Abstract
Background & Aims
Previous studies reported that black vs white disparities in survival among elderly patients with colorectal cancer (CRC) were due to differences in tumor characteristics (tumor stage, grade, nodal status, and comorbidity) rather than differences in treatment. We sought to determine the sequential contribution of differences in insurance, comorbidity, tumor characteristics, and treatment receipt to the black-white survival disparity among patients with CRC in 18–64 years old.
Methods
We used data from the National Cancer Database, a hospital-based cancer registry database sponsored by the American College of Surgeons and the American Cancer Society, on non-Hispanic black (black) and non-Hispanic white (white) patients, 18–64 years old, diagnosed from 2004 through 2012 with single or first primary invasive stage I–IV CRC. Blacks were sequentially matched by demographics, insurance, comorbidity, tumor characteristics, and treatment with 5 white partially overlapping subgroups using propensity score and greedy matching algorithm. We used the Kaplan-Meier method to estimate 5-year survival, and Cox proportional hazards models to generate hazard ratios (HR).
Results
The absolute 5-year survival difference between black and white unmatched patients with CRC was 9.2% (57.3% for black patients vs 66.5% for white patients; P < .0001). The absolute difference in survival did not change after patient groups were matched for demographics, but decreased to 4.9% (47% relative decrease [4.3% of 9.2%]) when they were matched for insurance and to 2.3% when they were matched for tumor characteristics (26% relative decrease [2.4% of 9.2%]). Further matching by treatment did not reduce the difference in 5-year survival between black and white patients. In proportional hazards model, insurance and tumor characteristics matching accounted for the 54% and 27% excess risk of death in black patients, respectively.
Conclusions
In an analysis of data from the National Cancer Database, we found that insurance coverage differences accounted for approximately one-half of the disparity in survival rate between black vs white patients with CRC in 18–64 years old; tumor characteristics accounted for a quarter of the disparity. Affordable health insurance coverage for all populations could substantially reduce differences in survival times of black vs white patients with CRC.
Keywords: colon cancer, mortality, race, comorbidity
Introduction
Colorectal cancer (CRC) is the third most commonly diagnosed cancer in both men and women in the United Sates.1 Although overall CRC incidence and mortality rates are decreasing in the United States,2 rates are increasing in the younger population.3 Notwithstanding these patterns, CRC incidence and mortality rates continue to be higher in blacks than in whites.4, 5 Multiple factors have been proposed as potential contributors to CRC racial survival disparity, including differences in age composition,6–11 stage,11–15 grade,6, 15, 16 tumor location,8, 9, 17 comorbidities,15, 18 provider characteristics,19 socioeconomic factors,16, 18, 20–22 and treatment receipt.11, 18, 23
Two recent studies reported that black-white differences in receipt of treatment contributed to only 0.1%–0.6% of the 8.3%–9.9% of the absolute black-white survival disparity among patients with colon cancer aged 65 years and older,12, 15 but these patients have relatively uniform health insurance coverage through Medicare. In this paper, we focus on the impact of access to care on black-white survival disparity by extending these analyses to black and white patients with CRC aged 18–64 years old, who have less uniform insurance status. Specifically, we estimate the contribution of difference in insurance coverage to the racial overall survival disparity, as well as the sequential contributions of differences in tumor characteristics and treatment receipt to the survival disparity.
Methods
Data Source and Patient Cohorts
We used data from the National Cancer Data Base (NCDB). The NCDB is a national hospital-based cancer registry database jointly sponsored by the American College of Surgeons and the American Cancer Society, which captures approximately 70% of newly diagnosed cancer cases in the U.S.24 We included patients diagnosed between 2004–2012 with single or first primary, invasive cancer at the primary site of colon and rectum (C18.0, C18.2–C18.9, C19.9, C20.9), treated at the reporting facility, histology codes (International Classification of Diseases for Oncology, 3rd edition [ICD-O-3]) for colorectal cancer (Supplementary Table 1), American Joint Committee on Cancer clinical stage I-IV, race/ethnicity (NH whites [whites], NH blacks [blacks]), aged 18–64 years, and no duplicate data from the NCDB. Patients with missing/unknown values for nodal status, insurance (additionally excluded government insurance other than Medicaid/Medicare), chemotherapy, radiotherapy, or surgery were excluded (n = 13,245) (Figure 1). Variables were coded according to the Facility Oncology Registry Data Standards (FORDS) Manual revised for 2012.25
Figure 1.
Flow chart of cohort inclusion/exclusion criteria for 18–64 years old patients with stage I–IV colorectal cancer in the National Cancer Data Base (2004–2012).
We sequentially added variables to generate a propensity score for each patient; here, the score predicts the probability that the patient is black, accounting for differences in covariates. The propensity scores were then used as a single scalar variable to match black and white patients using Greedy matching algorithm (Supplementary Material),15, 26 as a method to adjust or account for the variables used in calculating the score.27 We added variables as follows: we first used (1) demographics (age, sex, diagnosis year, and US census division) [demographic match]; then used (2) demographics, plus insurance status [insurance match]; then used (3) demographics, insurance status, plus comorbidity score [comorbidity match]; then used (4) demographics, insurance status, comorbidy score, plus tumor characteristics (tumor stage, grade, tumor location, margin status, and nodal status) [tumor characteristics match]; and finally used (5) demographics, insurance status, comorbidy score, and tumor characteristics, plus treatment (receipt of surgery, surgery of distant sites, chemotherapy, or radiotherapy) [treatment match]. We compared results from these sequential analyses to estimate the contribution of each factor or set of factors to the overall survival disparity between black and white CRC patients. In the propensity score, we included insurance status immediately after demographics because insurance affects survival through early stage at diagnosis and receipt of treatment, as well as through the number and severity of comorbidities.
Outcomes of Interest
Our main outcome of interest was overall 5-year survival probability. Follow-up time for calculating survival probability was from date of diagnosis until end-of-study date (December 31, 2013), last contact date, or death, whichever occurred first.
Independent Variables of Interest
Race/ethnicity was our primary independent variable of interest and categorized as NH white (white) and NH black (black) based on information reported by the patient and captured in the medical records. Further, we categorized whites into five partially overlapping subgroups by sequentially matching whites with blacks using propensity scores that were calculated after accounting for demographic characteristics, insurance status, comorbidity score, tumor characteristics, and treatment receipt. Other independent variables of interest include receipt of surgery, chemotherapy, and radiation therapy, and each was categorized into “no” or “yes” per receipt of treatment information.
Control Variables
Variables used in the demographic match were categorized as follows: age (18–39, 40–49, 50–64), gender (male, female), diagnosis year (2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012), US census division (New England, Middle Atlantic, South Atlantic, East North Central, East South Central, West North Central, West South Central, Mountain, Pacific) defined by location of reporting facility. Insurance status was categorized as uninsured, Medicaid, Medicare, or private. Comorbidity score was categorized as a score of 0, 1, or ≥ 2 based on the sum of weighted Charlson-Deyo Score.28 Tumor characteristics included were: grade (well differentiated, moderately differentiated, poorly differentiated, undifferentiated, missing), stage (I, II, III, IV), tumor location (right colon, transverse colon, left colon, unspecified colon, rectum), node status (positive, negative, not examined), and surgical margin (negative, positive, unknown/other). Treatment variables were categorized as: surgery (no, yes), distant site surgery (none, other site, distant site, missing/unknown), radiation (no, yes), and chemotherapy (no, yes). Facility case volume was categorized into tertiles of low, medium, or high based on the rank of each facility’s total number of CRC cases treated during the study period.
Statistical Analysis
We conducted descriptive analyses for characteristics of black and five sequentially matched subgroups of white CRC patients in the National Cancer Data Base. The subgroups of whites partially overlap with one another. We also calculated 5-year overall survival probabilities for blacks and five sequentially matched partially overlapping subgroups of whites using Kaplan-Meier method. We complemented these survival analyses with Cox proportional hazards models to generate hazard ratios (HR) of 5-year all-cause mortality for blacks compared with the entire population of whites and with each of the five partially overlapping subgroups of whites that were sequentially matched with blacks by propensity scores. In addition, we performed similar analyses stratified by stage and tumor location.
We conducted a sensitivity analysis to determine black-white 5-year survival differences for the treatment-matched subgroup when stratified by facility case volume. We also performed a sensitivity analysis to estimate the black-white 5-year survival differences by restricting anslysis to patients diagnosed between 2004–2008. Furthermore, we conducted a supplementary analysis to estimate the individual contribution of each of the five sets of factors to the survival disparity without sequential matching. Our study received exempt status from the Institutional Review Board of the Morehouse School of Medicine, Atlanta, GA. Statistical significance was determined based on the two-sided P value < 0.05. All statistical analyses were conducted using SAS software (version 9.4; SAS Institute, Cary, NC).
Results
There were 199,098 CRC patients aged 18 to 64 years old (16.7% NH blacks, 83.3% NH whites) in the study cohort. Black CRC patients were more likely to be younger (aged < 50 years, 28.1% vs 26.2%), female (49.7% versus 42.9%), and reside in the South Atlantic division (38.9% vs 22.4%) compared with white CRC patients (Table 1). Blacks were also more likely to present with stage IV disease (27.6% vs 22.6%), right-sided colon cancer (33.3% vs 24.1%), and comorbidity score ≥ 2 (7.3% vs 4.9%). Demographic matching did not change black-white differences in tumor characteristics, but further matching by insurance status reduced the black/white difference in proportion of metastatic CRC by 2.2% in absolute terms and by 41.5% in relative terms.
Table 1.
Descriptive characteristics of patients aged 18–64 years with stage I–IV colorectal cancer for NH blacks matched with NH whites by demographic, insurance, comorbidity, tumor characteristic, and treatment factors sequentially
| Variable | Category | NH black Patients, N (%) |
NH White Patients, N (%)
|
|||||
|---|---|---|---|---|---|---|---|---|
| All NH whites (Unmatched) |
Demographic- Matched |
Insurance- Matched |
Comorbidity- Match |
Tumor Characteristics- Matched |
Treatment- Matched |
|||
| N=33299 | N=165799 | N=33299 | N=33299 | N=33298 | N=33169 | N=33151 | ||
| N (col %) | N (col %) | N (col %) | N (col %) | N (col %) | N (col %) | N (col %) | ||
| Diagnosis age group, year | ||||||||
| 18–39 | 2170 (6.5) | 10232 (6.2) | 2157 (6.5) | 2170 (6.5) | 2123 (6.4) | 2258 (6.8) | 2243 (6.8) | |
| 40–49 | 7185 (21.6) | 33238 (20.0) | 7254 (21.8) | 7185 (21.6) | 7173 (21.5) | 7127 (21.5) | 7025 (21.2) | |
| 50–64 | 23944 (71.9) | 122329 (73.8) | 23888 (71.7) | 23944 (71.9) | 24002 (72.1) | 23784 (71.7) | 23883 (72.0) | |
| Sex | ||||||||
| Male | 16733 (50.3) | 94591 (57.1) | 16789 (50.4) | 16733 (50.3) | 16725 (50.2) | 16612 (50.1) | 16647 (50.2) | |
| Female | 16566 (49.7) | 71208 (42.9) | 16510 (49.6) | 16566 (49.7) | 16573 (49.8) | 16557 (49.9) | 16504 (49.8) | |
| Diagnosis year | ||||||||
| 2004 | 3122 (9.4) | 16721 (10.1) | 3122 (9.4) | 3233 (9.7) | 3173 (9.5) | 3231 (9.7) | 3217 (9.7) | |
| 2005 | 3185 (9.6) | 17076 (10.3) | 3185 (9.6) | 3333 (10.0) | 3300 (9.9) | 3327 (10.0) | 3272 (9.9) | |
| 2006 | 3480 (10.5) | 17388 (10.5) | 3480 (10.5) | 3388 (10.2) | 3385 (10.2) | 3338 (10.1) | 3410 (10.3) | |
| 2007 | 3550 (10.7) | 17913 (10.8) | 3606 (10.8) | 3547 (10.7) | 3547 (10.7) | 3502 (10.6) | 3428 (10.3) | |
| 2008 | 3774 (11.3) | 18487 (11.2) | 3705 (11.1) | 3698 (11.1) | 3681 (11.1) | 3661 (11.0) | 3659 (11.0) | |
| 2009 | 3885 (11.7) | 18855 (11.4) | 3898 (11.7) | 3724 (11.2) | 3761 (11.3) | 3730 (11.2) | 3762 (11.3) | |
| 2010 | 3959 (11.9) | 19021 (11.5) | 3959 (11.9) | 3992 (12.0) | 4019 (12.1) | 3990 (12.0) | 3992 (12.0) | |
| 2011 | 4178 (12.5) | 19869 (12.0) | 4178 (12.5) | 4134 (12.4) | 4170 (12.5) | 4152 (12.5) | 4119 (12.4) | |
| 2012 | 4166 (12.5) | 20469 (12.3) | 4166 (12.5) | 4250 (12.8) | 4262 (12.8) | 4238 (12.8) | 4292 (12.9) | |
| Region | ||||||||
| New England | 730 (2.2) | 10597 (6.4) | 730 (2.2) | 730 (2.2) | 728 (2.2) | 787 (2.4) | 826 (2.5) | |
| Middle Atlantic | 4125 (12.4) | 21965 (13.2) | 4138 (12.4) | 4125 (12.4) | 4113 (12.4) | 4142 (12.5) | 3937 (11.9) | |
| South Atlantic | 12967 (38.9) | 37205 (22.4) | 12967 (38.9) | 12967 (38.9) | 12972 (39) | 12930 (39) | 12745 (38.4) | |
| East North Central | 5313 (16.0) | 30507 (18.4) | 5300 (15.9) | 5313 (16.0) | 5316 (16.0) | 5152 (15.5) | 5316 (16.0) | |
| East South Central | 3256 (9.8) | 13421 (8.1) | 3256 (9.8) | 3256 (9.8) | 3256 (9.8) | 3205 (9.7) | 3241 (9.8) | |
| West North Central | 1125 (3.4) | 13791 (8.3) | 1125 (3.4) | 1125 (3.4) | 1122 (3.4) | 1152 (3.5) | 1164 (3.5) | |
| West South Central | 3942 (11.8) | 13405 (8.1) | 3942 (11.8) | 3942 (11.8) | 3953 (11.9) | 3823 (11.5) | 3893 (11.7) | |
| Mountain | 335 (1.0) | 7926 (4.8) | 335 (1.0) | 335 (1.0) | 335 (1.0) | 415 (1.3) | 405 (1.2) | |
| Pacific | 1506 (4.5) | 16982 (10.2) | 1506 (4.5) | 1506 (4.5) | 1503 (4.5) | 1563 (4.7) | 1624 (4.9) | |
| Insurance | ||||||||
| Uninsured | 4250 (12.8) | 10905 (6.6) | 2444 (7.3) | 4250 (12.8) | 4252 (12.8) | 4221 (12.7) | 4352 (13.1) | |
| Medicaid | 5698 (17.1) | 12029 (7.3) | 2377 (7.1) | 5698 (17.1) | 5704 (17.1) | 5636 (17.0) | 5649 (17.0) | |
| Medicare | 4091 (12.3) | 13846 (8.4) | 2721 (8.2) | 4091 (12.3) | 4085 (12.3) | 4255 (12.8) | 4219 (12.7) | |
| Private | 19260 (57.8) | 129019 (77.8) | 25757 (77.4) | 19260 (57.8) | 19257 (57.8) | 19057 (57.5) | 18931 (57.1) | |
| Comorbidity score | ||||||||
| 0 | 24826 (74.6) | 132112 (79.7) | 26453 (79.4) | 25859 (77.7) | 24862 (74.7) | 24656 (74.3) | 24611 (74.2) | |
| 1 | 6054 (18.2) | 25517 (15.4) | 5236 (15.7) | 5526 (16.6) | 6046 (18.2) | 6104 (18.4) | 6093 (18.4) | |
| ≥ 2 | 2419 (7.3) | 8170 (4.9) | 1610 (4.8) | 1914 (5.7) | 2390 (7.2) | 2409 (7.3) | 2447 (7.4) | |
| Stage | ||||||||
| I | 7086 (21.3) | 40982 (24.7) | 8113 (24.4) | 7501 (22.5) | 7480 (22.5) | 6952 (21.0) | 6995 (21.1) | |
| II | 7595 (22.8) | 39793 (24.0) | 8133 (24.4) | 8138 (24.4) | 8199 (24.6) | 7615 (23.0) | 7569 (22.8) | |
| III | 9411 (28.3) | 47589 (28.7) | 9623 (28.9) | 9492 (28.5) | 9480 (28.5) | 9367 (28.2) | 9351 (28.2) | |
| IV | 9207 (27.6) | 37435 (22.6) | 7430 (22.3) | 8168 (24.5) | 8139 (24.4) | 9235 (27.8) | 9236 (27.9) | |
| Grade | ||||||||
| Well differentiated | 2989 (9.0) | 14645 (8.8) | 2979 (8.9) | 2940 (8.8) | 2939 (8.8) | 2949 (8.9) | 3041 (9.2) | |
| Moderately differentiated | 21367 (64.2) | 105756 (63.8) | 21267 (63.9) | 21065 (63.3) | 21053 (63.2) | 21144 (63.7) | 21058 (63.5) | |
| Poorly differentiated | 4865 (14.6) | 26393 (15.9) | 5238 (15.7) | 5236 (15.7) | 5233 (15.7) | 4914 (14.8) | 4846 (14.6) | |
| Undifferentiated | 421 (1.3) | 2778 (1.7) | 551 (1.7) | 565 (1.7) | 577 (1.7) | 486 (1.5) | 495 (1.5) | |
| Missing/unknown | 3657 (11.0) | 16227 (9.8) | 3264 (9.8) | 3493 (10.5) | 3496 (10.5) | 3676 (11.1) | 3711 (11.2) | |
| Margin | ||||||||
| Margin negative | 25597 (76.9) | 133810 (80.7) | 26996 (81.1) | 26122 (78.4) | 26143 (78.5) | 25353 (76.4) | 25259 (76.2) | |
| Margin positive | 2523 (7.6) | 12161 (7.3) | 2458 (7.4) | 2669 (8.0) | 2671 (8.0) | 2571 (7.8) | 2692 (8.1) | |
| Unknown/other | 5179 (15.6) | 19828 (12.0) | 3845 (11.5) | 4508 (13.5) | 4484 (13.5) | 5245 (15.8) | 5200 (15.7) | |
| Tumor location | ||||||||
| Right colon | 11082 (33.3) | 39996 (24.1) | 8107 (24.3) | 8092 (24.3) | 8202 (24.6) | 10890 (32.8) | 11015 (33.2) | |
| Transverse colon | 2196 (6.6) | 8544 (5.2) | 1756 (5.3) | 1773 (5.3) | 1793 (5.4) | 2157 (6.5) | 2222 (6.7) | |
| Left colon | 9989 (30) | 47762 (28.8) | 9660 (29.0) | 9633 (28.9) | 9637 (28.9) | 9953 (30) | 9934 (30) | |
| Unspecified colon | 1472 (4.4) | 4892 (3.0) | 997 (3.0) | 1103 (3.3) | 1103 (3.3) | 1571 (4.7) | 1500 (4.5) | |
| Rectum | 8560 (25.7) | 64605 (39.0) | 12779 (38.4) | 12698 (38.1) | 12563 (37.7) | 8598 (25.9) | 8480 (25.6) | |
| Node status | ||||||||
| Positive | 13822 (41.5) | 66779 (40.3) | 13591 (40.8) | 13561 (40.7) | 13579 (40.8) | 13730 (41.4) | 13828 (41.7) | |
| Negative | 13478 (40.5) | 73997 (44.6) | 14888 (44.7) | 14348 (43.1) | 14373 (43.2) | 13464 (40.6) | 13396 (40.4) | |
| Chemotherapy | Not examined | 5999 (18.0) | 25023 (15.1) | 4820 (14.5) | 5390 (16.2) | 5346 (16.1) | 5975 (18.0) | 5927 (17.9) |
| No chemotherapy | 13630 (40.9) | 62015 (37.4) | 12474 (37.5) | 12462 (37.4) | 12541 (37.7) | 12839 (38.7) | 13565 (40.9) | |
| Chemotherapy | 19669 (59.1) | 103784 (62.6) | 20825 (62.5) | 20837 (62.6) | 20757 (62.3) | 20330 (61.3) | 19586 (59.1) | |
| Radiation treatment | ||||||||
| No radiation | 27619 (82.9) | 124531 (75.1) | 25195 (75.7) | 25113 (75.4) | 25198 (75.7) | 27213 (82.0) | 27375 (82.6) | |
| Radiation treatment | 5680 (17.1) | 41268 (24.9) | 8104 (24.3) | 8186 (24.6) | 8100 (24.3) | 5956 (18.0) | 5776 (17.4) | |
| Surgery | ||||||||
| No surgery | 7398 (22.2) | 33247 (20.1) | 6549 (19.7) | 7225 (21.7) | 7136 (21.4) | 7723 (23.3) | 7534 (22.7) | |
| Surgery | 25901 (77.8) | 132552 (79.9) | 26750 (80.3) | 26074 (78.3) | 26162 (78.6) | 25446 (76.7) | 25617 (77.3) | |
| Distant site surgery | ||||||||
| None | 30214 (90.7) | 151472 (91.4) | 30263 (90.9) | 30249 (90.8) | 30234 (90.8) | 30027 (90.5) | 29905 (90.2) | |
| Other sites | 1740 (5.2) | 7850 (4.7) | 1694 (5.1) | 1727 (5.2) | 1748 (5.2) | 1703 (5.1) | 1853 (5.6) | |
| Distant site | 1322 (4.0) | 6312 (3.8) | 1309 (3.9) | 1297 (3.9) | 1291 (3.9) | 1407 (4.2) | 1367 (4.1) | |
| Missing/unknown | 23 (0.1) | 165 (0.1) | 33 (0.1) | 26 (0.1) | 25 (0.1) | 32 (0.1) | 26 (0.1) | |
Abbreviations: NH, non-Hispanic
Note: For comorbidity, tumor characteristics, and treatment matching the total number of blacks matched with whites is slightly lower than the total number of blacks as unmatched cases were excluded
Treatment Outcomes
In the unmatched entire cohort, blacks had lower receipt of chemotherapy (59.1% vs 62.6%), radiation treatment (17.1% vs 24.9%) and surgery (only for rectal cancer [83.5% vs 88.8% for stage I and 35.3% vs 43.8% for stage IV]) compared with whites (Table 1 and Supplementary Table 2). These differences in receipt of treatment were not affected by demographic matching, but were affected by insurance matching for receipt of rectal cancer surgery (reduced difference by 33% in relative terms) and by tumor characteristics matching for receipt of radiation and of chemotherapy (reduced difference by 89% and 37% in relative terms, respectively).
Survival Outcomes
The absolute 5-year survival difference between black and white CRC patients in the unmatched entire cohort was 9.2% (57.3% vs 66.5%) (Table 2 and Figure 2). The absolute survival difference remained unchanged after demographic matching but decreased to 4.9% after insurance matching, to 4.7% after comorbidity matching, and to 2.3% after tumor characteristics matching; there was no further reduction after subsequent treament matching.
Table 2.
5-year survival for 18–64 years old NH blacks and sequentially matched NH whites with colorectal cancer
| Entire Cohort | Demographic Match | Insurance Match | Comorbidity Match | Tumor Characteristics Match | Treatment Match | |
|---|---|---|---|---|---|---|
| 5-year survival (95% CI) | ||||||
| NH white | 66.5 (66.3–66.8) | 66.5 (65.9–67.1) | 62.2 (61.6–62.8) | 62.0 (61.3–62.6) | 59.6 (59.0–60.2) | 59.6 (58.9–60.2) |
| NH black | 57.3 (56.6–57.9) | 57.3 (56.6–57.9) | 57.3 (56.6–57.9) | 57.3 (56.6–57.9) | 57.3 (56.7–57.9) | 57.3 (56.7–58.0) |
| Absolute difference (%) | 9.2 | 9.2 | 4.9 | 4.7 | 2.3 | 2.3 |
| Percentage difference explained (%) | ||||||
| Total | 0.0 | 46.7 | 48.9 | 75.0 | 75.0 | |
| Individual | 0.0 | 46.7 | 2.2 | 26.1 | 0.0 |
Abbreviations: CI, confidence interval; NH, non-Hispanic
Figure 2.
Overall 5-year survival curves for 18–64 years old NH blacks and five sequentially matched 18–64 years old NH whites with colorectal cancer.
In stratified analyses, the absolute survival difference between black and white CRC patients ranged from 3.8% in stage I patients to 8.5% in stage III patients, and from 5.3% for right colon cancer patients to 11.5% for rectal cancer patients (Supplementary Table 3). Similar to the overall analysis, matching by demographics did not substantially alter the absolute survival difference between blacks and whites for any of the stages or anatomic subsites. Likewise, subsequent matching by insurance substantially reduced the absolute survival difference for all categories, especially among stage II patients (from 5.5% to 1.9%) and among right colon cancer patients (from 5.4% to 3.5%).
In the proportional hazards model, overall blacks had a 40% (HR, 1.40; 95% confidence interval [CI], 1.37–1.43) excess risk of all-cause mortality compared with unmatched whites (Table 3 and Figure 3). The excess risk was reduced to 19% (HR, 1.19; 95% CI, 1.16–1.22) after insurance matching, to 18% (HR, 1.18; 95% CI, 1.15–1.21) after comorbidity matching, to 7% (HR, 1.07; 95% CI, 1.04–1.10) after tumor characteristics matching, and to 6% (HR, 1.06; 95% CI, 1.03–1.09) after treatment matching.
Table 3.
Hazard ratios of all-cause 5-year mortality risk for 18–64 years old NH black vs NH white patients with colorectal cancer for overall and according to tumor stage and location
| Overall
| |||||
|---|---|---|---|---|---|
| HR (95% CI) | |||||
|
|
|||||
| Crude | 1.40 (1.37–1.43) | ||||
| Demographic matcheda | 1.41 (1.37–1.45) | ||||
| Insurance matchedb | 1.19 (1.16–1.22) | ||||
| Comorbidity matchedc | 1.18 (1.15–1.21) | ||||
| Tumor characteristics matchedd | 1.07 (1.04–1.10) | ||||
| Treatment matchede | 1.06 (1.03–1.09) | ||||
|
| |||||
| Stratified by Stage | |||||
|
| |||||
| Stage I | Stage II | Stage III | Stage IV | ||
|
|
|||||
| HR (95% CI) | HR (95% CI) | HR (95% CI) | HR (95% CI) | ||
|
|
|||||
| Crude | 1.47 (1.35–1.59) | 1.39 (1.31–1.47) | 1.39 (1.33–1.45) | 1.24 (1.21–1.27) | |
| Demographic matcheda | 1.44 (1.29–1.61) | 1.38 (1.28–1.49) | 1.40 (1.32–1.48) | 1.24 (1.19–1.28) | |
| Insurance matchedb | 1.16 (1.05–1.30) | 1.11 (1.03–1.19) | 1.20 (1.14–1.27) | 1.13 (1.09–1.16) | |
| Comorbidity matchedc | 1.12 (1.01–1.25) | 1.09 (1.01–1.17) | 1.20 (1.13–1.27) | 1.11 (1.08–1.15) | |
| Tumor characteristics matchedd | 1.08 (0.97–1.20) | 1.10 (1.02–1.18) | 1.15 (1.09–1.22) | 1.06 (1.02–1.09) | |
| Treatment matchede | 1.07 (0.96–1.19) | 1.09 (1.01–1.17) | 1.12 (1.06–1.18) | 1.05 (1.02–1.09) | |
|
| |||||
| Stratified by tumor location | |||||
|
| |||||
| Right colon | Transverse colon | Left colon | Unspecified location colon | Rectum | |
|
|
|||||
| HR (95% CI) | HR (95% CI) | HR (95% CI) | HR (95% CI) | HR (95% CI) | |
|
|
|||||
| Crude | 1.18 (1.14–1.23) | 1.29 (1.19–1.40) | 1.43 (1.37–1.48) | 1.18 (1.09–1.28) | 1.55 (1.49–1.61) |
| Demographic matcheda | 1.20 (1.14 –1.27) | 1.30 (1.16–1.46) | 1.41 (1.34–1.48) | 1.27 (1.13–1.42) | 1.55 (1.48–1.63) |
| Insurance matchedb | 1.05 (1.00–1.10) | 1.08 (0.97–1.21) | 1.18 (1.12–1.24) | 1.08 (0.97–1.20) | 1.30 (1.24–1.36) |
| Comorbidity matchedc | 1.05 (1.00–1.11) | 1.08 (0.97–1.20) | 1.17 (1.11–1.23) | 1.06 (0.95–1.18) | 1.28 (1.22–1.34) |
| Tumor characteristics matchedd | 1.01 (0.96–1.06) | 1.07 (0.06–1.19) | 1.11 (1.05–1.16) | 0.99 (0.89–1.09) | 1.16 (1.11–1.22) |
| Treatment matchede | 1.02 (0.97–1.07) | 1.06 (0.96–1.18) | 1.12 (1.06–1.17) | 1.01 (0.91–1.11) | 1.10 (1.04–1.15) |
Abbreviations: CI, confidence interval; HR, hazard ratio; NH, non-Hispanic
Matched for diagnosis age, diagnosis year, sex, region
Additionally matched for insurance
Additionally matched for comorbidity
Additionally matched for stage, grade, margin, tumor location, node status
Additionally matched for chemotherapy, radiotherapy, surgery, surgery of other site
Note: stage not used for matching when used to stratify and tumor location not used for matching when used to stratify
Figure 3.
Hazard ratios of all-cause 5-year mortality risk for 18–64 years old NH blacks versus sequentially matched 18–64 years old NH whites for patients diagnosed with colorectal cancer Abbreviation: NH, non-Hispanic
In the stratified analyses, the excess risk of deaths in black patients in the umatched cohort ranged from 24% (HR, 1.24; 95% CI, 1.21–1.27) among stage IV patients to 47% (HR, 1.47; 95% CI, 1.35–1.59) among stage I patients, and from 18% (HR, 1.18; 95% CI, 1.14–1.23) among right colon cancer patients to 55% (HR, 1.55; 95% CI, 1.49–1.61) among rectal cancer patients (Table 3). Matching by insurance and tumor characteristics substantially reduced the excess risk of deaths in blacks in all stage and tumor location categories although the contributions were larger for insurance. Further matching by treatment substantially reduced the excess risk of death in blacks among rectal cancer patients but not among patients with other tumor locations (Table 3).
In a sensitivity five-year survival analysis restricted to patients diagnosed between 2004–2008, we found that the absolute overall survival difference between blacks and whites (Supplementary Table 4) was similar to those from our main analysis. In another sensitivity analysis, stratifying treatment-matched subgroup by facility case volume did not change the black/white 5-year survival differences (Supplementary Table 5).
In the supplementary analysis to estimate the individual contribution of each of the five sets of factors to the survival disparity without sequential matching, insurance coverage alone accounted for the largest overall black-white survival disparity (50%), closely followed by tumor characteristics alone (49%), followed by treatment receipt alone (20%), and comorbidity alone (5%) (Supplementary Table 6; percentages are not necessarily additive to 100%).
Discussion
In a large nationwide oncologic outcomes database, we showed that differences in health insurance coverage accounted for nearly one-half and differences in tumor characteristics accounted for approximately a quarter of the black-white survival disparity in CRC patients aged 18–64 years old. These findings reinforce the importance of equitable access to care to mitigate the survival disparity between black and white CRC patients in this age range, and underscore the need for further studies to elucidate reasons for racial differences in tumor characteristics.
Several previous studies demonstrated significant disparities in cancer stage at diagnosis and survival for major cancers by insurance status.29, 30 For instance, Halpern et al reported that uninsured or Medicaid-insured patients had two-fold higher odds of being diagnosed with advanced-stage CRC compared with privately insured patients.29 Certainly, lack of health insurance coverage for patients could greatly impact access to timely diagnosis and treatment,30 with lower receipt of screening tests and follow-up for abnormal test results.31 Compared with privately-insured patients, uninsured patients were less likely to have a primary source of care for preventive services, more likely to be diagnosed with advanced stage diseases, less likely to receive standard of care, and more likely to die of their diseases.29, 30, 32 A recent study reported a narrowing of the black-white disparity in insurance coverage following the implementation of the Affordable Care Act (ACA),33 and whether this leads to reducing black-white CRC survival disparity remains to be seen.
Our findings of the contribution of black-white differences in tumor characteristics to the racial survival disparity among CRC patients are in line with previous studies,12, 15, 34 although in relative terms the contribution in our study is much lower compared to those from two recent studies of elderly (≥ 65 years) colon cancer patients (40%–50% in those two studies compared to 26% in our study).12, 15 In addition to differences in age (≥ 65 years vs 18–64 years old patients), our study differs from those of Lai et al12 and Silber et al15 in some other important ways. For example, Lai et al12 and Silber et al15 did not account for insurance coverage, possibly because of the nearly universal health insurance coverage of elderly patients through Medicare, nor did they adjust for differences in primary tumor location despite the fact that blacks are known to have a higher proportion of right-sided colon tumors (which have less favorable outcomes) compared to whites.35 Because of the higher risk of right-sided colon cancer in younger blacks, the American College of Gastroenterology’s Committee of Minority Affairs and Cultural Diversity recently recommended CRC screening to start at age 45 years instead of 50 years in blacks and to use colonoscopy instead of sigmoidoscopy.36
In the stratified analysis, accounting for insurance coverage difference substantially reduced the black-white survival disparities across disease stages and tumor locations, reflecting the importance of insurance coverage to earlier diagnosis and/or receipt of timely and high quality treatment. Several previous studies documented black-white disparity in receipt of screening and adjuvant therapy for CRC.6, 37, 38
Our findings of large differences in receipt of surgical resection in black versus white patients with stage I and stage IV rectal cancer, despite accounting for differences in insurance and tumor stage, were in agreement with previous studies.18 Reasons for these differences may reflect differences in access to surgeons, high volume hospitals, or patient/physician preferences .22, 39, 40 Several studies showed that blacks were more likely to reside in counties with lower availability of colorectal surgeons,40 or receive treatment in low volume hospitals,41 although a sensitivity analysis stratifying by hospital volume did not alter our results. Likewise, a number of studies have documented lower awareness about medical procedures and higher refusal of invasive procedures such as proctocolectomy and other forms of treatment in blacks compared with whites.39
Although we matched blacks with partially overlapping subgroups of whites by demographic characteristics, insurance status, comorbidity status, tumor characteristics, and treatment, approximately 25% of the black-white total survival difference (2.3% of 9.2%) among the unmatched CRC patients remains unexplained. Factors that were not considered in our study and that may contribute to the unexplained racial survival disparity include differences in completion of treatment, quality of treatment, response to treatment, adverse events, socioeconomic indicators other than insurance, and prognostic biological characteristics.22, 42–46 For example, Carethers et al43 found that microsatellite instability of CRC (a marker of favorable prognosis) was twice as common in white patients (14%) as in black patients (7%). However, the upstream factors that contribute to such molecular tumor differences require more research.
A strength of our study is the use of large nationwide oncologic outcomes database for patients to estimate the contribution of demographic characteristics, insurance coverage, comorbidity score, tumor characteristics, and treatment to the overall survival disparity between black and white CRC patients aged 18–64 years old using a sequential matching method. However, there are several limitations in our study. First, there could be ascertainment errors and differences between patients included versus excluded in the analysis. However, data collection in the NCDB was highly standardized and performed by certified tumor registrars,25 and the demographic and clinical characteristics of patients were generally similar between included and excluded patients though with some differences in tumor characteristics, and comorbidity (data not shown). Second, we matched by broad age groups and stage categories rather than single age interval and stage subcategories. Third, we were also unable to account for differences in other barriers of access to care besides insurance coverage, including copayments, geographic availability of providers, and inability to take time off from work for healthcare services. Fourth, the relationship between access to medical care, insurance coverage, comorbidity, and tumor characteristics may be complex and interpretations of our findings could be affected by this somewhat. Fifth, except for radiation, the NCDB collects data on initiation of treatment, but not dose or completion of treatment. Finally, the NCDB is a hospital-based registry rather than population-based cancer registry and findings may not be generalizable to all U.S. patients. However, sociodemographic and clinical characteristics of patients and survival outcomes in the NCDB for common cancers including CRC were comparable with those of the National Cancer Institute’s population-based Surveillance Epidemiology and End Results (SEER) data.47, 48 For example, the 5-year overall survival for white and black 18–64 years old CRC patients diagnosed from 2004–2008 in the SEER database were 65.1% and 53.2%, respectively, which are comparable to our findings based on the NCDB (66.5% vs 57.3%).49 We also note that we did not study those 65 years of age and older, in part allowing us to better investigate the role of variations in access to care. However, further study of these older age groups may be useful as well.
Conclusion
Health insurance coverage differences accounted for nearly one-half of the black-white survival disparity in 18–64 years old CRC patients, whereas tumor characteristics explained a quarter of the disparity. Our findings reinforce the importance of equitable health insurance coverage to mitigate the black-white survival disparity in 18–64 years old CRC patients. Future research should explore the biological mechanisms for the differences in tumor characteristics and implications for treatment.
Supplementary Material
Acknowledgments
Funding: This work was supported by the American Cancer Society Intramural Research (no grant number applicable to HMS, WDF, OWB, AJ) ; and National Institutes of Health/National Cancer Institute (NIH/NCI R01 CA205406 to KN).
The data used in the study are derived from a limited data set of the National Cancer Data Base (NCDB). The authors acknowledge the efforts of the American College of Surgeons, the Commission on Cancer and the American Cancer Society in the creation of the National Cancer Data Base. 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 authors.
Abbreviations
- CRC
colorectal cancer
- HR
hazard ratio
- NCDB
National Cancer Data Base
- NH
non-Hispanic
Footnotes
Conflict of Interest: The authors disclose no conflicts of interest.
The preliminary finding of this study was presented at the American Society of Clinical Oncology 2017 Annual Conference on the in Chicago, Illinois, June 3, 2017.
Author Contributions: Study concept and design (Helmneh M. Sineshaw, Kimmie Ng, Ahmedin Jemal); acquisition of data (Helmneh M. Sineshaw, Ahmedin Jemal); analysis and interpretation of data (all authors); drafting of the manuscript (Helmneh M. Sineshaw, Ahmedin Jemal); critical revision of the manuscript for important intellectual content (all authors); statistical analysis (Helmneh M. Sineshaw, W. Dana Flanders, Ahmedin Jemal); administrative, technical, or material support (Helmneh M. Sineshaw, Ahmedin Jemal); study supervision (Ahmedin Jemal)
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
Author names in bold designate shared co-first authorship.
- 1.Siegel RL, Miller KD, Jemal A. Cancer statistics, 2017. CA Cancer J Clin. 2017;67:7–30. doi: 10.3322/caac.21387. [DOI] [PubMed] [Google Scholar]
- 2.Kohler BA, Sherman RL, Howlader N, et al. Annual report to the nation on the status of cancer, 1975–2011, featuring incidence of breast cancer subtypes by race/ethnicity, poverty, and state. J Natl Cancer Inst. 2015;107:djv048. doi: 10.1093/jnci/djv048. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Siegel RL, Fedewa SA, Anderson WF, et al. Colorectal Cancer Incidence Patterns in the United States, 1974–2013. J Natl Cancer Inst. 2017;109:djw322. doi: 10.1093/jnci/djw322. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Siegel RL, Miller KD, Jemal A. Cancer statistics, 2016. CA Cancer J Clin. 2016;66:7–30. doi: 10.3322/caac.21332. [DOI] [PubMed] [Google Scholar]
- 5.Rahman R, Schmaltz C, Jackson CS, et al. Increased risk for colorectal cancer under age 50 in racial and ethnic minorities living in the United States. Cancer Med. 2015;4:1863–1870. doi: 10.1002/cam4.560. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Jessup JM, Stewart A, Greene FL, et al. Adjuvant chemotherapy for stage III colon cancer: implications of race/ethnicity, age, and differentiation. JAMA. 2005;294:2703–2711. doi: 10.1001/jama.294.21.2703. [DOI] [PubMed] [Google Scholar]
- 7.Potosky AL, Harlan LC, Kaplan RS, et al. Age, sex, and racial differences in the use of standard adjuvant therapy for colorectal cancer. J Clin Oncol. 2002;20:1192–1202. doi: 10.1200/JCO.2002.20.5.1192. [DOI] [PubMed] [Google Scholar]
- 8.Holowatyj AN, Ruterbusch JJ, Rozek LS, et al. Racial/Ethnic Disparities in Survival Among Patients With Young-Onset Colorectal Cancer. J Clin Oncol. 2016;34:2148–2156. doi: 10.1200/JCO.2015.65.0994. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Wallace K, Hill EG, Lewin DN, et al. Racial disparities in advanced-stage colorectal cancer survival. Cancer Causes Control. 2013;24:463–471. doi: 10.1007/s10552-012-0133-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Wallace K, Sterba KR, Gore E, et al. Prognostic factors in relation to racial disparity in advanced colorectal cancer survival. Clin Colorectal Cancer. 2013;12:287–293. doi: 10.1016/j.clcc.2013.08.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Phatak UR, Kao LS, Millas SG, et al. Interaction between age and race alters predicted survival in colorectal cancer. Ann Surg Oncol. 2013;20:3363–3369. doi: 10.1245/s10434-013-3045-z. [DOI] [PubMed] [Google Scholar]
- 12.Lai Y, Wang C, Civan JM, et al. Effects of Cancer Stage and Treatment Differences on Racial Disparities in Survival From Colon Cancer: A United States Population-Based Study. Gastroenterology. 2016 doi: 10.1053/j.gastro.2016.01.030. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Soneji S, Iyer SS, Armstrong K, et al. Racial disparities in stage-specific colorectal cancer mortality: 1960–2005. Am J Public Health. 2010;100:1912–1916. doi: 10.2105/AJPH.2009.184192. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Mayberry RM, Coates RJ, Hill HA, et al. Determinants of Black/White Differences in Colon Cancer Survival. J Natl Cancer Inst. 1995;87:1686–1693. doi: 10.1093/jnci/87.22.1686. [DOI] [PubMed] [Google Scholar]
- 15.Silber JH, Rosenbaum PR, Ross RN, et al. Racial disparities in colon cancer survival: a matched cohort study. Ann Intern Med. 2014;161:845–854. doi: 10.7326/M14-0900. [DOI] [PubMed] [Google Scholar]
- 16.Marcella S, Miller JE. Racial differences in colorectal cancer mortality: the importance of stage and socioeconomic status. J Clin Epidemiol. 2001;54:359–366. doi: 10.1016/s0895-4356(00)00316-4. [DOI] [PubMed] [Google Scholar]
- 17.Wray CM, Ziogas A, Hinojosa MW, et al. Tumor subsite location within the colon is prognostic for survival after colon cancer diagnosis. Dis Colon Rectum. 2009;52:1359–66. doi: 10.1007/DCR.0b013e3181a7b7de. [DOI] [PubMed] [Google Scholar]
- 18.Le H, Ziogas A, Lipkin SM, et al. Effects of socioeconomic status and treatment disparities in colorectal cancer survival. Cancer Epidemiol Biomarkers Prev. 2008;17:1950–62. doi: 10.1158/1055-9965.EPI-07-2774. [DOI] [PubMed] [Google Scholar]
- 19.Breslin TM, Morris AM, Gu N, et al. Hospital factors and racial disparities in mortality after surgery for breast and colon cancer. J Clin Oncol. 2009;27:3945–3950. doi: 10.1200/JCO.2008.20.8546. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Tehranifar P, Neugut AI, Phelan JC, et al. Medical Advances and Racial/Ethnic Disparities in Cancer Survival. Cancer Epidemiol Biomarkers Prev. 2009;18:2701–2708. doi: 10.1158/1055-9965.EPI-09-0305. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Du XL, Fang S, Vernon SW, et al. Racial disparities and socioeconomic status in association with survival in a large population-based cohort of elderly patients with colon cancer. Cancer. 2007;110:660–669. doi: 10.1002/cncr.22826. [DOI] [PubMed] [Google Scholar]
- 22.Haas JS, Brawarsky P, Iyer A, et al. Association of area sociodemographic characteristics and capacity for treatment with disparities in colorectal cancer care and mortality. Cancer. 2011;117:4267–4276. doi: 10.1002/cncr.26034. [DOI] [PubMed] [Google Scholar]
- 23.Simpson DR, Martínez ME, Gupta S, et al. Racial disparity in consultation, treatment, and the impact on survival in metastatic colorectal cancer. J Natl Cancer Inst. 2013;105:1814–1820. doi: 10.1093/jnci/djt318. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.American College of Surgeons. [Accessed January 26, 2017];National Cancer Data Base: 2017 update. Available at: https://www.facs.org/quality-programs/cancer/ncdb.
- 25.Commission on Cancer. [Accessed on November 16, 2016];FORDS: Facility Oncology Registry Data Standards—Revised for 2012. Available at: https://www.facs.org/~/media/files/quality%20programs/cancer/coc/fords/fords%202012c.ashx.
- 26.Parsons LS. Reducing bias in a propensity score matched-pair sample using greedy matching techniques. Proceedings of the Twenty-sixth Annual SAS Users group international conference; Cary, NC: SAS Institute; 2001. pp. 214–226. [Google Scholar]
- 27.Rosenbaum PR, Rubin DB. Reducing bias in observational studies using subclassification on the propensity score. J Am Stat Assoc. 1984;79:516–524. [Google Scholar]
- 28.Charlson ME, Pompei P, Ales KL, et al. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40:373–383. doi: 10.1016/0021-9681(87)90171-8. [DOI] [PubMed] [Google Scholar]
- 29.Halpern MT, Ward EM, Pavluck AL, et al. Association of insurance status and ethnicity with cancer stage at diagnosis for 12 cancer sites: a retrospective analysis. Lancet Oncol. 2008;9:222–231. doi: 10.1016/S1470-2045(08)70032-9. [DOI] [PubMed] [Google Scholar]
- 30.Ward E, Halpern M, Schrag N, et al. Association of insurance with cancer care utilization and outcomes. CA Cancer J Clin. 2008;58:9–31. doi: 10.3322/CA.2007.0011. [DOI] [PubMed] [Google Scholar]
- 31.Sabatino SA, White MC, Thompson TD, et al. Cancer screening test use—United States, 2013. MMWR Morb Mortal Wkly Rep. 2015;64:464–468. [PMC free article] [PubMed] [Google Scholar]
- 32.Ayanian JZ, Weissman JS, Schneider EC, et al. Unmet health needs of uninsured adults in the United States. JAMA. 2000;284:2061–2069. doi: 10.1001/jama.284.16.2061. [DOI] [PubMed] [Google Scholar]
- 33.McMorrow S, Long SK, Kenney GM, et al. Uninsurance disparities have narrowed for Black and Hispanic adults under the Affordable Care Act. Health Aff (Millwood) 2015;34:1774–1778. doi: 10.1377/hlthaff.2015.0757. [DOI] [PubMed] [Google Scholar]
- 34.Doubeni CA, Field TS, Buist DS, et al. Racial differences in tumor stage and survival for colorectal cancer in an insured population. Cancer. 2007;109:612–20. doi: 10.1002/cncr.22437. [DOI] [PubMed] [Google Scholar]
- 35.Thornton JG, Morris AM, Thornton JD, et al. Racial variation in colorectal polyp and tumor location. J Natl Med Assoc. 2007;99:723. [PMC free article] [PubMed] [Google Scholar]
- 36.Agrawal S, Bhupinderjit A, Bhutani MS, et al. Colorectal cancer in African Americans. Am J Gastroenterol. 2005;100:515–523. doi: 10.1111/j.1572-0241.2005.41829.x. [DOI] [PubMed] [Google Scholar]
- 37.Baldwin L-M, Dobie SA, Billingsley K, et al. Explaining black–white differences in receipt of recommended colon cancer treatment. J Natl Cancer Inst. 2005;97:1211–1220. doi: 10.1093/jnci/dji241. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Ayanian JZ, Zaslavsky AM, Fuchs CS, et al. Use of adjuvant chemotherapy and radiation therapy for colorectal cancer in a population-based cohort. J Clin Oncol. 2003;21:1293–1300. doi: 10.1200/JCO.2003.06.178. [DOI] [PubMed] [Google Scholar]
- 39.Demissie K, Oluwole OO, Balasubramanian BA, et al. Racial differences in the treatment of colorectal cancer: a comparison of surgical and radiation therapy between Whites and Blacks. Ann Epidemiol. 2004;14:215–21. doi: 10.1016/j.annepidem.2003.08.002. [DOI] [PubMed] [Google Scholar]
- 40.Hayanga AJ, Waljee AK, Kaiser HE, et al. Racial clustering and access to colorectal surgeons, gastroenterologists, and radiation oncologists by African Americans and Asian Americans in the United States: a county-level data analysis. Arch Surg. 2009;144:532–535. doi: 10.1001/archsurg.2009.68. [DOI] [PubMed] [Google Scholar]
- 41.Huang LC, Tran TB, Ma Y, et al. What Factors Influence Minority Use of High Volume Hospitals for Colorectal Cancer Care. Dis Colon Rectum. 2015;58:526. doi: 10.1097/DCR.0000000000000353. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Phipps AI, Limburg PJ, Baron JA, et al. Association between molecular subtypes of colorectal cancer and patient survival. Gastroenterology. 2015;148:77–87. e2. doi: 10.1053/j.gastro.2014.09.038. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Carethers JM, Murali B, Yang B, et al. Influence of race on microsatellite instability and CD8+ T cell infiltration in colon cancer. PLoS One. 2014;9:e100461. doi: 10.1371/journal.pone.0100461. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Ball JK, Elixhauser A. Treatment differences between blacks and whites with colorectal cancer. Med Care. 1996;34:970–984. doi: 10.1097/00005650-199609000-00008. [DOI] [PubMed] [Google Scholar]
- 45.Sanoff HK, Sargent DJ, Green EM, et al. Racial differences in advanced colorectal cancer outcomes and pharmacogenetics: a subgroup analysis of a large randomized clinical trial. J Clin Oncol. 2009;27:4109–4115. doi: 10.1200/JCO.2009.21.9527. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Yoon HH, Shi Q, Alberts SR, et al. Racial differences in BRAF/KRAS mutation rates and survival in stage III Colon cancer patients. J Natl Cancer Inst. 2015;107:djv186. doi: 10.1093/jnci/djv186. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Mettlin CJ, Menck HR, Winchester DP, et al. A comparison of breast, colorectal, lung, and prostate cancers reported to the National Cancer Data Base and the Surveillance, Epidemiology, and End Results Program. Cancer. 1997;79:2052–2061. doi: 10.1002/(sici)1097-0142(19970515)79:10<2052::aid-cncr29>3.0.co;2-s. [DOI] [PubMed] [Google Scholar]
- 48.Ries L, Melbert D, Krapcho M, et al. SEER Cancer Statistics Review, 1975–2014. National Cancer Institute; Bethesda, MD, USA: [Google Scholar]
- 49.Surveillance Epidemiology, and End Results (SEER) Program Research Data. National Cancer Institute, DCCPS, Surveillance Research Program; 1973–2014. [Accessed on August 2, 2017]. released April 2017, based on the November 2016 submission. Available at: www.seer.cancer.gov. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.