Abstract
Background:
Access to health care is associated with cancer survival. The U.S. military health system (MHS) provides universal health care to beneficiaries, reducing barriers to medical care access. However, it is unknown whether the universal care has translated into improved survival among colon cancer patients. We compared survival of colon cancer patients in the MHS to that in the U.S. general population and assessed whether stage at diagnosis differed between the two populations and thus could contribute to survival difference.
Methods:
The data were from Department of Defense’s (DoD) Automated Central Tumor Registry (ACTUR) and the National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) program, respectively. The ACTUR (N=11,907) and SEER patients (N=23,814) were matched to on demographics and diagnosis year with a matching ratio of 1:2. Multivariable Cox regression model was used to estimate all-cause mortality for ACTUR compared to SEER.
Results:
ACTUR patients exhibited better survival than their SEER counterparts (hazard ratio = 0.82, 95% Confidence Interval= 0.79–0.87) overall and in most subgroups by age, in both men and women, and in whites and blacks. The better survival remained when the comparison was stratified by tumor stage.
Conclusions:
Colon cancer patients in a universal health care system had better survival than patients in the general population.
Impact:
Universal care access is important to improve survival of colon cancer patients.
Keywords: colon cancer, universal health care, survival, military health system, SEER
Introduction
Colorectal cancer is the third most commonly diagnosed cancer and the third leading cause of cancer death in the United States. (1) In the U.S. general population, access to health care, as reflected by insurance status, is associated with survival outcome among cancer patients. (2–7) Colon cancer patients without health insurance or with Medicaid have poorer survival than those with private insurance, (2, 3, 5) which is often attributed to lower accessibility to care or care quality. (2, 5, 8) Barriers to health care access may lead to a later tumor stage at diagnosis (2, 8, 9), lower adherence to treatment guidelines, (10, 11) and lower quality of care received. (4) These factors may contribute to the poorer survival.
The U.S. military health system (MHS) provides universal health care access to its beneficiaries, including active duty service members, National Guard and Reserve members, retirees and their family members. The beneficiaries receive health care free of charge or with minimal out-of-pocket cost. (12) A presumable benefit of the universal care access is improved health outcome. However, it is unknown whether the universal health care in the MHS has translated into improved survival for colon cancer. In this study, we hypothesized that colon cancer patients in the MHS have better survival than patients in the U.S. general population. Since patients in a universal health system may have higher accessibility to cancer screening and less financial barrier to get diagnosis, we further hypothesized that patients in MHS may have an earlier tumor stage at diagnosis. As tumor stage at diagnosis is an important factor for cancer prognosis, it may contribute to the survival difference. The first aim of this study was to compare overall survival of colon cancer patients in the MHS with colon cancer cases in the U.S. general population. The second aim was to assess whether the difference in tumor stage at diagnosis could contribute to the survival difference by comparing tumor stage between the two populations and examining the survival difference by tumor stage.
Materials and Methods
Data Sources
The data sources for this study were the Department of Defense’s (DoD) Automated Central Tumor Registry (ACTUR) and the National Cancer Institute’s (NCI) Surveillance, Epidemiology, and End Results (SEER) program. The ACTUR is the DoD’s cancer registry that tracks cancer patients who are diagnosed and/or receive cancer treatment at military treatment facilities (MTFs). MTFs are required to report to the ACTUR cancer patients diagnosed or treated at their facilities. The ACTUR contains information on demographics, tumor characteristics, cancer treatment, follow up, vital status and other characteristics. ACTUR follows all patients for vital status according to the Commission on Cancer (CoC) of the American College of Surgeons (ACoS) standards.(13) The ACTUR complies with the uniform data standards set by the North American Association of Central Cancer Registries (NAACCR).(14) The SEER collects population-based data on cancer cases within the areas served by SEER cancer registries and contains data on demographics, tumor characteristics, first course of treatment, follow up, vital status and other information. SEER cancer registries follow the CoC requirements on follow up and vital status.(15) In this study, we used data from the SEER-18 registries (Atlanta, Connecticut, Detroit, Hawaii, Iowa, New Mexico, San Francisco-Oakland, Seattle-Puget Sound, Utah, Los Angeles, San Jose-Monterey, Rural Georgia, the Alaska Native, Greater California, Greater Georgia, Kentucky, Louisiana, and New Jersey ), which represents 28% of the U.S. general population.(16)
This study was based on the non-identifiable ACTUR data approved by the institutional review board of Walter Reed National Military Medical Center and the SEER data de-identified for public use.
Study Population
Patients included in the study were those histologically diagnosed with colon adenocarcinoma between January 1, 1987 and December 31, 2013 from ACTUR and aged 18 years or older at diagnosis. Colon adenocarcinoma was defined with the cancer site codes (C18.0, C18.2-C18.9) and morphology codes for adenocarcinoma according to the International Classification of Diseases for Oncology, third edition (ICD-O-3), (17) or converted to ICD-O-3 for patients diagnosed in earlier periods. Patients with diagnosis from death certificate or autopsy were excluded. Cases with multiple primary cancers were excluded to minimize possible effects of other cancers on the study outcomes. The same inclusion and exclusion criteria were applied to both ACTUR and SEER cases.
To reduce potential confounding effects from age, sex, race, and diagnosis year on the survival outcome, we matched SEER cases to ACTUR cases on age (within 5 years), sex (male and female), race (White, Black, Asian/Pacific islander), and diagnosis year group (1987–1989, 1990–1994, 1995–1999, 2000–2004, 2005–2009, 2010–2013) with a matching ratio of 2:1. Cases with missing values on a matching variable were excluded.
Study Variables
Data extracted on each case were demographics, tumor characteristics, site-specific surgery, vital status, and follow up variables as appropriate. Tumor stages were defined I, II, III, and IV according to the American Joint Committee on Cancer (AJCC)’s criteria. Tumor grade was defined as well-differentiated (grade I), moderately-differentiated (grade II), poorly differentiated (grade III), non-differentiated (grade IV), (18) and unknown. Tumor location was grouped as right, transverse, left, overlapping and unknown. Surgery is recommended as the primary treatment for stages I/II/III colon cancer. (19) Site-specific cancer-directed surgery codes were used define cancer-directed surgery types according to SEER guidelines,(20) and then grouped into “cancer-directed surgery received”, “no cancer-directed surgery” or “unknown or missing” according to SEER guidelines. Survival was defined based on all-cause death because data on cancer-specific mortality from ACTUR was incomplete. Other variables included in the study were Hispanic ethnicity (Non-Hispanic, Hispanic, or unknown), region of diagnosis (Northeast, South, Midwest, West, Other) and age as a continuous variable.
Statistical Analysis
We first compared the distributions of demographic and tumor characteristics between ACTUR and SEER patients using the Chi-square test. We then conducted survival analysis. The study outcome was all cause mortality during five years after diagnosis. If a patient died during the 5-year period after diagnosis, the follow up time was calculated as the time from diagnosis to death. If death was not observed during the period, follow-up time was censored at the end of the fifth year. Patients who were not dead through the end of the study without a full five-year follow-up time were censored on the study ending date, December 31, 2013. In survival analysis, we first used Kaplan-Meier curve to compare overall survival between cases from ACTUR and SEER. We then used multivariable Cox proportional hazards model for matched data to estimate hazard ratios (HRs) and 95% CIs for ACTUR relative to SEER. The proportional hazards assumption was checked by plotting the log-log survival curves. (21) To further control for potential confounding variables that were not matched, we adjusted for demographic variables other than the matching variables, including region of diagnosis and continuous age, tumor stage, tumor grade, tumor location and cancer-directed surgery.
Since SEER contains data on insurance type from the year 2007, we further compared survival of the ACTUR patients with that of the SEER patients by insurance type, using the 2007–2013 data. Insurance types in the SEER data included “insured”, “insurance/no specifics”, “any Medicaid”, “uninsured”, and “unknown insurance status”.
To assess if the survival differences were independent of tumor stage, the Cox survival models were stratified by tumor stage (stage I, stage II, stage III, or unknown). Tumor stages were divided as early stage (stages I and II combined) and late stage (stages III and IV combined). Prevalence ratios (PRs) and 95% confidence intervals (95% CIs) were estimated. Potential confounding variables were adjusted for in the analysis.
All statistical analyses were conducted using the SAS software version 9.4.0 (SAS Institute, Inc.). All reported p values are two sided with the significance level set at p<0.05.
Results
There were 11,907 eligible colon cancer cases identified from ACTUR and 23,814 matched cases from SEER. Table 1 shows the distributions of the characteristics between ACTUR and SEER. There was difference between the two populations in region of diagnosis variable. In tumor variables, ACTUR cases were more likely to present with stage I disease (22.67% vs. 18.64%) and less likely to be with stage IV disease (18.74% vs. 21.63%) than SEER cases; they were more likely to have grade I tumor (15.36% vs. 9.71%) and less likely to have grade III (14.54% vs. 18.41%) or grade IV (0.29% vs. 0.95%) tumor. The two populations also differed by tumor location with a higher proportion of tumors of unknown location in ACTUR than SEER (8.43% vs. 2.16%). The percentage of not receiving surgery was lower in ACTUR than SEER (5.45% vs. 7.24%), while the percentage of receiving surgery was similarly high in both populations (92.9% and 92.4% in ACTUR and SEER, respectively) (Table 1).
Table 1.
Demographic and tumor characteristics of colon cancer cases diagnosed during 1987–2013 in ACTUR and SEER registries
| ACTUR (N=11,907) | SEER (N=23,814) | p-value | |||
|---|---|---|---|---|---|
| N | % | N | % | ||
| Age group | 1.00 | ||||
| 18–39 | 622 | 5.22 | 1244 | 5.22 | |
| 40–49 | 1127 | 9.47 | 2254 | 9.47 | |
| 50–64 | 4793 | 40.25 | 9586 | 40.25 | |
| 65–74 | 3228 | 27.11 | 6456 | 27.11 | |
| 75 or older | 2137 | 17.95 | 4274 | 17.95 | |
| Sex | 1.00 | ||||
| Male | 7316 | 61.44 | 14632 | 61.44 | |
| Female | 4591 | 38.56 | 9182 | 38.56 | |
| Race | 1.00 | ||||
| White | 9505 | 79.83 | 19010 | 79.83 | |
| Black | 1729 | 14.52 | 3458 | 14.52 | |
| Asian or Pacific Islander | 673 | 5.65 | 1346 | 5.65 | |
| Year of diagnosis | 1.00 | ||||
| 1987–1989 | 1484 | 12.46 | 2968 | 12.46 | |
| 1990–1994 | 3201 | 26.88 | 6414 | 26.88 | |
| 1995–1999 | 2566 | 21.55 | 5132 | 21.55 | |
| 2000–2004 | 2116 | 17.77 | 4238 | 17.77 | |
| 2005–2009 | 1597 | 13.41 | 3204 | 13.41 | |
| 2010–2013 | 943 | 7.92 | 1888 | 7.92 | |
| Region of Diagnosis | <0.001 | ||||
| Northeast | 267 | 2.24 | 3585 | 15.05 | |
| South | 6593 | 55.37 | 3424 | 14.38 | |
| Midwest | 918 | 7.71 | 5007 | 21.03 | |
| West | 3699 | 31.07 | 11798 | 49.54 | |
| Other | 430 | 3.61 | 0 | 0.00 | |
| Tumor stage | <0.001 | ||||
| Stage I | 2699 | 22.67 | 4438 | 18.64 | |
| Stage II | 2839 | 23.84 | 6156 | 25.85 | |
| Stage III | 3066 | 25.75 | 5943 | 24.96 | |
| Stage IV | 2231 | 18.74 | 5152 | 21.63 | |
| Unknown | 1072 | 9.00 | 2125 | 8.92 | |
| Tumor grade | <0.001 | ||||
| Well differentiated, grade1 | 1829 | 15.36 | 2312 | 9.71 | |
| Moderately differentiated, grade 2 | 7128 | 59.86 | 14298 | 60.04 | |
| Poorly differentiated, grade 3 | 1731 | 14.54 | 4384 | 18.41 | |
| Undifferentiated, grade 4 | 35 | 0.29 | 227 | 0.95 | |
| Unknown | 1184 | 9.94 | 2742 | 10.89 | |
| Tumor location | <0.001 | ||||
| Right | 4620 | 38.80 | 9876 | 41.47 | |
| Transverse | 912 | 7.66 | 2003 | 8.41 | |
| Left | 5200 | 43.67 | 11177 | 46.93 | |
| Overlapping | 171 | 1.44 | 243 | 1.02 | |
| Unknown | 1004 | 8.43 | 515 | 2.16 | |
| Cancer-directed surgery | <0.001 | ||||
| Yes | 11063 | 92.91 | 21989 | 92.34 | |
| No | 649 | 5.45 | 1723 | 7.24 | |
| Unknown | 195 | 1.64 | 102 | 0.43 | |
The median follow-up times for ACTUR and SEER cases were 56 months and 49 months, respectively. Among deceased patients, the median follow-up times for ACTUR and SEER cases were 18 months and 14 months, respectively. The median follow-up time for non-deceased patients were 60 months in both patient populations. Kaplan-Meier survival analysis showed that ACTUR cases exhibited significantly better overall survival than SEER cases (log-rank p<0.0001) (Figure 1). In Cox regression model for matched data further adjusted for age (continuous), region at diagnosis, tumor characteristics and surgery, ACTUR cases had 18% lower morality risk than SEER cases. The HR was 0.82 (95% CI=0.79–0.87) (Table 2). In stratified analysis by matching variables, the lower mortality in ACTUR than SEER was observed in nearly all subgroups stratified by age group, gender, race (except Asian/Pacific Islander) and diagnosis year (except in 2010–2013) group. The reduced morality tended to be more evident in blacks than whites in all models, while the 95% CIs were overlapped between the two racial groups.
Figure 1.
Kaplan-Meier survival curve for ACTUR and SEER patients with colon cancer diagnosed during 1987–2013
Table 2.
Overall and stratified hazard ratios of all-cause mortality comparing ACTUR with SEER among colon cancer cases diagnosed between 1987 and 2013
| Variables | Numbers | Unadjusted HR (95% CI)a | Adjusted HR (95% CI)b | |
|---|---|---|---|---|
| All cases | Deaths | |||
| Overall | ||||
| SEER | 23814 | 10991 | 1.00 (ref.) | 1.00 (ref.) |
| ACTUR | 11907 | 4898 | 0.81 (0.78–0.84) | 0.82 (0.78–0.87) |
| By Age | ||||
| 18–39 | ||||
| SEER | 1244 | 504 | 1.00 (ref.) | 1.00 (ref.) |
| ACTUR | 622 | 221 | 0.81 (0.68–0.96) | 0.79 (0.60–1.03) |
| 40–49 | ||||
| SEER | 2254 | 870 | 1.00 (ref.) | 1.00 (ref.) |
| ACTUR | 1127 | 420 | 0.88 (0.78–1.00) | 0.78 (0.64–0.96) |
| 50–64 | ||||
| SEER | 9586 | 3970 | 1.00 (ref.) | 1.00 (ref.) |
| ACTUR | 4793 | 1726 | 0.79 (0.74–0.84) | 0.79 (0.72–0.87) |
| 65–74 | ||||
| SEER | 6456 | 3095 | 1.00 (ref.) | 1.00 (ref.) |
| ACTUR | 3228 | 1342 | 0.77 (0.72–0.83) | 0.83 (0.74–0.92) |
| 75 or older | ||||
| SEER | 4274 | 2552 | 1.00 (ref.) | 1.00 (ref.) |
| ACTUR | 2137 | 1189 | 0.86 (0.80–0.93) | 0.88 (0.79–0.98) |
| By Gender | ||||
| Male | ||||
| SEER | 14632 | 7032 | 1.00 (ref.) | 1.00 (ref.) |
| ACTUR | 7316 | 3154 | 0.81 (0.77–0.85) | 0.81 (0.76–0.86) |
| Female | ||||
| SEER | 9182 | 3959 | 1.00 (ref.) | 1.00 (ref.) |
| ACTUR | 4591 | 1744 | 0.81 (0.76–0.86) | 0.85 (0.77–0.93) |
| By Race | ||||
| White | ||||
| SEER | 19010 | 8783 | 1.00 (ref.) | 1.00 (ref.) |
| ACTUR | 9505 | 3961 | 0.82 (0.78–0.85) | 0.82 (0.77–0.87) |
| Black | ||||
| SEER | 3458 | 1730 | 1.00 (ref.) | 1.00 (ref.) |
| ACTUR | 1729 | 703 | 0.70 (0.64–0.78) | 0.74 (0.64–0.85) |
| Asian/Pacific Islander | ||||
| SEER | 1346 | 478 | 1.00 (ref.) | 1.00 (ref.) |
| ACTUR | 673 | 234 | 0.99 (0.83–1.17) | 1.04 (0.82–1.32) |
| By Diagnosis Year | ||||
| 1987–1989 | ||||
| SEER | 2968 | 1556 | 1.00 (ref.) | 1.00 (ref.) |
| ACTUR | 1484 | 660 | 0.76 (0.69–0.84) | 0.80 (0.69–0.94) |
| 1990–1994 | ||||
| SEER | 6402 | 3273 | 1.00 (ref.) | 1.00 (ref.) |
| ACTUR | 3201 | 1408 | 0.77 (0.72–0.83) | 0.82 (0.74–0.91) |
| 1995–1999 | ||||
| SEER | 5132 | 2568 | 1.00 (ref.) | 1.00 (ref.) |
| ACTUR | 2566 | 1115 | 0.77 (0.72–0.84) | 0.83 (0.73–0.94) |
| 2000–2004 | ||||
| SEER | 4232 | 1862 | 1.00 (ref.) | 1.00 (ref.) |
| ACTUR | 2116 | 860 | 0.87 (0.79–0.94) | 0.86 (0.76–0.97) |
| 2005–2009 | ||||
| SEER | 3194 | 1299 | 1.00 (ref.) | 1.00 (ref.) |
| ACTUR | 1597 | 594 | 0.86 (0.77–0.95) | 0.84 (0.72–0.98) |
| 2010–2013 | ||||
| SEER | 1886 | 433 | 1.00 (ref.) | 1.00 (ref.) |
| ACTUR | 943 | 261 | 1.08 (0.89–1.30) | 1.01 (0.77–1.33) |
All HRs were estimated from multivariable Cox proportional hazard model for matched data
HRs were adjusted for age as a continuous variable, region at diagnosis, tumor stage, tumor grade, tumor location and surgery
HR=Hazard ratio; CI=Confidence Interval
For comparisons of ACTUR with SEER by insurance type using the 2007–2013 data, relative to the ACTUR patients, the adjusted HRs were 0.94 (95% CI=0.79–1.12), 0.98 (95% CI=0.75–1.27), 1.31 (95% CI=1.05–1.63), 1.73 (95% CI=1.31–2.29), and 1.12 (0.65–1.95) for those with “insured”, “insurance/no specifics”, “any Medicaid”, “uninsured”, and “unknown insurance status” in SEER, respectively (Supplementary Table 1).
The comparison of tumor stage at diagnosis showed that ACTUR cases were significantly less likely to be diagnosed at a later stage with a PR of 0.90 (95% CI=0.85–0.94) adjusted for the potential confounders (Table 3). This association was significant in subgroups of age 50–64, 65–74, in whites and blacks, and in men (borderline significant in women). Table 4 shows that the significantly better survival of the ACTUR patients was observed for all tumor stages except unknown stage.
Table 3.
Overall and stratified prevalence ratios of stage at diagnosis comparing ACTUR with SEER among colon cases
| Variables | Tumor Stage | Adjusted PRa (95% CI) | |
|---|---|---|---|
| Stages I and II | Stages III and IV | ||
| Overall | |||
| SEER | 10594 | 11095 | 1.00 (ref.) |
| ACTUR | 5538 | 5297 | 0.90 (0.85–0.94) |
| By Age | |||
| 18–39 | |||
| SEER | 428 | 719 | 1.00 (ref.) |
| ACTUR | 208 | 373 | 1.02 (0.80-1.29) |
| 40–49 | |||
| SEER | 862 | 1237 | 1.00 (ref.) |
| ACTUR | 411 | 643 | 1.00 (0.85-1.19) |
| 50–64 | |||
| SEER | 4113 | 4619 | 1.00 (ref.) |
| ACTUR | 2189 | 2181 | 0.86 (0.79–0.93) |
| 65–74 | |||
| SEER | 2987 | 2825 | 1.00 (ref.) |
| ACTUR | 1612 | 1288 | 0.88 (0.79–0.98) |
| 75 or older | |||
| SEER | 2204 | 1695 | 1.00 (ref.) |
| ACTUR | 1118 | 812 | 0.93 (0.82–1.06) |
| By Gender | |||
| Male | |||
| SEER | 6515 | 6743 | 1.00 (ref.) |
| ACTUR | 3415 | 3206 | 0.87 (0.82–0.94) |
| Female | |||
| SEER | 4079 | 4352 | 1.00 (ref.) |
| ACTUR | 2123 | 2091 | 0.93 (0.86–1.02) |
| By Race | |||
| White | |||
| SEER | 8630 | 8537 | 1.00 (ref.) |
| ACTUR | 4518 | 4094 | 0.89 (0.84–0.95) |
| Black | |||
| SEER | 1361 | 1810 | 1.00 (ref.) |
| ACTUR | 730 | 871 | 0.85 (0.75–0.98) |
| Asian/Pacific Islander | |||
| SEER | 603 | 648 | 1.00 (ref.) |
| ACTUR | 290 | 332 | 1.05(0.85-1.31) |
In the overall analysis, PR was adjusted for continuous age, sex, race, year of diagnosis, region of diagnosis, and tumor location. In stratified analysis, all variables in the overall analysis were adjusted for the variables in the overall analysis except the stratification variable itself. PR=Prevalence ratio; CI=95% confidence interval
Table 4.
Stratified hazard ratios of all-cause mortality by tumor stage comparing ACTUR with SEER among colon cancer cases
| Variables | Numbers | Adjusted HRa (95% CI) | |
|---|---|---|---|
| All Cases | Deaths | ||
| Stage I | |||
| SEER | 4438 | 818 | 1.00 (ref.) |
| ACTUR | 2699 | 473 | 0.89 (0.78–1.02) |
| Stage II | |||
| SEER | 6156 | 1811 | 1.00 (ref.) |
| ACTUR | 2839 | 743 | 0.83 (0.75–0.92) |
| Stage III | |||
| SEER | 5943 | 2733 | 1.00 (ref.) |
| ACTUR | 3066 | 1202 | 0.77 (0.72–0.84) |
| Stage IV | |||
| SEER | 5152 | 4662 | 1.00 (ref.) |
| ACTUR | 2231 | 1984 | 0.91 (0.85–0.97) |
| Unknown stage | |||
| SEER | 2125 | 967 | 1.00 (ref.) |
| ACTUR | 1072 | 496 | 0.98 (0.86–1.12) |
All HRs were estimated from multivariable Cox proportional hazard model with matching variables, age (as continuous variable), region of diagnosis, tumor grade, tumor location and surgery in the model. HR=Hazard ratio; CI=Confidence Interval
Discussion
In this study, we observed significant better survival among ACTUR patients with colon cancer than among those in the SEER population. In addition, the ACTUR cases were more likely than SEER cases to have an earlier stage at diagnosis.
To the best of our knowledge, the current study is the first one comparing survival of colon cancer patients between MHS and SEER. However, there was a study comparing survival of cancer patients in the Veteran Health Administration (VHA) system and U.S. general population, which reported better survival among older men with colon cancer in the VHA system than those in the general population. (22) The study did not present results, however, from subgroup analysis. In our study, the better survival in ACTUR than SEER tended to be more evident among black patients than white patients. This may reflect the reduction of a larger gap in care access among black persons because blacks have poorer access to care than whites in the general population, (23, 24) and blacks thus may benefit more than whites from the universal health care provided by the MHS. The survival benefit of blacks in our study suggests that universal health care system may be helpful to reduce racial disparity.
In keeping with our finding, the better survival of the ACTUR population than the SEER population was observed in our previous studies of non-small cell lung cancer(25) and invasive breast cancer.(26) The consistent results of improved survival outcome of the MHS beneficiaries suggest that the findings are not cancer-site specific, and that the MHS’s universal health care may have translated into the improved survival outcome across different cancer sites.
The better survival in ACTUR than SEER suggests the positive impact of better access to care in the universal healthcare system on cancer outcome. This was further demonstrated with comparisons of ACTUR to different types of insurance in SEER. Although follow-up time was short due to unavailability of the data prior to 2007, significant worse survival for no insurance or Medicaid but not for private insurance or other insurances than ACTUR supports the notion that better access to care and quality of care may benefit cancer outcome. The specific factors related to health care access contributing to the improved survival remain to be identified. Literature suggests various factors related to health care access and those beyond health care access may operate at the levels of patient, provider, and health care delivery system.(2, 4, 5, 8, 9, 11) These factors include out-of-pocket costs for medical services,(2, 5, 8) adherence to screening and treatment guidelines,(5, 8, 9, 11) cancer stage,(2, 8, 9) comorbidities, health belief and literacy and social support, (2, 5, 9) quality of care received,(4) healthcare coverage disruptions (27) and other health delivery factors.(8) Among the possible factors, two factors related to both access to care and survival are tumor stage at diagnosis and receipt of cancer-directed treatment. In the general population, earlier stage at diagnosis was found in patients with private insurance than those with Medicaid, or no insurance,(2, 5, 9) and in Medicare patients with a private FEE-For-Service (FFS) supplement compared to other Medicare plans.(28)
Consistent with our result of an earlier tumor stage at diagnosis among MHS beneficiaries than cases in the general population, two studies reported an earlier stage at diagnosis of colon cancer among older patients (65 or older) in the VHA system, which provides care to veterans probably due to improved preventive care of VHA.(22, 29) Data from the TRICARE (the Department of Defense’s health insurance program) report to Congress showed that colorectal cancer screening rates in MHS exceeded the national standards at the 90th percentile continuously over multiple periods of time.(12) In addition, a study reported higher colorectal screening rate (71%) than the national average rate (50%−60%) in a TRICARE population aged 50 or older.(30) The higher utilization of cancer screening services of MHS beneficiaries may have contributed to earlier stage at diagnosis in our data, as we showed that earlier stage at diagnosis started to appear at age 50, the age recommended for colorectal cancer screening.
It is noteworthy that the better survival of MHS cases was observed across all cancer stages. These findings imply that factors other than tumor stage may play a role in the survival differences between the two populations. Although specific factors are yet to be identified, it is possible that, colon cancer treatments, treatment combinations and modalities (10, 11, 31), some of them were not available in the SEER data, might be different between ACTUR and SEER patients. This might partially account for the identified differences in survival. Second, generally better care might contribute to the differences, such as compliance with the treatment guidelines, time between diagnosis and treatment, and so on. Better survival on other cancers as shown in our previous study on lung cancer (25) and breast cancer (26) supports the possibility of better general care across cancer sites in the MHS.
The large numbers of colon cancer cases from the ACTUR and SEER cancer registries allowed stratified analysis by demographic and tumor characteristics in our study. However, due to the limitation or incompleteness of cancer registry data, we were unable to study or control for the effects of some factors such as chemotherapy, quality of care delivery, provider factors, and patients’ ethnicity, socioeconomic status and comorbidities on the study outcomes. Regarding comorbidities, active-duty service members are healthier than the general population as they are required to pass military mandated fitness tests and medical examinations. However, among the ACTUR patients included in this study, 94% were non-active duty patients (family members of active duty, retirees and their family members), who are more similar to the general population. Thus, the potential impact of healthier active-duty members on the study findings might be limited, although we were unable to evaluate the impact directly. In addition, since all-cause mortality rather than cancer-specific death was used as the outcome, the potential effects from causes other than colon cancer on the survival difference cannot be excluded. Finally, for patients whose follow-up time was shorter than five years at the end of the study period, the follow-up time might not be long enough to observe the actual five-year survival.
In conclusion, MHS beneficiaries with colon cancer had better survival than their counterparts in the general population. The tendency of more survival benefit among blacks than whites may suggest the potential impact of universal health care on the reduction of racial disparity in survival among colon cancer patients. Future studies are warranted to identify factors contributing to the improved survival.
Supplementary Material
Acknowledgements
The authors thank the Joint Pathology Center and the Surveillance, Epidemiology, and End Results (SEER) program for the use of the cancer registry data.
Financial Support:
This project was supported by John P. Murtha Cancer Center Research Program, Uniformed Services University of the Health Sciences under the auspices of the Henry M. Jackson Foundation for the Advancement of Military Medicine and by the intramural research program of the National Cancer Institute.
Footnotes
Disclaimers:
The contents of this publication are the sole responsibility of the author(s) and do not necessarily reflect the views, opinions or policies of Uniformed Services University of the Health Sciences (USUHS), The Henry M. Jackson Foundation for the Advancement of Military Medicine, Inc., the Department of Defense (DoD) or the Departments of the Army, Navy, or Air Force. Mention of trade names, commercial products, or organizations does not imply endorsement by the U.S. Government.
Conflicts of Interest: The authors declare no potential conflicts of interest.
References
- 1.Siegel RL, Miller KD, Jemal A. Cancer statistics, 2018. CA Cancer J Clin 2018;68:7–30. [DOI] [PubMed] [Google Scholar]
- 2.Walker GV, Grant SR, Guadagnolo BA, Hoffman KE, Smith BD, Koshy M, et al. Disparities in stage at diagnosis, treatment, and survival in nonelderly adult patients with cancer according to insurance status. J Clin Oncol 2014; 32:3118–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Robbins AS, Pavluck AL, Fedewa SA, Chen AY, Ward EM. Insurance status, comorbidity level, and survival among colorectal cancer patients age 18 to 64 years in the National Cancer Data Base from 2003 to 2005. J Clin Oncol 2009;27:3627–33. [DOI] [PubMed] [Google Scholar]
- 4.Amini A, Yeh N, Jones BL, Bedrick E, Vinogradskiy Y, Rusthoven CG, et al. Perioperative Mortality in Nonelderly Adult Patients With Cancer: A Population-based Study Evaluating Health Care Disparities in the United States According to Insurance Status. Am J Clin Oncol 2018;41:476–84. [DOI] [PubMed] [Google Scholar]
- 5.Ward EM, Fedewa SA, Cokkinides V, Virgo K. The association of insurance and stage at diagnosis among patients aged 55 to 74 years in the national cancer database. Cancer J (Sudbury, Mass). 2010;16:614–21. [DOI] [PubMed] [Google Scholar]
- 6.Lee JM, Wang X, Ojha RP. The effect of health insurance on childhood cancer survival in the United States. Cancer 2017;123:4878–85. [DOI] [PubMed] [Google Scholar]
- 7.Aizer AA, Falit B, Mendu ML, Chen MH, Choueiri TK, Hoffman KE, et al. Cancer-specific outcomes among young adults without health insurance. J Clin Oncol 2014;32:2025–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Kirsner RS, Ma F, Fleming L, Federman DG, Trapido E, Duncan R, et al. The effect of medicare health care delivery systems on survival for patients with breast and colorectal cancer. Cancer Epidemiol Biomarkers Prev 2006;15:769–73. [DOI] [PubMed] [Google Scholar]
- 9.Halpern MT, Ward EM, Pavluck AL, Schrag NM, Bian J, Chen AY. Association of insurance status and ethnicity with cancer stage at diagnosis for 12 cancer sites: a retrospective analysis. Lancet Oncol 2008; 9:222–31. [DOI] [PubMed] [Google Scholar]
- 10.Chagpar R, Xing Y, Chiang YJ, Feig BW, Chang GJ, You YN, et al. Adherence to stage-specific treatment guidelines for patients with colon cancer. J Clin Oncol 2012;30:972–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Boland GM, Chang GJ, Haynes AB, Chiang YJ, Chagpar R, Xing Y, et al. Association between adherence to National Comprehensive Cancer Network treatment guidelines and improved survival in patients with colon cancer. Cancer 2013;119:1593–601. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.The Evaluation of the TRICARE Program: Fiscal Year 2019 Report to Congress: https://health.mil/Military-Health-Topics/Access-Cost-Quality-and-Safety/Health-Care-Program-Evaluation/Annual-Evaluation-of-the-TRICARE-Program; Accessed July 24, 2019.
- 13.Cancer Program Standards 2012: Ensuraing Patient-Centered Care. https://www.facs.org/~/media/files/quality%20programs/cancer/coc/programstandards2012.ashx. AccessedJanuary 16, 2020.
- 14.Tryo J User’s Guide for ACTUR cancer registry software system abstacting module. 2007.
- 15.National Cancer Institute SEER Taining Models. Cancer Patients Follo-up: Requirements of CoC/Governing Agencies. https://training.seer.cancer.gov/followup/intro/requirements.html. AccessedJanuary 16, 2020.;
- 16.SEER*Stat Databases: November 2015 Submission. Case Listing and Frequency Sessions: https://seer.cancer.gov/data-software/documentation/seerstat/nov2015/Surveillance, Epidemiology, and End Results Program. AccessedJuly 31, 2019. [Google Scholar]
- 17.International Classification of Diseases for Oncology, 3rd Edition(ICD-O-3). World Health Organization. https://www.who.int/classifications/icd/adaptations/oncology/en/; AccessedJuly 31, 2019. [Google Scholar]
- 18.American Joint Committee on Cancer. AJCC Cancer Staging Manual, sixth Edition. Greene FL, Page DL, Flemming ID, et al. (eds.) Chicago, IL: Springer; 2002. [Google Scholar]
- 19.National Comprehensive Cancer Network: NCCN Clinical Practice Guidelines in Oncology: Colon Cancer. https://www.nccn.org/professionals/physician_gls/pdf/colon.pdf; AccessedSpetember 10, 2019. [DOI] [PubMed]
- 20.SEER Progrm Coding and Staging Manual 2015: https://seer.cancer.gov/archive/manuals/2015/AppendixC/colon/surgery_codes.pdf; AccessedFebruary 11, 2019.
- 21.Lee EW, Wang JW. Statistical Methods for Survival Data Analysis, Third Edition. Hoboken, New Jersey: John Wiley & Sons, Inc., 2003. [Google Scholar]
- 22.Landrum MB, Keating NL, Lamont EB, Bozeman SR, Krasnow SH, Shulman L, et al. Survival of older patients with cancer in the Veterans Health Administration versus fee-for-service Medicare. J Clin Oncol 2012;30:1072–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Shavers VL, Brown ML. Racial and ethnic disparities in the receipt of cancer treatment. J Natl Cancer Inst 2002;94:334–57. [DOI] [PubMed] [Google Scholar]
- 24.DeSantis CE, Siegel RL, Sauer AG, Miller KD, Fedewa SA, Alcaraz KI, et al. Cancer statistics for African Americans, 2016: Progress and opportunities in reducing racial disparities. CA Cancer J Clin 2016;66:290–308. [DOI] [PubMed] [Google Scholar]
- 25.Lin J, Kamamia C, Brown D, Shao S, McGlynn KA, Nations JA, et al. Survival among Lung Cancer Patients in the U.S. Military Health System: A Comparison with the SEER Population. Cancer Epidemiol Biomarkers & Prev 2018; 27:673–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Ru Y, Liu J, Fantacone-Campbell JL, Zhu K, Kovatich AJ, Hooke JA, et al. Comparative Survival Analysis of Invasive Breast Cancer Patients Treated by a U.S. Military Medical Center and Matched Patients From the U.S. General Population. Mil Med 2017;182:e1851–e8. [DOI] [PubMed] [Google Scholar]
- 27.Yabroff KR, Reeder-Hayes K, Zhao J, Halpern MT, Lopez AM, Bernal-Mizrachi L, et al. Health Insurance Coverage Disruptions and Cancer Care and Outcomes: Systematic Review of Published Research. J Natl Cancer Inst. 2020;112:671–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Lee-Feldstein A, Feldstein PJ, Buchmueller T. Health care factors related to stage at diagnosis and survival among Medicare patients with colorectal cancer. Med Care 2002;40:362–74. [DOI] [PubMed] [Google Scholar]
- 29.Keating NL, Landrum MB, Lamont EB, Bozeman SR, Krasnow SH, Shulman LN, et al. Quality of care for older patients with cancer in the Veterans Health Administration versus the private sector: a cohort study. Ann Intern Med 2011;154:727–36. [DOI] [PubMed] [Google Scholar]
- 30.Brounts LR, Lehmann RK, Lesperance KE, Brown TA, Steele SR. Improved rates of colorectal cancer screening in an equal access population. Am J Surg 2009;197:609–12; discussion 12–3. [DOI] [PubMed] [Google Scholar]
- 31.Casadaban L, Rauscher G, Aklilu M, Villenes D, Freels S, Maker AV. Adjuvant chemotherapy is associated with improved survival in patients with stage II colon cancer. Cancer 2016;122:3277–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.