Abstract
Background
Annual surveillance mammography is recommended after breast cancer diagnosis. Previous studies have suggested that surveillance mammography varies by demographics and initial tumor characteristics, which are related to access to healthcare. The Department of Defense's Military Health System (MHS) provides beneficiaries with equal healthcare access and thus offers an excellent opportunity to assess whether racial differences in surveillance mammography persist when access to care is equal.
Methods
Among female beneficiaries with a history of breast cancer logistic regression was used to assess racial/ethnic variation in surveillance mammography during three 12-month periods, beginning one year after diagnosis adjusting for demographic, tumor and health characteristics.
Results
Overall surveillance mammography decreased from 70% during the first year to 59% during the third year (p<0.01). Although there was an overall tendency for surveillance mammography to be higher among minority women compared to non-Hispanic white women, after adjusting for covariates, the difference was significant only during the first year among blacks (odds ratio (OR)=1.46; 95% Confidence Interval (CI)=1.10-1.95) and the second year among Asian Pacific Islanders (OR=2.29; 95%CI=1.52-3.44) and Hispanics (OR=1.92; 95%CI=1.17-3.18). When stratified by age at diagnosis and type of breast cancer surgery, significant racial differences tended to be observed among younger women (<50 years) and only among women who had mastectomies.
Conclusion
Minority women were equally or more likely than non-Hispanic white women to receive surveillance mammography within the MHS. The racial disparities in surveillance mammography reported in other studies were not observed in a system with equal access to care.
Keywords: breast cancer, mammography, surveillance, survivor, epidemiology, healthcare access
Introduction
Although female breast cancer is the most common cancer among U.S. women and the second most common cause of cancer death,1 the overall five-year relative survival rate is high (89%).2 Thus, there are more than 2.7 million women alive in the U.S. with a history of breast cancer,3 all of whom are at risk for recurrence and an increased risk for new primary tumors and ultimately breast cancer mortality.4-8
Breast cancer mortality rates vary by race/ethnicity. Notably, overall age-adjusted mortality rates are higher among black women than white women, even though the reverse is true for overall age-adjusted incidence rates.1 Additionally, Asian/Pacific Islander (API) women and Hispanic women have lower mortality rates, as well as incidence rates, than non-Hispanic white women.9 Racial/ethnic variations in mortality rates are likely due to multiple factors possibly including differences in follow-up care.
Mammography is an effective screening tool to diagnose breast cancer at early stages and is associated with lower breast cancer mortality rates.10 Although the American Society for Clinical Oncology (ASCO) recommends annual surveillance mammography for all women with a history of breast cancer,11 studies have indicated that a sizable proportion of these women do not receive adequate screening, that usage decreases significantly with time since diagnosis and that many factors, including demographics, tumor characteristics and cancer treatment, affect usage.12-17 Some of these studies14, 16, 17 have found lower usage among minority women than white/non-Hispanic white women. However, differential access to care by race/ethnicity in these previous studies may have confounded the racial comparisons. Additionally, many of these previous studies were conducted among Medicare beneficiaries12, 14, 16 or at least among predominately post-menopausal women13 and thus, may not be generalizable to younger women.
The Department of Defense's (DoD's) Military Healthcare System (MHS) provides a unique environment to conduct disparities research given that all DoD beneficiaries are provided equal healthcare access. Additionally, because the MHS is not restricted to women of a certain age, conducting research among its beneficiaries may provide insights into the behaviors of young women after breast cancer diagnosis. The objectives of this study were two-fold: first, to examine if annual surveillance mammography varied by race/ethnicity after adjustment for covariates among female DoD beneficiaries of any age with a history of breast cancer and second, to describe how annual surveillance mammography varied with time since diagnosis.
Methods
Linked data from the DoD's Central Cancer Registry (CCR) and the MHS Data Repository (MDR), the DoD's medical claims database, were used for this study. The CCR contains information for all DoD cancer patients diagnosed or treated at military treatment facilities (MTFs), including active-duty and retired military personnel and their dependents. Duplicate records pertaining to the same tumor were consolidated following North America Association of Central Cancer Registries guidelines. The MDR includes administrative and medical claims information from the DoD heath care program, known as Tricare, beneficiaries, including direct care received at military treatment facilities (MTFs) or indirect care received at non-MTFs that is paid for by the DoD. The MDR database includes information on clinical diagnoses of all medical conditions, which are coded using the International Classification of Diseases – Ninth Revision (ICD-9), and diagnostic and treatment procedures, which are coded using ICD-9, Current Procedural Terminology (CPT) or Healthcare Common Procedure Coding System (HCPCS) codes. The current study began in 2005 as radiology records, which were the primary source of mammography documentation, became available that year.
Women were eligible for the study if they had surgically treated, histologically confirmed, first primary, malignant breast cancer diagnosed between 2005 and 2007 (the most recent years of the linked data). The racial/ethnic groups compared were non-Hispanic white, non-Hispanic black (black), non-Hispanic Asian Pacific Islander (API) and Hispanic (regardless of race). Annual surveillance mammography usage was assessed by race/ethnicity for three 12-month periods that began one year after diagnosis. Analyses for each year were restricted to women who had complete follow-up data for that year. Women were excluded if they left the MHS (year 1 n=17; year 2 n=455; year 3 n=595) or died (year 1 n=60; year 2 n=41; year 3 n=22) prior to or during each assessment year. Annual surveillance mammography is not recommended for women who have had bilateral mastectomies; therefore women were also excluded if there was indication that they had bilateral mastectomies (concurrently or sequentially) prior to each assessment year (year 1 n= 133; year 2 n=22; year 3 n=12).
To minimize the possibility of counting diagnostic mammograms, a mammogram was considered to have been for surveillance if there was a recorded bilateral mammogram in MDR among women who had undergone breast conserving surgery (BCS) (CPT: 76091, 76092, 77056, 77057; HCPCS: G0202, G0203; ICD-9: V76.10, V76.11, V72.12) or a unilateral mammogram (CPT: 76090, 77055) among women who had undergone unilateral mastectomies and no diagnosis within the two preceding months of a breast mass or other breast symptom diagnoses (ICD-9: 611.72, 611.79). Additionally, the identified mammogram had to be conducted at least 6 months after a previous mammogram.
Hormone receptor status was considered positive if either estrogen receptor (ER) or progesterone receptor (PR) status was recorded as positive, negative if both ER and PR were negative, and unknown if neither measure was recorded. Comorbidities were considered to be present if a diagnosis was recorded in the MDR during the 12 months prior to each assessment year. To minimize the possibility of false comorbidity diagnoses, codes had to be recorded in the outpatient datasets at least three times. The level of comorbidity present was categorized according to the Charlson Comorbidity Index18 and breast cancer diagnoses were excluded from the calculation.
Chi-square tests were used to compare demographics, initial breast cancer tumor characteristics and cancer treatments between women who did and did not undergo surveillance mammography during each assessment year and to compare annual surveillance mammography usage across the three years. Variables assessed included age at diagnosis; year of diagnosis; active duty status; beneficiary type (Tricare prime: HMO component/Tricare non-prime/unknown); marital status (married/not married: never married, separated, divorced/unknown); service branch; sponsor's service rank; tumor characteristics of the initial breast cancer (tumor stage, tumor grade, hormone receptor status); cancer treatments (surgery/radiation and chemotherapy) and comorbidity index. Logistic regression was used to assess receipt of annual surveillance mammography by race/ethnicity after adjustment for potential confounding by variables that were found to be significant for any year during univariate analysis, separately for the three years. Effect modification by age (<50 vs. >50 years) and breast cancer surgery type were assessed by creating stratified regression models. All statistical analyses were performed using SAS (version 9.3; SAS Institute Inc, Cary, NC) and the two-sided significance level was set at p<0.05.
Approval for this project was approved by the institutional review boards of the Walter Reed National Military Medical Center, Tricare Management Activity, and the National Institutes of Health Office of Human Subjects Research.
Results
Overall annual surveillance mammography decreased from 70% during the first year to 59% during the third year (p<0.01; Table 1). When stratified by race/ethnicity the largest temporal decrease (−15%; p<0.01) was observed among black women and the smallest temporal decrease (−7%; p>0.05) was observed among API women.
Table 1.
Comparison of characteristics by receipt of annual mammography during three consecutive years that started 12 months post-cancer diagnosis, among women diagnosed with breast cancer between 2005 and 2007 in the Military Health System.
| Year 1 (n=1951) Mammography |
Year 2 (n=1433) Mammography |
Year 3 (n=823) Mammography |
||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| No | Yes | No | Yes | No | Yes | |||||||||||
| Characteristic1 | N | % | N | % | p2 | N | % | N | % | p2 | N | % | N | % | p2 | |
| All | 589 | 30% | 1362 | 70% | 464 | 32% | 969 | 68% | 334 | 41% | 489 | 59% | ||||
| Race/ethnicity | Non-Hispanic White | 385 | 32% | 834 | 68% | 0.27 | 311 | 35% | 582 | 65% | <0.01 | 213 | 41% | 304 | 59% | 0.62 |
| Non-Hispanic Black | 94 | 27% | 258 | 73% | 94 | 35% | 176 | 65% | 65 | 42% | 88 | 58% | ||||
| Non-Hispanic Asian Pacific Islander | 68 | 28% | 177 | 72% | 35 | 20% | 139 | 80% | 35 | 35% | 66 | 65% | ||||
| Hispanic | 42 | 31% | 93 | 69% | 24 | 25% | 72 | 75% | 21 | 40% | 31 | 60% | ||||
| Age | 19-39 | 66 | 40% | 98 | 60% | <0.01 | 50 | 40% | 74 | 60% | <0.01 | 34 | 45% | 41 | 55% | 0.01 |
| 40-49 | 171 | 35% | 311 | 65% | 138 | 38% | 223 | 62% | 99 | 47% | 110 | 53% | ||||
| 50-59 | 161 | 30% | 378 | 70% | 118 | 30% | 278 | 70% | 81 | 35% | 149 | 65% | ||||
| 60-69 | 112 | 23% | 365 | 77% | 103 | 29% | 253 | 71% | 83 | 40% | 126 | 60% | ||||
| 70-79 | 56 | 27% | 155 | 73% | 37 | 25% | 109 | 75% | 21 | 29% | 51 | 71% | ||||
| 80-96 | 23 | 29% | 55 | 71% | 18 | 36% | 32 | 64% | 16 | 57% | 12 | 43% | ||||
| Marital Status | Married | 476 | 30% | 1112 | 70% | 0.37 | 365 | 31% | 800 | 69% | 0.20 | 270 | 39% | 416 | 61% | 0.27 |
| Not married | 103 | 30% | 237 | 70% | 95 | 37% | 161 | 63% | 61 | 47% | 69 | 53% | ||||
| Unknown | 10 | 43% | 13 | 57% | 4 | 33% | 8 | 67% | 3 | 43% | 4 | 57% | ||||
| Duty Status* | Non-Active Duty | 545 | 29% | 1308 | 71% | <0.01 | 431 | 32% | 930 | 68% | 0.01 | 317 | 41% | 462 | 59% | 0.79 |
| Active Duty | 44 | 45% | 54 | 55% | 33 | 46% | 39 | 54% | 17 | 39% | 27 | 61% | ||||
| Rank* | Enlisted | 394 | 29% | 948 | 71% | <0.01 | 320 | 32% | 684 | 68% | 0.01 | 229 | 40% | 347 | 60% | 0.03 |
| Officer | 174 | 30% | 405 | 70% | 130 | 32% | 276 | 68% | 94 | 41% | 138 | 59% | ||||
| Other/Unknown | 21 | 70% | 9 | 30% | 14 | 61% | 9 | 39% | 11 | 73% | 4 | 27% | ||||
| Service Branch | Army | 246 | 32% | 520 | 68% | 0.15 | 185 | 34% | 367 | 66% | 0.65 | 111 | 38% | 181 | 62% | 0.56 |
| Air Force | 178 | 30% | 406 | 70% | 144 | 34% | 283 | 66% | 108 | 42% | 152 | 58% | ||||
| Navy | 97 | 25% | 290 | 75% | 82 | 29% | 201 | 71% | 67 | 40% | 101 | 60% | ||||
| Other | 59 | 31% | 130 | 69% | 46 | 32% | 100 | 68% | 44 | 48% | 48 | 52% | ||||
| Unknown | 9 | 36% | 16 | 64% | 7 | 28% | 18 | 72% | 4 | 36% | 7 | 64% | ||||
| Beneficiary Plan | Prime “HMO” | 444 | 30% | 1028 | 70% | <0.01 | 349 | 32% | 741 | 68% | 0.54 | 259 | 41% | 366 | 59% | 0.65 |
| Not Prime | 123 | 28% | 317 | 72% | 104 | 33% | 213 | 67% | 68 | 38% | 113 | 62% | ||||
| Unknown | 22 | 56% | 17 | 44% | 11 | 42% | 15 | 58% | 7 | 41% | 10 | 59% | ||||
| Diagnosis Year | 2005 | 199 | 28% | 511 | 72% | 0.26 | 215 | 32% | 464 | 68% | 0.84 | 267 | 40% | 393 | 60% | 0.88 |
| 2006 | 198 | 32% | 420 | 68% | 194 | 33% | 397 | 67% | 67 | 41% | 96 | 59% | ||||
| 2007 | 192 | 31% | 431 | 69% | 55 | 34% | 108 | 66% | 0 | 0% | 0 | 0% | ||||
| Tumor Stage | I | 272 | 27% | 729 | 73% | <0.01 | 224 | 30% | 526 | 70% | 0.10 | 166 | 40% | 253 | 60% | 0.82 |
| IIA | 152 | 30% | 351 | 70% | 122 | 33% | 248 | 67% | 89 | 40% | 132 | 60% | ||||
| IIB | 74 | 36% | 129 | 64% | 54 | 37% | 92 | 63% | 40 | 45% | 49 | 55% | ||||
| III | 91 | 37% | 153 | 63% | 64 | 38% | 103 | 62% | 39 | 41% | 55 | 59% | ||||
| Tumor Grade | I | 128 | 28% | 322 | 72% | 0.45 | 103 | 30% | 235 | 70% | 0.79 | 81 | 41% | 119 | 60% | 0.90 |
| II | 225 | 29% | 544 | 71% | 181 | 32% | 377 | 68% | 127 | 39% | 197 | 61% | ||||
| III-IV | 207 | 33% | 429 | 67% | 157 | 34% | 307 | 66% | 108 | 42% | 149 | 58% | ||||
| Unknown | 29 | 30% | 67 | 70% | 23 | 32% | 50 | 68% | 18 | 43% | 24 | 57% | ||||
| Hormone Receptor | ER+ and/or PR+ | 424 | 29% | 1035 | 71% | 0.17 | 335 | 31% | 736 | 69% | 0.13 | 252 | 41% | 362 | 59% | 0.85 |
| ER- and PR- | 141 | 33% | 282 | 67% | 106 | 34% | 203 | 66% | 71 | 39% | 112 | 61% | ||||
| Unknown | 24 | 35% | 45 | 65% | 23 | 43% | 30 | 57% | 11 | 42% | 15 | 58% | ||||
| Surgery/Radiation | BCS | 30 | 31% | 68 | 69% | <0.01 | 29 | 45% | 36 | 55% | <0.01 | 17 | 44% | 22 | 56% | <0.01 |
| BCS + Radiation | 204 | 22% | 737 | 78% | 192 | 27% | 531 | 73% | 145 | 35% | 264 | 65% | ||||
| Mastectomy | 233 | 40% | 352 | 60% | 158 | 39% | 247 | 61% | 116 | 49% | 119 | 51% | ||||
| Mastectomy + Radiation | 122 | 37% | 205 | 63% | 85 | 35% | 155 | 65% | 56 | 40% | 84 | 60% | ||||
| Chemotherapy | Yes | 388 | 33% | 795 | 67% | <0.01 | 300 | 34% | 570 | 66% | 0.03 | 205 | 53% | 181 | 47% | 0.64 |
| No | 201 | 26% | 567 | 74% | 164 | 29% | 399 | 71% | 129 | 30% | 308 | 70% | ||||
| Comorbidity Index | 0 | 398 | 30% | 928 | 70% | 0.78 | 331 | 32% | 699 | 68% | 0.41 | 233 | 40% | 356 | 60% | 0.07 |
| 1 | 121 | 32% | 263 | 68% | 81 | 31% | 182 | 69% | 54 | 38% | 89 | 62% | ||||
| >1 | 70 | 29% | 171 | 71% | 52 | 37% | 88 | 63% | 47 | 52% | 44 | 48% | ||||
BCS: breast conserving surgery; ER: estrogen receptor; HMO: health management organization; PR: progesterone receptor
Sponsor's duty status and service rank, if the patient was a dependent.
At diagnosis unless otherwise stated.
Chi-Square test.
3 During the first year post-diagnosis.
4 During the preceding 12 months.
During all three years there was a tendency for annual surveillance mammography to be higher among minority women compared to non-Hispanic white women. During the first year, the receipt of annual surveillance mammography was significantly higher among black women than non-Hispanic white women (odds ratio (OR)=1.46; 95% Confidence Interval (CI)=1.10-1.95; Table 2). During the second year, the receipt of annual surveillance mammography was significantly higher only among API women (OR=2.29; 95% CI=1.52-3.44) and Hispanic women (OR=1.92; 95% CI=1.17-3.18) in comparison with non-Hispanic white women. During the third year, no significant racial/ethnic differences were observed.
Table 2.
Adjusted odds of annual surveillance mammography during three consecutive years that started 12 months post-cancer diagnosis, among women diagnosed with breast cancer between 2005 and 2007 in the Military Health System
| Year 1 |
Year 2 |
Year 3 |
||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Characteristic1 | OR2 | 95% CI | OR2 | 95% CI | OR2 | 95% CI | ||||
| Race/ethnicity | Non-Hispanic White | 1.00 | ref | 1.00 | ref | 1.00 | ref | |||
| Non-Hispanic Black | 1.46 | 1.10 | 1.95 | 1.18 | 0.87 | 1.60 | 0.91 | 0.62 | 1.35 | |
| Non-Hispanic Asian Pacific Islander | 1.33 | 0.96 | 1.83 | 2.29 | 1.52 | 3.44 | 1.29 | 0.81 | 2.05 | |
| Hispanic | 1.19 | 0.79 | 1.78 | 1.92 | 1.17 | 3.18 | 1.05 | 0.57 | 1.92 | |
| Age | 19-39 | 0.96 | 0.65 | 1.40 | 1.01 | 0.65 | 1.57 | 1.08 | 0.62 | 1.88 |
| 40-49 | 1.00 | ref | 1.00 | ref | 1.00 | ref | ||||
| 50-59 | 1.28 | 0.97 | 1.68 | 1.38 | 1.01 | 1.89 | 1.70 | 1.14 | 2.53 | |
| 60-69 | 1.93 | 1.39 | 2.67 | 1.78 | 1.23 | 2.57 | 1.51 | 0.96 | 2.36 | |
| 70-79 | 2.10 | 1.23 | 3.59 | 3.38 | 1.81 | 6.31 | 2.32 | 1.05 | 5.10 | |
| 80-96 | 2.00 | 1.01 | 3.96 | 2.48 | 1.12 | 5.48 | 0.83 | 0.31 | 2.26 | |
| Duty Status* | Non-Active Duty | 1.00 | ref | 1.00 | ref | 1.00 | ref | |||
| Active Duty | 0.62 | 0.40 | 0.97 | 0.71 | 0.43 | 1.19 | 1.32 | 0.67 | 2.59 | |
| Rank* | Enlisted | 1.00 | ref | 1.00 | ref | 1.00 | ref | |||
| Officer | 1.00 | 0.80 | 1.26 | 1.06 | 0.81 | 1.38 | 0.97 | 0.70 | 1.36 | |
| Other/unknown | 0.23 | 0.10 | 0.54 | 0.36 | 0.14 | 0.90 | 0.23 | 0.07 | 0.77 | |
| Beneficiary Plan | Prime “HMO” | 1.00 | ref | 1.00 | ref | 1.00 | ref | |||
| Not Prime | 0.75 | 0.51 | 1.11 | 0.57 | 0.37 | 0.87 | 1.17 | 0.70 | 1.96 | |
| Unknown | 0.37 | 0.18 | 0.77 | 0.73 | 0.30 | 1.81 | 1.24 | 0.41 | 3.74 | |
| Tumor Stage | I | 1.00 | ref | 1.00 | ref | 1.00 | ref | |||
| IIA | 1.06 | 0.81 | 1.39 | 0.99 | 0.73 | 1.35 | 1.00 | 0.68 | 1.47 | |
| IIB | 0.82 | 0.57 | 1.18 | 0.79 | 0.52 | 1.21 | 0.73 | 0.43 | 1.25 | |
| III | 0.82 | 0.55 | 1.21 | 0.71 | 0.45 | 1.13 | 0.73 | 0.41 | 1.31 | |
| Surgery/Radiation3 | BCS + Radiation | 1.00 | ref | 1.00 | ref | 1.00 | ref | |||
| BCS | 0.56 | 0.34 | 0.91 | 0.37 | 0.21 | 0.64 | 0.71 | 0.34 | 1.45 | |
| Mastectomy | 0.40 | 0.32 | 0.51 | 0.52 | 0.40 | 0.68 | 0.56 | 0.40 | 0.79 | |
| Mastectomy + Radiation | 0.58 | 0.41 | 0.81 | 0.84 | 0.57 | 1.24 | 0.95 | 0.59 | 1.54 | |
| Chemotherapy3 | Yes | 1.00 | ref | 1.00 | ref | 1.00 | ref | |||
| No | 1.10 | 0.84 | 1.44 | 1.10 | 0.81 | 1.50 | 0.76 | 0.52 | 1.12 | |
BCS: breast conserving surgery; CI: Confidence Interval; HMO: health management organization; OR: odds ratio
Sponsor's duty status and service rank, if the patient was a dependent.
At diagnosis unless otherwise stated.
Adjusted for all variables listed.
During the first year post-diagnosis.
In addition to the observed racial/ethnic variations, annual surveillance mammography varied by covariates (Tables 2). The receipt of surveillance mammography tended to increase with age. In comparison to women aged 40-49 years at diagnosis, older women were more likely to have received annual surveillance mammography with the highest levels appearing to be obtained by those who were aged 70-79 years at diagnosis. In comparison to patients who had an enlisted sponsor, those who had sponsors of ‘other/unknown’ rank were less likely to receive annual surveillance mammography (OR: 0.23-0.36). The combination of surgery type and radiation appeared to be associated with receipt of annual surveillance mammography. For example, during the first assessment year, women who had BCS without radiotherapy and unilateral mastectomies with or without radiotherapy were less likely (OR: 0.40-0.58) to receive annual surveillance mammography than women who had received BCS with radiotherapy. Finally, depending on the assessment year, there were indications that receipt of annual surveillance mammography varied by sponsor's duty status and beneficiary plan.
Effect modification of the relationship between race/ethnicity and surveillance mammography was assessed by age and type of breast cancer surgery (Table 3). When stratified by age, racial/ethnic variation tended to be observed among young women (<50 years). During the first year young black and API women (OR: 2.20-3.11) were more likely than their non-Hispanic white counterparts to receive a surveillance mammography. During the second year young API and Hispanic women (OR: 2.44-5.31) were also more likely than their non-Hispanic white counterparts to receive a surveillance mammography. An increased likelihood of surveillance mammography among older API women (OR:1.61), in comparison to non-Hispanic white women, was also observed. When stratified by type of breast cancer surgery, significant racial/ethnic variation was only observed among women who received mastectomies and the tendency was for minority women to be more likely to receive surveillance mammography than non-Hispanic white women.
Table 3.
Racial/ethnic variation in annual surveillance mammography by age and surgery type among women diagnosed with breast cancer between 2005 and 2007 in the Military Health System.
| Period |
||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Stratified variable | Race | OR1 | 95% CI | OR1 | 95% CI | OR1 | 95% CI | |||
| Age | ||||||||||
| <50 years | Non-Hispanic White | 1.00 | ref | 1.00 | ref | 1.00 | ref | |||
| Non-Hispanic Black | 2.20 | 1.39 | 3.46 | 1.08 | 0.68 | 1.72 | 0.87 | 0.47 | 1.62 | |
| Non-Hispanic Asian Pacific Islander | 3.11 | 1.67 | 5.79 | 5.31 | 2.26 | 12.50 | 2.11 | 0.90 | 4.95 | |
| Hispanic | 1.19 | 0.65 | 2.20 | 2.44 | 1.11 | 5.34 | 0.80 | 0.29 | 2.19 | |
| ≥50 years | Non-Hispanic White | 1.00 | ref | 1.00 | ref | 1.00 | ref | |||
| Non-Hispanic Black | 1.06 | 0.73 | 1.54 | 1.19 | 0.78 | 1.82 | 0.88 | 0.52 | 1.47 | |
| Non-Hispanic Asian Pacific Islander | 0.89 | 0.61 | 1.29 | 1.61 | 1.01 | 2.57 | 1.04 | 0.60 | 1.81 | |
| Hispanic | 1.27 | 0.72 | 2.24 | 1.53 | 0.80 | 2.94 | 1.21 | 0.55 | 2.64 | |
| Surgery | ||||||||||
| Mastectomy | Non-Hispanic White | 1.00 | ref | 1.00 | ref | 1.00 | ref | |||
| Non-Hispanic Black | 1.93 | 1.28 | 2.89 | 1.53 | 0.96 | 2.43 | 1.10 | 0.60 | 2.04 | |
| Non-Hispanic Asian Pacific Islander | 1.52 | 1.00 | 2.29 | 2.74 | 1.56 | 4.78 | 2.45 | 1.19 | 5.05 | |
| Hispanic | 1.94 | 1.10 | 3.42 | 2.35 | 1.20 | 4.60 | 1.33 | 0.58 | 3.05 | |
| Breast Conserving Surgery | Non-Hispanic White | 1.00 | ref | 1.00 | ref | 1.00 | ref | |||
| Non-Hispanic Black | 1.06 | 0.70 | 1.59 | 0.86 | 0.57 | 1.29 | 0.77 | 0.46 | 1.28 | |
| Non-Hispanic Asian Pacific Islander | 1.12 | 0.67 | 1.87 | 1.73 | 0.95 | 3.15 | 0.72 | 0.38 | 1.34 | |
| Hispanic | 0.72 | 0.41 | 1.28 | 1.55 | 0.72 | 3.31 | 0.85 | 0.35 | 2.09 | |
CI: Confidence Interval; OR: odds ratio.
Adjusted for age at diagnosis, active duty status at diagnosis, beneficiary plan type, rank of active duty member/sponsor, tumor stage, chemotherapy and surgery/radiation. Stratified variables were not included in stratified analysis.
Discussion
Overall surveillance mammography decreased from 70% during the first year to 59% during the third year (p<0.01). Although there was a tendency for surveillance mammography to be higher among minority women compared to non-Hispanic white women, after adjusting for covariates, the difference was significant only during the first year among black women and during the second year among API and Hispanic women. Furthermore, variation in the receipt of annual surveillance mammography by racial/ethnic tended to be confined to young women (<50 years) and those who underwent mastectomies. Surveillance mammography also appeared to vary by sponsor's duty status, sponsor's service rank and beneficiary plan.
In comparison to previous studies on surveillance mammography usage based on medical claims data and/or abstraction of medical records, the overall rate of surveillance mammography that was observed in this study was lower than some13-16 but higher than others.12, 17 However, due to differences in study population characteristics, study calendar years and definition of surveillance mammography it is difficult to know whether any of the observed differences among the studies are important. For example, we excluded diagnostic mammograms in defining surveillance mammography. This may have resulted in lower estimates than previous studies that used less restrictive definitions of surveillance mammography. Therefore, caution should be taken when making comparisons to these previous studies. Caution should also be taken when comparing the current findings to survey estimates19 because surveys rely on participant self-report, which has been shown to overestimate actual mammography usage.20, 21
Contrary to previous studies,14, 16, 17 we observed higher rather than lower surveillance mammography among minority women in comparison to non-Hispanic white women. Given that the MHS provides equal healthcare access it was hypothesized that surveillance mammography would be similar across racial/ethnic groups; instead, we observed that minority groups, particularly API women, tended to be more likely than non-Hispanic white women to have surveillance mammography. Previous studies have indicated that having no health insurance and thus limited access to health care may be a greater barrier to obtaining mammography among API women than other minority women;22-24 therefore, equal access healthcare in the current study may have been able to overcome the major obstacle to mammography among API women. It is also possible that API women may have developed closer relationships with their healthcare providers during cancer treatment because there have been indications that they are more likely to experience adverse effects of chemotherapy,25 which likely results in closer follow-up during treatment and possibly after treatment. Previous cancer screening is a strong predictor of future screening;23 therefore, given that the women in the later assessment years were included in the previous year(s), the relatively consistent finding of higher rates among API women across all three years may be due to the same women continuing screening adherence.
It is possible that incomplete mammography data may have affected our findings; however, it seems unlikely that completeness was differential by race/ethnicity and thus an explanation of our findings. If beneficiaries have other health insurance and seek care that is not paid for by the DoD then there would be no record of this care in MDR. Albeit not always significant, this may partially explain the lower surveillance mammography usage among women who did not have Tricare prime coverage (HMO component), assuming that women who have Tricare prime coverage are less likely to have other health insurance and thus have more complete data in MDR. However, under this same assumption, incompleteness of mammography data does not appear to explain the observed racial/ethnic differences; during sensitivity analyses when restricted to Tricare prime beneficiaries, similar racial/ethnic differences were observed (data not shown).
As in previous studies, we found that surveillance mammography was associated with other demographic, tumor and health characteristics. In agreement with Geller et al.,15 in comparison with women aged 40-49 years, we observed higher surveillance mammography among older women. However, we observed the highest usage among women aged 70-79 years as opposed to those aged 60-69 years and we did not observe a significant decrease among the eldest age group (80+ years), which had also been observed by others.12, 14 Similar to previous findings,12-16 we also observed differences by surgery-radiotherapy such that women who did not receive radiotherapy following BCS were less likely to receive surveillance mammography, particularly during the first year. In contrast to previous studies,12-16 we did not observe an inverse relationship between receipt of surveillance mammography and tumor stage at diagnosis. As in previous studies,12-14, 16 we observed an indication that women with comorbidities were less likely to receive surveillance mammography, which may be, as others have hypothesized,12, 16 an indication that these women and their providers were more concerned with other health issues. Finally, in agreement with previous studies,12-14, 19 we observed that surveillance mammography declined with time since diagnosis.
The main strengths of this study were that it allowed for the assessment of racial/ethnic disparities within an equal access healthcare system and allowed for the assessment of surveillance mammography across a wider age range than many previous studies. Another strength was the ability to adjust for demographics, tumor characteristics and comorbidities was made possible by combining cancer registry and medical administrative databases. Limitations of the study included those inherent to using medical administrative databases, which include coding inaccuracies. Another possible study limitation was the unavailability of data on possible confounders, including whether follow-up care was provided by a specialist (oncologist/radiologist), which has been associated with increased surveillance mammography.13, 14, 16, 26 However, unless minority women in the MHS are more likely to be seen by specialists than non-Hispanic white women this is an unlikely explanation for our observations by race/ethnicity. Finally, because we did not have data on the reason why a mammogram was conducted, some diagnostic mammograms may have been counted as surveillance but our classification scheme should have minimized this possibility.
In summary, among women with equal healthcare access, minority women were equally or more likely than non-Hispanic white women to receive surveillance mammography. The reasons for these findings are not clear. However, even with equal healthcare it is apparent that a large proportion of breast cancer survivors do not receive the recommended level of surveillance, particularly as time passes after diagnosis. Efforts should be taken to educate breast cancer survivors and their healthcare providers on the importance of continued surveillance.
Condensed Abstract.
Minority women are often less likely than non-Hispanic white women to receive surveillance mammography; however this was not observed within an equal access healthcare system. The racial disparities in surveillance mammography reported in other studies may be due to racial/ethnic variations in healthcare access.
Acknowledgments
This study was supported by John P. Murtha Cancer Center, Walter Reed National Military Cancer Center via the Uniformed Services University of the Health Sciences under the auspices of the Henry M. Jackson Foundation for the Advancement of Military Medicine. Drs McGlynn, Zahm and Anderson were supported by the Intramural Research Program of the DCEG, National Cancer Institute.
Footnotes
Disclaimer: The opinions and assertions expressed in this article represent the private views of the authors and do not reflect the official views of the U.S. Departments of the Army, Navy, or Defense, National Cancer Institute, or U.S. Government. Nothing in the presentation implies any Federal/Department of Defense/Department of the Navy endorsement.
Financial disclosures: There are no financial disclosures from any authors.
References
- 1.American Cancer Society . Cancer facts and figures 2012. American Cancer Society; Atlanta: 2012. [Google Scholar]
- 2.American Cancer Society . Cancer facts & figures. American Cancer Society; Atlanta: 2012. [Google Scholar]
- 3.Howlader N, Noone AM, Krapcho M, et al. SEER Cancer Statistics Review, 1975-2009 (Vintage 2009 Populations) National Cancer Institute; Bethesda, MD: http://seer.cancer.gov/csr/1975_2009_pops09/, based on November 2011 SEER data submission, posted to the SEER web site, April 2012.
- 4.Bernstein JL, Thompson WD, Risch N, Holford TR. Risk factors predicting the incidence of second primary breast cancer among women diagnosed with a first primary breast cancer. Am J Epidemiol. 1992;136(8):925–36. doi: 10.1093/oxfordjournals.aje.a116565. [DOI] [PubMed] [Google Scholar]
- 5.Broet P, de la Rochefordiere A, Scholl SM, et al. Contralateral breast cancer: annual incidence and risk parameters. J Clin Oncol. 1995;13(7):1578–83. doi: 10.1200/JCO.1995.13.7.1578. [DOI] [PubMed] [Google Scholar]
- 6.Harvey EB, Brinton LA. Second cancer following cancer of the breast in Connecticut, 1935-82. Natl Cancer Inst Monogr. 1985;68:99–112. [PubMed] [Google Scholar]
- 7.Rubino C, Arriagada R, Delaloge S, Le MG. Relation of risk of contralateral breast cancer to the interval since the first primary tumour. Br J Cancer. 2010;102(1):213–9. doi: 10.1038/sj.bjc.6605434. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Soerjomataram I, Louwman WJ, Lemmens VE, de Vries E, Klokman WJ, Coebergh JW. Risks of second primary breast and urogenital cancer following female breast cancer in the south of The Netherlands, 1972-2001. Eur J Cancer. 2005;41(15):2331–7. doi: 10.1016/j.ejca.2005.01.029. [DOI] [PubMed] [Google Scholar]
- 9.Altekruse SF, Kosary CL, Krapcho M, et al. SEER Cancer Statistics Review, 1975-2007. National Cancer Institute; Bethesda, MD: http://seer.cancer.gov/csr/1975_2007/, based on November 2009 SEER data submission, posted to the SEER web site, 2010.
- 10.Cady B, Michaelson JS, Chung MA. The “tipping point” for breast cancer mortality decline has resulted from size reductions due to mammographic screening. Ann Surg Oncol. 2011;18(4):903–6. doi: 10.1245/s10434-011-1557-y. [DOI] [PubMed] [Google Scholar]
- 11.Khatcheressian JL, Wolff AC, Smith TJ, et al. American Society of Clinical Oncology 2006 Update of the Breast Cancer Follow-Up and Management Guidelines in the Adjuvant Setting. JCO. 2006;24(31):5091–97. doi: 10.1200/JCO.2006.08.8575. [DOI] [PubMed] [Google Scholar]
- 12.Carcaise-Edinboro P, Bradley CJ, Dahman B. Surveillance mammography for Medicaid/Medicare breast cancer patients. J Cancer Surviv. 2010;4(1):59–66. doi: 10.1007/s11764-009-0107-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Doubeni CA, Field TS, Ulcickas Yood M, et al. Patterns and predictors of mammography utilization among breast cancer survivors. Cancer. 2006;106(11):2482–8. doi: 10.1002/cncr.21893. [DOI] [PubMed] [Google Scholar]
- 14.Field TS, Doubeni C, Fox MP, et al. Under utilization of surveillance mammography among older breast cancer survivors. J Gen Intern Med. 2008;23(2):158–63. doi: 10.1007/s11606-007-0471-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Geller BM, Kerlikowske K, Carney PA, et al. Mammography surveillance following breast cancer. Breast Cancer Res Treat. 2003;81(2):107–15. doi: 10.1023/A:1025794629878. [DOI] [PubMed] [Google Scholar]
- 16.Keating NL, Landrum MB, Guadagnoli E, Winer EP, Ayanian JZ. Factors related to underuse of surveillance mammography among breast cancer survivors. J Clin Oncol. 2006;24(1):85–94. doi: 10.1200/JCO.2005.02.4174. [DOI] [PubMed] [Google Scholar]
- 17.Mandelblatt JS, Lawrence WF, Cullen J, et al. Patterns of care in early-stage breast cancer survivors in the first year after cessation of active treatment. J Clin Oncol. 2006;24(1):77–84. doi: 10.1200/JCO.2005.02.2681. [DOI] [PubMed] [Google Scholar]
- 18.Charlson ME, Pompei P, Ales KL, CR. M. A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chronic Dis. 1987;40:373–83. doi: 10.1016/0021-9681(87)90171-8. [DOI] [PubMed] [Google Scholar]
- 19.Sabatino SA, Thompson TD, Richardson LC, Milller J. Health insurance and other factors associated with mammography surveillance among breast cancer survivors: results from a national survey. Medical Care. 2012;50(3):270–76. doi: 10.1097/MLR.0b013e318244d294. [DOI] [PubMed] [Google Scholar]
- 20.Rauscher GH, Johnson TPCY, Walk JA. Accuracy of self-reported cancer-screening histories: a meta-analysis. Cancer Epidemiol Biomarkers Prev. 2008;17(4):748–57. doi: 10.1158/1055-9965.EPI-07-2629. [DOI] [PubMed] [Google Scholar]
- 21.Njai R, Siegel PZ, Miller JW, Liao Y. Misclassification of survey responses and black-white disparity in mammography use, behavioral risk factor surveillance system, 1995-2006. Prev Chronic Dis. 2011;8(3):A59. [PMC free article] [PubMed] [Google Scholar]
- 22.Pourat N, Kagawa-Singer M, Breen N, Sripipatana A. Access versus acculturation: identifying modifiable factors to promote cancer screening among Asian American women. Med Care. 2010;48(12):1088–96. doi: 10.1097/MLR.0b013e3181f53542. [DOI] [PubMed] [Google Scholar]
- 23.Schueler KM, Chu PW, Smith-Bindman R. Factors associated with mammography utilization: a systematic quantitative review of the literature. Journal of Women's Health. 2008;17(9):1477–98. doi: 10.1089/jwh.2007.0603. [DOI] [PubMed] [Google Scholar]
- 24.Moy B, Park ER, Feibelman S, Chiang S, Weissman JS. Barriers to repeat mammography: cultural perspectives of African American, Asian, and Hispanic women. Psycho-Oncology. 2006;15:623–34. doi: 10.1002/pon.994. [DOI] [PubMed] [Google Scholar]
- 25.Han HS, Reis IM, Zhao W, et al. Racial differences in acute toxicities of neoadjuvant or adjuvant chemotherapy in patients with early-stage breast cancer. Eur J Cancer. 2011;47(17):2537–45. doi: 10.1016/j.ejca.2011.06.027. [DOI] [PubMed] [Google Scholar]
- 26.Hollowell K, Olmsted CL, Richardson AS, et al. American Society of Clinical Oncology-recommended surveillance and physician specialty among long-term breast cancer survivors. Cancer. 2010;116(9):2090–8. doi: 10.1002/cncr.25038. [DOI] [PubMed] [Google Scholar]