Abstract
Background
Increasingly, patients with rectal cancer receive nonoperative management. A growing body of retrospective evidence supporting the safety of this approach has likely contributed to its growing popularity. However, patients may also undergo nonoperative management because of refusal of surgical resection. We hypothesize that patients who refuse surgery are more likely to be from groups who traditionally face barriers accessing care.
Methods
We used the National Cancer Database (2006–2017) to analyze patients with nonmetastatic rectal adenocarcinoma who underwent nonoperative management following radiation. We identified 2 groups: (1) planned nonoperative management and (2) nonoperative management because of refusal of surgery. We performed logistic regression to compare the groups along patient, socioeconomic, and facility-level factors.
Results
In total, 9,613 and 2,039 patients were included in the planned nonoperative management and refused nonoperative management groups, respectively. Of the total study cohort (ie, planned nonoperative management + refused nonoperative management), 21% of these patients diagnosed in 2017 underwent refused nonoperative management, versus 12% in 2006. Patients who were Black (adjusted odds ratio 1.47, 95% confidence interval 1.26–1.71) or Asian/Pacific Islander (adjusted odds ratio 1.51, 95% confidence interval 1.18–1.92), age ≥65 years (adjusted odds ratio 1.55, 95% confidence interval 1.37–1.77), with more advanced disease stage (stage III adjusted odds ratio 1.30, 95% confidence interval 1.10–1.53), and government insurance (adjusted odds ratio 1.19, 95% confidence interval 1.04–1.36) were associated with increased utilization of refused nonoperative management. Conversely, lower education (adjusted odds ratio 0.62, 95% confidence interval 0.50–0.76) and female sex (adjusted odds ratio 0.88, 95% confidence interval 0.79–0.97) were associated with planned nonoperative management.
Conclusion
Our findings suggest that the refused nonoperative management group is demographically distinct. Outreach efforts to better understand the rationale behind patient decision making in rectal cancer will be paramount to ensuring appropriate implementation of nonoperative management.
Introduction
Worldwide, there are approximately 700,000 new cases of rectal cancer and more than 300,000 deaths associated with rectal cancer annually.
1
,2
The current standard of care for nonmetastatic, locally advanced rectal cancer is to provide neoadjuvant therapy, followed by total mesorectal excision (TME) with either low anterior resection (LAR) or abdominoperineal resection (APR).3
However, an increasing body of evidence suggests that patients who experience a clinical complete response to neoadjuvant therapy may safely undergo nonoperative management (NOM), avoiding significant morbidities associated with surgery.National Comprehensive Cancer Network. Rectal cancer (version 2.2021).
https://www.nccn.org/professionals/physician_gls/pdf/rectal.pdf
Date accessed: November 26, 2021
4
, 5
, 6
, 7
, 8
A key concern is whether NOM is performed in clinically appropriate, protocolized settings.
9
, 10
, 11
Previously, data have suggested that socioeconomic factors commonly associated with disadvantaged backgrounds are associated with increased utilization of NOM, raising concerns that health disparities may be a major driving force behind the observed increase in NOM of rectal cancer. In particular, patients undergoing NOM may be refusing recommended surgery, independent of clinical appropriateness for a nonoperative approach. Indeed, prior studies have demonstrated that Black race, older age, male sex, and socioeconomic factors such as insurance status are associated with increased likelihood of refusing colon and rectal surgery.12
, 13
, 14
As NOM becomes an increasingly accepted approach for selected patient groups, including by the NCCN, it will likely be more widely adopted.
3
Therefore, it is critical to understand both patient and provider factors that are related to the decision to opt against surgical management. We set out to perform a descriptive analysis looking at patients who did not receive surgery after neoadjuvant treatment, grouped by the provided reason surgery was not performed. By examining and comparing patients based on the rationale for NOM, we aim to better understand the present real-world non–surgically managed population. Identifying patient, socioeconomic, and facility factors associated with specific reasons for nonsurgical management may provide insight into the nuanced and complex decision-making process underlying choice of treatment. We hope to provide context to inform future interventions that ensure equitable access and patient receptivity to recommended care, whether necessary surgery or protocolized NOM. We hypothesize that patients who traditionally experience barriers to care were more likely to have refused surgery than patients undergoing planned NOM. To our knowledge, this is the first article to directly compare patients with rectal cancer undergoing NOM based on the stated reason for no surgery.National Comprehensive Cancer Network. Rectal cancer (version 2.2021).
https://www.nccn.org/professionals/physician_gls/pdf/rectal.pdf
Date accessed: November 26, 2021
Methods
Data
The National Cancer Database (NCDB) is a national oncology outcomes database containing data from more than 1,500 Commission on Cancer (CoC)-accredited facilities, representing more than 70% of newly diagnosed cases nationwide. The NCDB is a joint project of the Commission on Cancer of the American College of Surgeons and the American Cancer Society. The NCDB includes patient, social, and facility-level data in addition to diagnosis, treatment, and clinical outcomes for patients with malignant diseases. These data are derived from a deidentified NCDB file. The American College of Surgeons and the 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. Because of the deidentified nature of these data, this study was exempt from Institutional Review Board approval.
Cohort
We identified all patients diagnosed with nonmetastatic rectal adenocarcinoma between 2006 and 2017 who received radiation therapy and did not undergo surgery of the primary disease site. We used the “reason for no surgery of primary site” variable, included in the NCDB, to define the study cohorts. This variable is coded based on whether the patient received surgery or not, and if the patient did not receive surgery is further defined based on the rationale for no surgery documented in the patient chart. Our study cohort consisted of 2 groups: (1) planned NOM (pNOM), in which surgery was “not part of the planned first course treatment”; and (2) refused NOM (rNOM), in which surgery was not performed because it was “refused by the patient.” We excluded patients <40 years of age as several facility-level variables are suppressed in these patients (eg, facility type) because of privacy concerns. In addition, we also excluded patients with unknown disease stage or unknown surgery status. Furthermore, we excluded patients who “[died] before planned or recommended surgery,” those in which “no reason was noted” for NOM, and patients who underwent NOM “due to patient risk factors.” We chose to exclude these patients and focus on the pNOM and rNOM groups as the decision to enter NOM is more likely to be informed by patient and provider preferences and decision making. Because the NCDB does not include information on response to neoadjuvant therapy, we are not able to limit our analysis to patients with a clinical complete response. A flow diagram is shown in Figure 1.
Figure 1Flow diagram.
Variables
Year of diagnosis was categorized into 3 groups over the 12-year study period. Patient comorbidity was represented by Charlson-Deyo score. Our cohort was limited to patients with stage 0–III disease. Age was dichotomized to 40–64 and ≥65 years. Race was categorized as White, Black, Asian/Pacific Islander, and other/unknown. Spanish or Hispanic origin was included separately, as this designation can apply to patients of any race. Insurance status was categorized as private, government, none, and unknown. Median income and education is categorized in the NCDB as quartiles based on the patient’s domicile. In addition, urbanity was determined by the population density of the patient’s domicile. Distance traveled was based on the distance from patient’s ZIP code centroid to reporting hospital, split into quartiles. Treatment facility type was categorized as academic, community, comprehensive community, and integrate network cancer program. No “unknown” facility type values were included as patients <40 years of age, for whom this variable is suppressed, were excluded from the analysis.
Statistical analysis
Categorical variables were compared using χ2 test. The threshold for inclusion in the logistic regression analysis was set at P < .20 a priori. Each variable meeting this threshold was included in a multivariable logistic regression analysis with backward stepwise selection to identify predictors of rNOM versus pNOM. Statistical tests were 2-sided with a significance level set at P < .05. All analyses were performed using RStudio (version 3.4.1, R Foundation, Vienna, Austria).
Results
A total of 11,652 patients was included in the analysis, consisting of 9,613 pNOM and 2,039 rNOM patients. The relative proportion of patients classified as rNOM rose over time, rising from 12% in 2006 to 21% in 2017, whereas the proportion of the pNOM patients decreased from 88% in 2006 to 79% in 2017. The composition of the study population over time is illustrated in Figure 2. The number of patients undergoing surgery listed in Figure 2 reflect patients who meet all other inclusion criteria but underwent surgery after neoadjuvant therapy. The characteristics of the pNOM and rNOM groups are summarized in Table I.
Figure 2Change in study cohort size and corresponding surgery cohort over study period.
Table IPatient demographics, socioeconomic factors, and hospital characteristics
| Total (%) | pNOM (%) | rNOM (%) | |
|---|---|---|---|
| All patients | 11,652 | 9,613 (82.5) | 2,039 (17.5) |
| Year diagnosed | |||
| 2006–2009 | 2,634 (22.6) | 2,261 (23.5) | 373 (18.3) |
| 2010–2013 | 3,700 (31.8) | 3089 (32.1) | 611 (30.0) |
| 2014–2017 | 5,318 (45.6) | 4,263 (44.3) | 1,055 (51.7) |
| Age | |||
| <65 | 4,555 (39.1) | 3,950 (41.1) | 605 (29.7) |
| ≥65 | 7,097 (60.9) | 5,663 (58.9) | 1,434 (70.3) |
| Sex | |||
| Male | 7,124 (61.1) | 5,846 (60.8) | 1,278 (62.7) |
| Female | 4,528 (38.9) | 3,767 (39.2) | 761 (37.3) |
| Race | |||
| White | 9,611 (82.5) | 7,978 (83.0) | 1,633(80.1) |
| Black | 1,334 (11.4) | 1,070 (11.1) | 264 (12.9) |
| Asian/Pacific Islander | 429 (3.7) | 335 (3.5) | 94 (4.6) |
| Other/unknown | 278 (2.4) | 230 (2.4) | 48 (2.4) |
| Hispanic origin | |||
| No | 10,455 (89.7) | 8,606 (89.5) | 1,849 (90.7) |
| Yes | 745 (6.4) | 632 (6.6) | 113 (5.5) |
| Unknown | 452 (3.9) | 375 (3.9) | 77 (3.8) |
| Insurance | |||
| Private | 3,410 (29.3) | 2,931 (30.5) | 479 (23.5) |
| Government | 7,520 (64.5) | 6,059 (63.0) | 1,461 (71.7) |
| None | 532 (4.6) | 462 (4.8) | 70 (3.4) |
| Unknown | 190 (1.6) | 161 (1.7) | 29 (1.4) |
| Median income | |||
| Fourth quartile (highest) | 3,273 (28.1) | 2,686 (27.9) | 587 (28.8) |
| Third quartile | 2,723 (23.4) | 2,209 (23.0) | 514 (25.2) |
| Second quartile | 2,577 (22.1) | 2,165 (22.5) | 412 (20.2) |
| First quartile (lowest) | 2,251 (19.3) | 1,917 (19.9) | 334 (16.4) |
| Unknown | 828 (7.1) | 636 (6.6) | 192 (9.4) |
| High school diploma | |||
| Fourth quartile (highest) | 2,182 (18.7) | 1,753 (18.2) | 429 (21.0) |
| Third quartile | 3,172 (27.2) | 2,596 (27.0) | 576 (28.2) |
| Second quartile | 3,019 (25.9) | 2511 (26.1) | 508 (24.9) |
| First quartile (lowest) | 2,473 (21.2) | 2,129 (22.1) | 344 (16.9) |
| Unknown | 806 (6.9) | 624 (6.5) | 182 (8.9) |
| Metro/urban/rural | |||
| Metro | 9,367 (80.4) | 7,717 (80.3) | 1,650 (80.9) |
| Urban | 1,719 (14.8) | 1,441 (15.0) | 278 (13.6) |
| Rural | 225 (1.9) | 182 (1.9) | 43 (2.1) |
| Unknown | 341 (2.9) | 273 (2.8) | 68 (3.3) |
| Distance traveled (miles) | |||
| ≤4.0 | 2,790 (23.9) | 2,262 (23.5) | 528 (25.9) |
| 4.1–8.8 | 2,677 (23.0) | 2,220 (23.1) | 457 (22.4) |
| 8.9–20.2 | 2,713 (23.3) | 2,256 (23.5) | 457 (22.4) |
| >20.2 | 2,721 (23.4) | 2,296 (23.9) | 425 (20.8) |
| Unknown | 751 (6.4) | 579 (6.0) | 172 (8.4) |
| Charlson/Deyo score | |||
| 0 | 9,002 (77.3) | 7,454 (77.5) | 1,548 (75.9) |
| 1 | 1,838 (15.8) | 1,506 (15.7) | 332 (16.3) |
| 2 | 514 (4.4) | 415 (4.3) | 99 (4.9) |
| ≥3 | 298 (2.6) | 238 (2.5) | 60 (2.9) |
| Clinical stage | |||
| I | 1,533 (13.2) | 1,311 (13.6) | 222 (10.9) |
| II | 4,900 (42.1) | 4,000 (41.6) | 900 (44.1) |
| III | 5,137 (44.1) | 4,226 (44.0) | 911 (44.7) |
| 0 | 82 (0.7) | 76 (0.8) | 6 (0.3) |
| Facility type | |||
| Academic | 4,089 (35.1) | 3,415 (35.5) | 674 (33.1) |
| Community | 1,497 (12.8) | 1,284 (13.4) | 213 (10.4) |
| Comprehensive community | 4,674 (40.1) | 3,794 (39.5) | 880 (43.2) |
| Integrated network | 1,392 (11.9) | 1,120 (11.7) | 272 (13.3) |
pNOM, planned nonoperative management; rNOM, refused nonoperative management.
∗ Percentage of total row; all other percentages are within column.
Comorbidity score, Spanish or Hispanic origin, and metro/urban/rural status failed to meet the a priori threshold of P < .20 in the χ2 analysis. Table II shows both univariable and multivariable logistic regression analyses. The remaining variables were assessed via multivariable logistic regression analysis to identify predictors of rNOM versus pNOM. Distance traveled was excluded from the final multivariable logistic regression model by backward stepwise selection.
Table IIUnivariable and multivariable regression analysis to identify factors associated with refusal of surgery
| Univariable | Multivariable | |||
|---|---|---|---|---|
| OR (95% CI) | P value | OR (95% CI) | P value | |
| Year diagnosed | ||||
| 2006–2009 | Ref. | Ref. | Ref. | Ref. |
| 2010–2013 | 1.20 (1.04–1.38) | .011 | 1.19 (1.03–1.37) | .017 |
| 2014–2017 | 1.50 (1.32–1.71) | <.001 | 1.46 (1.28–1.67) | <.001 |
| Age | ||||
| <65 | Ref. | Ref. | Ref. | Ref. |
| ≥65 | 1.65 (1.49–1.83) | <.001 | 1.55 (1.37–1.77) | <.001 |
| Sex | ||||
| Male | Ref. | Ref. | Ref. | Ref. |
| Female | 0.92 (0.84–1.02) | .117 | 0.88 (0.79–0.97) | .012 |
| Race | ||||
| White | Ref. | Ref. | Ref. | Ref. |
| Black | 1.21 (1.04–1.39) | .011 | 1.47 (1.26–1.71) | <.001 |
| Asian/Pacific Islander | 1.37 (1.08–1.73) | .008 | 1.51 (1.18–1.92) | .001 |
| Other/unknown | 1.02 (0.74–1.38) | .904 | 1.16 (0.84–1.59) | .356 |
| Insurance | ||||
| Private | Ref. | Ref. | Ref. | Ref. |
| Government | 1.48 (1.32–1.65) | <.001 | 1.19 (1.04–1.36) | .011 |
| None | 0.93 (0.70–1.21) | .582 | 1.05 (0.79–1.38) | .718 |
| Unknown | 1.10 (0.72–1.63) | .639 | 1.04 (0.68–1.55) | .848 |
| Median income | ||||
| Fourth quartile (highest) | Ref. | Ref. | Ref. | Ref. |
| Third quartile | 1.06 (0.93–1.21) | .348 | 1.18 (1.02–1.37) | .026 |
| Second quartile | 0.87 (0.76–1.00) | .050 | 1.04 (0.88–1.23) | .658 |
| First quartile (lowest) | 0.80 (0.69–0.92) | .002 | 1.03 (0.84–1.25) | .796 |
| Unknown | 1.38 (1.15–1.66) | .001 | 4.79 (1.96–11.5) | <.001 |
| High school diploma | ||||
| Fourth quartile (highest) | Ref. | Ref. | Ref. | Ref. |
| Third quartile | 0.91 (0.79–1.04) | .167 | 0.89 (0.76–1.03) | .119 |
| Second quartile | 0.83 (0.72–0.95) | .009 | 0.81 (0.68–0.97) | .020 |
| First quartile (lowest) | 0.66 (0.57–0.77) | <.001 | 0.62 (0.50–0.76) | <.001 |
| Unknown | 1.19 (0.98–1.45) | .079 | 0.24 (0.10–0.60) | .002 |
| Distance traveled (miles) | ||||
| ≤4.0 | Ref. | Ref. | - | - |
| 4.1–8.8 | 0.88 (0.77–1.01) | .075 | - | - |
| 8.9–20.2 | 0.87 (0.76–1.00) | .044 | - | - |
| >20.2 | 0.79 (0.69–0.91) | .001 | - | - |
| Unknown | 1.27 (1.05–1.54) | .015 | - | - |
| Clinical stage | ||||
| I | Ref. | Ref. | Ref. | Ref. |
| II | 1.33 (1.13–1.56) | <.001 | 1.31 (1.12–1.55) | .001 |
| III | 1.27 (1.09–1.50) | .003 | 1.30 (1.10–1.53) | .002 |
| 0 | 0.47 (0.18–1.00) | .076 | 0.48 (0.18–1.03) | .087 |
| Facility type | ||||
| Academic | Ref. | Ref. | Ref. | Ref. |
| Community | 0.84 (0.71–0.99) | .041 | 0.86 (0.72–1.01) | .073 |
| Comprehensive community | 1.18 (1.05–1.31) | .004 | 1.17 (1.05–1.31) | .007 |
| Integrated network | 1.23 (1.05–1.44) | .009 | 1.16 (0.99–1.36) | .071 |
Variables removed by χ2 test: Charlson/Deyo Score, metro/urban/rural, Hispanic origin.
Variables removed from final multivariable model by backward stepwise selection: distance traveled.
Patient variables independently associated with rNOM in the final multivariable regression analysis include Black race (adjusted OR [aOR] 1.47, 95% CI 1.26–1.71) and Asian/Pacific Islander race (aOR 1.51, 95% CI 1.18–1.92), age ≥65 (aOR 1.55, 95% CI 1.37–1.77), more advanced disease stage (stage III aOR 1.30, 95% CI 1.10–1.53), and government insurance (aOR 1.19, 95% CI 1.04–1.36). Diagnosis in 2014–2017 (aOR 1.46, 95% CI 1.28–1.67) and 2010–2013 (aOR 1.19, 95% CI 1.03–1.37) were both also associated with rNOM. Compared to treatment at an academic cancer center, treatment at a comprehensive community cancer program was associated with rNOM (aOR 1.17, 95% CI 1.05–1.31). Increasingly lower levels of educational attainment (lowest quartile of HS diploma attainment: aOR 0.62, 95% CI 0.50–0.76) and female sex (aOR 0.88, 95% CI 0.79–0.97) were associated with decreased likelihood of rNOM.
Discussion
The number of patients with rectal cancer undergoing NOM increased during the study period, consistent with other studies.
9
,11
The proportion of patients who refused recommended surgery increased disproportionately over this period, growing from 12% to 21% of the combined total from 2006 to 2017. In multivariable analysis, we found that Black and Asian/Pacific Islander race, age ≥65, male sex, more advanced clinical stages, government insurance, and treatment at comprehensive community cancer programs were independently associated with rNOM, whereas lower educational attainment was associated with pNOM.We showed that Black and Asian/Pacific Islander patients are disproportionately likely to have refused recommended surgery rather than to have undergone planned NOM. Potential explanations include ineffective patient–provider communication for non-White patients, less receptivity to surgery or tolerance for associated morbidities among non-White patients, or, more likely, a combination of these and other considerations. Data suggest that racially discordant physician–patient interactions are associated with significantly less information exchange and active participation, supporting that discordant race can contribute to ineffective communication.
15
Indeed, many have described the association of non-White race with both pNOM and rNOM compared to patients undergoing surgery.9
,11
,13
,14
That race remains a powerful predictor between rNOM and pNOM supports our hypothesis that cultural identity and norms, for which race is an admittedly inexact proxy, may play a role in surgical hesitancy and patient–provider misalignment. However, these data do not include information regarding racial identity of the provider, so we are unable to conclusively point to discordant patient–provider race as a factor.In addition, age ≥65 years has also been associated with both rNOM and pNOM versus surgery.
9
,11
,13
,14
We found that age is also a significant predictor of rNOM versus pNOM. This is consistent with other data indicating older age is associated with increased rates of refusal of guideline-concordant cancer care across treatment modalities.16
Older patients may be especially resistant to surgery as these patients may be particularly focused on preserving function, independence, and quality of life rather than pursuing highest probability of cure. Given the nature of rectal cancer surgery complications (eg, genitourinary complications, bowel issues), older patients intent on maintaining independent function and/or lacking access to robust support may forego recommended surgery to avoid these potentially permanent sequalae.Male sex was also independently associated with rNOM versus pNOM, concordant with some prior studies indicating male patients may be more likely to refuse rectal surgery.
13
,14
The NCDB database does not provide sufficient patient-level details for an appropriately nuanced assessment of how sex may influence patient decision making. However, a speculative explanation we considered may be differences in perceived quality of life effects that vary by sex. For example, though male and female patients appear to experience similar rates of postoperative sexual dysfunction, male patients may be more likely to receive pretreatment counseling on risk of sexual dysfunction. An interpretation of these data could be that female patients may not be provided similar counseling and information regarding risk of sexual dysfunction.17
, 18
, 19
Additional factors that may contribute to the observed differences by sex to consider include the potential effect of sex on physician recommendations, possible differences in treatment adherence, and more.20
,21
We must consider that sex may be a confounder for a variable not included in this analysis. Because of limitations of the database, this analysis is not equipped to fully address these potential explanations for the observed results. Nevertheless, given these findings, we contend that targeted and culturally literate interventions may have substantial clinical benefit in rectal cancer care by strengthening patient–provider channels of communication.Our final multivariable model also indicated that government insurance (Medicare, Medicaid, and other) is independently associated with rNOM. Existing literature demonstrates that nonprivate insurance is associated with refusal compared to surgery.
12
, 13
, 14
Similarly, we found that government insurance was associated with rNOM, suggesting that differences in patient out-of-pocket expenditure may create access barriers that influence patients’ refusal of recommended surgery.Interestingly, our model found a trend toward increasing likelihood of pNOM in ZIP codes of lower educational attainment, with the difference between the most educated quartile and the 2 least educated quartiles reaching statistical significance. However, that this relationship was not significant across all quartiles indicates interpretation of this result be made with caution. Existing literature has demonstrated that surgery refusal is associated with lower educational attainment versus surgery in colon and rectal cancer.
12
Conversely, our analysis suggests that pNOM patients are disproportionately drawn from less educated communities compared to the rNOM group. Of great concern is if patients of lower perceived health literacy are not offered curative surgery at the same rate, resulting in disproportionate representation among the pNOM group.The increase in rNOM over the study period is concerning as patients may be refusing potentially curative interventions, borne out by decreased survival among rNOM versus surgery patients.
12
, 13
, 14
While beyond the scope of this article to assess, the increase in rNOM over the study period may reflect interplay between new treatment strategies for rectal cancer and perception of NOM from patients and surgeons. Alternatively, another possible explanation for this increase in rNOM may include the documented decline of trust in physicians and health care institutions.22
Cultural and personal values and needs affect any decision-making process. As such, patient-centered care has emerged as an essential tenet of high-quality care.
23
Therefore, we must also consider that the increasing trend in refusal of surgery may reflect a growing emphasis on patient-centered decision making; perhaps Black, Asian/Pacific Islander, and elderly patients are more likely to believe surgery represents an unacceptable risk/benefit tradeoff. In either case, future studies will be needed to clarify the underlying reasons for this trend, likely using a mixed-methods approach to capture the level of granularity and nuance involved in these complex decisions.Our study is limited by its retrospective nature, which can limit the variables available for analysis. For rectal cancer, the NCDB does not include a variable for response to neoadjuvant therapy. While we limited our analysis to nonmetastatic patients, we are unable to include solely patients with a clinical complete response. In addition, we are unable to explicitly determine the specific neoadjuvant radiation regimen (eg, long-course versus short-course), the tumor location within the rectum, treatment-related complications, or occurrence of metastatic disease during treatment. The NCDB also does not provide insight on provider experience with the disease of interest and surgeon preference or bias toward specific treatments. In addition, the NCDB lacks granular data that contributes to the complex decision-making process in rectal cancer treatment, especially certain sociodemographic, economic, and psychosocial aspects of care. Another key limitation of our analysis is variability in how the “reason for no surgery” variable is coded, as it relies on a partly subjective interpretation by physicians and abstractors. Lastly, the NCDB only contains patients treated at CoC-accredited centers, which may limit generalizability.
In conclusion, our findings indicate that the rNOM group is demographically distinct from the pNOM group along age, sex, race, and educational attainment. Given the distinct demographic characteristics of the rNOM and pNOM groups, developing and widely deploying culturally competent resources to help patients and physicians navigate treatment options is paramount to ensuring appropriate implementation of NOM.
Funding/Support
None declared.
Conflict of interest/Disclosure
The authors have no conflicts of interest to declare.
References
- Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries.CA Cancer J Clin. 2018; 68: 394-424
- Epidemiology of colorectal cancer: incidence, mortality, survival, and risk factors.Prz Gastroenterol. 2019; 14: 89-103
- National Comprehensive Cancer Network. Rectal cancer (version 2.2021).https://www.nccn.org/professionals/physician_gls/pdf/rectal.pdfDate accessed: November 26, 2021
- A watch-and-wait approach for locally advanced rectal cancer after a clinical complete response following neoadjuvant chemoradiation: a systematic review and meta-analysis.Lancet Gastroenterol Hepatol. 2017; 2: 501-513
- Operative versus nonoperative treatment for stage 0 distal rectal cancer following chemoradiation therapy: long-term results.Ann Surg. 2004; 240: 711-717
- Assessment of a watch-and-wait strategy for rectal cancer in patients with a complete response after neoadjuvant therapy.JAMA Oncol. 2019; 5: e185896
- Long-term outcome in patients with a pathological complete response after chemoradiation for rectal cancer: a pooled analysis of individual patient data.Lancet Oncol. 2010; 11: 835-844
- Cost-effectiveness and quality-adjusted survival of watch and wait after complete response to chemoradiotherapy for rectal cancer.J Natl Cancer Inst. 2020; 112: 792-801
- National trends in nonoperative management of rectal adenocarcinoma.J Clin Oncol. 2016; 34: 1644-1651
- No surgery after chemoradiation is not equal to nonoperative management after complete clinical response and chemoradiation.J Clin Oncol. 2016; 34: 4051
- Recent trends and overall survival of young versus older adults with stage ii to iii rectal cancer treated with and without surgery in the United States, 2010–2015.Am J Clin Oncol. 2020; 43: 694-700
- Sociodemographic predictors of surgery refusal in patients with stage I–III colon cancer.J Surg Oncol. 2020; 121: 1306-1313
- Treatment of stage I–III rectal cancer: who is refusing surgery?.J Surg Oncol. 2020; 121: 990-1000
- Refusal of colorectal cancer surgery in the United States: predictors and associated cancer-specific mortality in a Surveillance, Epidemiology, and End Results (SEER) cohort.Surg Open Sci. 2020; 2: 12-18
- Racial differences in doctors’ information-giving and patients’ participation.Cancer. 2006; 107: 1313-1320
- Reasons for underuse of recommended therapies for colorectal and lung cancer in the Veterans Health Administration.Cancer. 2012; 118: 3345-3355
- Complications after rectal cancer surgery.in: Chang G. Rectal Cancer. Springer, Cham, Switzerland2018: 335-354
- Prevalence of male and female sexual dysfunction is high following surgery for rectal cancer.Ann Surg. 2005; 242: 212-223
- Long term sexual function following rectal cancer treatment.Am J Surg. 2020; 220: 1258-1263
- Adherence to treatment guidelines and survival for older patients with stage II or III colon cancer in Texas from 2001 through 2011.Cancer. 2018; 124: 679-687
- Association of sex with toxic effects, treatment adherence, and oncologic outcomes in the CAO/ARO/AIO-94 and CAO/ARO/AIO-04 phase 3 randomized clinical trials of rectal cancer.JAMA Oncol. 2020; 6: 294-296
- A trust initiative in health care: why and why now?.Acad Med. 2019; 94: 463-465
- When does nonadherence indicate a deviation from patient-centered care?.Am J Manag Care. 2021; 27: e141-e144
Article Info
Publication History
Published online: August 28, 2022
Accepted:
May 30,
2022
Footnotes
Abstract titled “Factors Associated With Refusing Surgery Versus Planned Nonoperative Management for Rectal Cancer” was presented virtually as an oral presentation at the 17th Annual Academic Surgical Congress, February 2, 2022 (Orlando, FL).
Kurt Pianka and Beiqun Zhao made equal contributions to this article.
Identification
Copyright
© 2022 The Author(s). Published by Elsevier Inc.
User License
Creative Commons Attribution (CC BY 4.0) | How you can reuse 
Elsevier's open access license policy
Creative Commons Attribution (CC BY 4.0)
Permitted
- Read, print & download
- Redistribute or republish the final article
- Text & data mine
- Translate the article
- Reuse portions or extracts from the article in other works
- Sell or re-use for commercial purposes
Elsevier's open access license policy
00421-4&id=gr1.jpg){kind=link}
00421-4&id=gr2.jpg){kind=link}