 1). Patients who underwent surgery had one-year and five-year OS rates of 96% and 56%, respectively, while those who refused surgery had rates of 87% and 34% (Table III).
Discussion
This study assessed the impact of surgery refusal on OS in patients with PTC and identified factors associated with refusal. This analysis revealed that patients who refused surgery had lower one-year and five-year OS rates and shorter median survival times than patients who underwent surgery. Surgery refusal was associated with older age, male sex, Black and Asian race, Hispanic ethnicity, lower income, and uninsured or non-private insurance status.
Patients who refused surgery were significantly older, with a mean age of 55.2 years, which was 6.5 years older than the mean age of 48.7 years among all patients with PTC. These results align with existing research demonstrating higher treatment refusal rates among older adults (14). Previous studies have found that concerns about side effects are a significant factor influencing older adults’ decisions to refuse cancer treatment (7, 14). This is particularly relevant as older adults are at a higher risk of experiencing complications from thyroid surgery compared to younger patients (15, 16). A reduced desire to prolong life and concerns regarding safety, impacts on quality of life, and treatment efficacy also contribute to higher refusal rates among older adults (7, 8). Addressing these concerns is crucial to enhancing surgical care utilization and improving survival outcomes in this demographic.
A larger proportion of males declined surgery compared to females, a trend consistent with prior research on thyroid cancer (6). Sex disparities in surgical utilization have also been observed in other cancer types, such as lung and colon cancer, reflecting broader patterns in healthcare where males are generally less likely to seek medical care compared to females (9, 17).
This analysis also found higher surgery refusal rates among Black, Asian, and Hispanic patients. This finding is consistent across studies examining racial disparities in surgery refusal for thyroid cancer and other solid tumors, including breast, colorectal, lung, gynecologic, and prostate cancers (6, 9-11, 18). Studies have demonstrated higher cancer surgery refusal rates among Black and Asian patients, even when controlling for insurance status and socioeconomic factors (10, 11).
Addressing these disparities will likely require concerted efforts to bridge gaps in access, utilization, and outcomes through culturally competent communication and education. Prior studies indicate that shared decision-making utilizing decisional aids can significantly benefit racial and ethnic minority patients, increasing their medical knowledge and likelihood of undergoing surgery (19). Decisional aids tailored for PTC are currently being developed to aid patients in making informed choices about surgical intervention (20, 21).
Lower income and lack of insurance were strongly associated with higher surgery refusal rates. Out-of-pocket costs can create a significant financial barrier to surgical treatment for uninsured and low-income patients (22). Non-private insurance coverage, mainly Medicaid and Medicare, was also associated with surgery refusal. These findings underscore the urgent need for interventions aimed at improving affordability and expanding insurance coverage to reduce barriers to essential surgical care. Patient navigation programs, which assist patients in overcoming barriers to cancer care, have been shown to increase the likelihood of patients receiving recommended cancer treatment (23). Addressing financial and insurance barriers through policy changes and support programs can improve access to surgical care for lower-income and uninsured patients. Efforts to expand insurance coverage and reduce out-of-pocket costs for surgical procedures are critical to mitigating disparities in treatment acceptance.
Patients who refused surgery for PTC had lower median survival times and one-year and five-year OS rates compared to those who underwent surgical intervention. The impact of surgery refusal on OS rates highlights the urgency of addressing disparities in the acceptance of surgical treatment to ensure optimal outcomes for all patients. These results are consistent with prior literature demonstrating inferior survival outcomes for patients who refuse surgery for thyroid and other cancers (6, 9-11, 18).
This study contributes to the existing literature by specifically focusing on PTC, which presents distinct clinical features and management considerations compared to other types of thyroid cancer. By analyzing recent data from the NCDB, a database with extensive case coverage, this research provides updated insights into surgery refusal patterns and outcomes.
Future prospective research is needed to explore patients’ reasons for refusing surgery. Interventions such as patient navigator programs and decisional aids aimed at reducing surgery refusals and enhancing treatment utilization are necessary to address disparities and improve patient outcomes. By building upon existing knowledge and leveraging comprehensive database resources like the NCDB, further studies can advance our understanding of surgery refusal in PTC and inform strategies to mitigate these disparities with the ultimate goal to improve patient care.
This study included a robust sample size derived from an extensive, multicenter database covering approximately 70% of patients with cancer. However, we acknowledge that there are important limitations to consider. The NCDB lacks data on recurrence, cause of death, or specific reasons for surgery refusal, limiting deeper analysis into these aspects. Additionally, factors such as cultural and religious backgrounds, which could influence surgery refusal, are not captured in NCDB data. Patient comorbidities were not systematically assessed, although patients unable to undergo surgery due to comorbidities or frailty were analyzed separately from those who refused recommended surgery. The retrospective design of this study may have also affected its ability to reflect current treatment practices, including shifts towards active surveillance for PTC, and evolving patient attitudes toward surgery, potentially impacting the generalizability of findings to contemporary clinical settings.
Conclusion
This study found that patients with PTC who refused recommended surgery had worse OS rates compared to those who underwent surgical treatment. Surgery refusal was associated with older age, Black and Asian race, Hispanic ethnicity, male sex, lower income, uninsured status, and non-private insurance. These findings highlight significant disparities in surgery refusal, emphasizing the need for interventions to ensure equitable access to surgical treatment. Addressing these disparities is essential for improving survival outcomes and achieving better equity in the management of PTC.
Footnotes
Authors’ Contributions
JCB was responsible for writing the original draft. JCB and EMG were responsible for conceptualization, methodology, investigation, and visualization. EMG was responsible for validation, formal analysis, resources, data curation, supervision, and funding acquisition. JCB, BPS, SRR, SHR, RS, RP, KBS, GK, PBJ, FM, EPE, KP, SB, SS, HKM, and EMG were responsible for reviewing and editing.
Conflicts of Interest
The Authors declare no conflicts of interest in relation to this study.
Funding
This work was supported by Roswell Park Cancer Institute and National Cancer Institute (NCI) grant P30CA016056.
- Received October 3, 2024.
- Revision received October 16, 2024.
- Accepted October 17, 2024.
- Copyright © 2024 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).