Use of breast MRI and other advanced imaging is increasing among patients newly diagnosed with breast cancer; individual patient and insurance-related factors are associated with receipt of these imaging tests.
Abstract
Introduction:
Because receipt of breast imaging likely occurs in nonrandom patterns, selection bias is an important issue in studies that attempt to elucidate associations between imaging and breast cancer outcomes. The purpose of this study was to analyze use of advanced diagnostic imaging in a cohort of patients with breast cancer insured by commercial, managed care, and public health plans by demographic, health insurance, and clinical variables from 2002 to 2009.
Methods:
We identified women with breast cancer diagnoses from a Surveillance Epidemiology and End Results (SEER) registry whose data could be linked to claims from participating health plans. We examined imaging that occurred between cancer diagnosis and initiation of treatment and classified patients according to receipt of (1) mammography or ultrasound only; (2) breast magnetic resonance imaging (MRI); and (3) other advanced imaging (computed tomography [CT] of the chest, abdoment, and pelvis; positron emission tomography [PET]; or PET-CT). We used logistic regression to identify factors associated with receipt of breast MRI as well as other advanced imaging.
Results:
Commercial health plan, younger age, and later year of diagnosis were strongly associated with receipt of breast MRI and other advanced imaging. Women with prescription drug plans and those who had less comorbidities were more likely to have received breast MRI.
Conclusion:
Use of breast MRI and other advanced imaging is increasing among patients newly diagnosed with breast cancer; individual patient and insurance-related factors are associated with receipt of these imaging tests. Whether use of diagnostic advanced imaging affects outcomes such as re-excision, cancer recurrence, mortality rates, and costs of breast cancer treatment remains to be determined.
Introduction
Among the estimated 209,060 women newly diagnosed with breast cancer in 2010,1 costs related to imaging are expected to be one of the most rapidly increasing expenditures.2 In particular, one study found that the costs of imaging for breast cancer increased 9.9% per year between 1999 and 2006.2 Over the past decade, the use of advanced breast imaging technologies, including magnetic resonance imaging (MRI), computed tomography (CT), positron emission tomography (PET), and PET-CT, has increased dramatically.3–6 In particular, increasing use of breast MRI has been widely described,7–9 and beginning in 2001, breast MRI was listed as an optional method of imaging for staging evaluation and treatment planning for women with clinical stages I, IIA, and IIB (stage III was added in 2004)10–12 breast cancer from the National Comprehensive Cancer Network (NCCN). However, the NCCN guidelines stipulate that false positive findings are common, and no studies have shown that use of MRI to determine local therapy improves rates of breast cancer recurrence or survival.13
Several single-institution studies have reported variations in receipt of breast MRI by factors such as patient age and receipt of genetic testing,14,15 but to our knowledge, the only population-based studies to describe variations in receipt of peridiagnostic breast MRI and other advanced imaging were limited to patients insured by Medicare.9,16 Using population-based claims data from a stakeholder group linked with Surveillance Epidemiology and End Results (SEER) data, we examined diagnostic imaging receipt by demographic, clinical, and health plan characteristics. We used data from the Advancing Innovative Comparative Effectiveness research (ADVICE) project,18 the purpose of which was to perform community-based comparative effectiveness studies with the rationale that research that reflected the priorities of regional stakeholders would have the greatest likelihood of leading to implementation of evidence-based clinical practices.
Because receipt of advanced breast imaging almost certainly occurs in nonrandom patterns,17 selection bias is an important issue in studies that attempt to elucidate associations between advanced imaging and breast cancer outcomes. The purpose of this study was to provide information on the variables associated with receipt of advanced diagnostic imaging for breast cancer. Future studies will be able to use these data to account for selection bias and more accurately determine whether receipt of advanced imaging is associated with differences in cost, treatment patterns, and recurrence and survival rates compared with receipt of only conventional imaging.
Methods
Study Population
We used health plan enrollment files to identify women at least 18 years old and linked these files to Western Washington's SEER registry to create a data set of women with breast carcinoma diagnosed between January 1, 2002, and December 31, 2009. These plans included the major commercial health plans in our region (Regence Blue Shield, Premera Blue Cross, and Uniform Medical Plan), publicly funded systems (Medicaid and Medicare), and an integrated health care deliverer (Group Health Cooperative). When available, claims were included in our data set for women who were enrolled in multiple participating health plans.
We used claims data to define date of breast cancer diagnosis as the date of breast biopsy (or lumpectomy/mastectomy if no biopsy was recorded; Data Supplement Table 1). We used this definition of the date of diagnosis because SEER dates were inconsistent over the study period (eg, > 75% of women diagnosed in 2002-2003 were missing day of diagnosis in SEER); use of administrative data to define date of diagnosis has been validated against registry data.19,20 To be included, women must have been enrolled in at least one of the participating insurance plans for 12 months before the date of diagnosis and for at least 6 months after, with gaps in coverage of no more than 60 days (eg, disenrollment with re-enrollment, either in the same or another health plan, took place within 60 days; n = 316 patients with 1-60 days of disenrollment). Women with prior (determined by SEER sequence number) cancer diagnoses except nonmelanoma skin cancers were excluded. We excluded grades B-cell or B-precursor cancers as well as Paget's disease, lymphoma, and unspecified histologies because we hypothesized that women with these characterizations were atypical compared with women with more common grades and histologies. Finally, we excluded women whose dates of diagnosis were the same as the dates of beginning treatment because they had no opportunity to receive advanced imaging, and we believed that women who received treatment immediately were likely different from women who did not, so excluding these women created the most clearly defined comparison groups. Numbers of patients excluded are presented in Appendix Figure A1 (online only). This study was approved by the institutional review boards at participating institutions.
Classification of Imaging
Because we were interested in imaging used for staging and treatment planning, we examined imaging that occurred between the breast cancer biopsy date and the initiation of treatment. The treatment initiation date was the date on which the first treatment claim for surgery, chemotherapy, or radiation was observed (Data Supplement Table 2) or 4 months after the biopsy, if no claims were found (n = 504 [5.5%] of women in our population). We chose to end the period of interest at 4 months because a previous study indicated that 75% of patients have begun adjuvant therapy by then21 and because we did not want to include imaging that was used for treatment monitoring.
We compared women who received breast MRI with those who received only mammography or ultrasound. In addition, we were interested in examining variations in use of advanced imaging technologies that are described as optional modalities in the NCCN guidelines, including CT of the chest, abdomen, and pelvis; PET; and PET-CT. With the exception of PET and PET-CT, which are covered only for metastatic breast cancer, each of these modalities was covered by the Centers for Medicare and Medicaid Services (CMS) throughout the period of our study.10 Although many private insurers require preauthorization for diagnostic breast MRI, it is reimbursed in at least some cases,22 as are PET scans.23
To avoid including women who received imaging for screening in our “imaged” groups, we categorized women who received imaging before biopsy in the “no imaging besides mammography or ultrasound” group. To ensure that this categorization did not alter our results, we also performed the multivariate regressions excluding women who received breast MRI (n = 339) or other advanced imaging (n = 294) in the 90 days before biopsy but not in the peridiagnostic period. Because results were essentially identical, we present only results that included the entire study population.
Variables
We considered demographic, tumor, and health plan characteristics. For patient demographic variables, we examined age at diagnosis, race/ethnicity, whether the patient had a rural or urban home address, the median income of the US Census tract of the patient's home address, and comorbidity (defined using the Klabunde score24 and categorized as zero, one, or two or more comorbidities). Health insurance characteristics included health plan, whether each woman's health plan included prescription drug coverage, and whether the health plan was a high-deductible plan. Women with claims from only one health plan throughout the study period were counted as having either Medicare; commercial plans A, B, or C; managed care; or Medicaid. Some women (n = 395) had claims from more than one of these plans throughout the study period and were classified into a separate health plan category. Tumor characteristics consisted of year of diagnosis, American Joint Committee on Cancer (AJCC) stage (0-IV; specific to year of diagnosis to accommodate changes in AJCC staging that took place beginning in 2004), cancer grade (I-IV), histology (categorized as ductal, lobular, ductal plus lobular, and other [which included medullary, mucinous, tubular, papillary, and inflammatory histologies]), and estrogen and progesterone receptor (ER and PR) status. In addition to these factors, we examined the number of days between diagnosis and initiation of treatment.
Statistical Analysis
To examine the question of whether use of advanced imaging varied by patient demographic, health insurance, and clinical characteristics, we classified women as receiving (1) no imaging besides mammography or ultrasound, (2) breast MRI, and (3) other advanced imaging (CT of the chest, abdomen, and pelvis; PET; or PET-CT; Data Supplement Table 3). Next, we conducted univariate logistic regressions for each demographic, tumor, and health plan characteristic with either receipt of breast MRI or receipt of other advanced imaging as the outcome variables. We used results of these analyses to construct multivariable logistic regression models comparing the receipt of breast MRI and receipt of other advanced imaging to receipt of no advanced imaging. Women who were enrolled in more than one health plan (n = 395) were excluded from the multivariable model to allow comparison among women enrolled in only one type of health plan.
We investigated whether the odds of receipt of advanced imaging had changed significantly over time depending on the health insurance plan by including an interaction term between type of health plan and the year of breast cancer diagnosis. As in the multivariate analysis, we included patients who were enrolled in only one type of health plan.
Results
Of the 26,984 women with a breast cancer diagnosis in the SEER database, 9,196 were included in these analyses. Frequencies for demographic and health plan variables are presented in Table 1, classified by type of imaging received. Of women who received some type of advanced imaging, 468 received both breast MRI and CT, or PET or PET-CT. Women who received breast MRIs and other types of advanced imaging tended to be younger than women who received only mammography or ultrasound. Receipt of breast MRI was similar across the different race groups but black patients were more likely to have received CT of the chest, abdomen, and pelvis; PET; or PET-CT compared with whites. Women in urban areas were more likely to have received breast MRI compared with those in rural areas.
Table 1.
Demographic and Clinical Characteristics, Stratified by Receipt of Breast MRI and Any Advanced Imaging
| Characteristic | No Advanced Imaging (n = 6,162) |
Breast MRI (n = 2,481) |
P: No Advanced Imaging v Breast MRI | Other Advanced Imaging* (n = 1,021) |
P: No Advanced Imaging v Other Advanced Imaging | Total (N = 9,196) | |||
|---|---|---|---|---|---|---|---|---|---|
| No. | % (row) | No. | % (row) | No. | % (row) | ||||
| Age at diagnosis, years | |||||||||
| 18-40 | 115 | 48 | 108 | 45 | 51 | 21 | 241 | ||
| 41-50 | 526 | 52 | 443 | 43 | 137 | 13 | 1,021 | ||
| 51-64 | 1,622 | 59 | 952 | 35 | 352 | 13 | 2,741 | ||
| 65-70 | 1,157 | 71 | 391 | 24 | 168 | 10 | 1,637 | ||
| 71-80 | 1,689 | 73 | 469 | 20 | 222 | 9.6 | 2,314 | ||
| ≥ 81 | 1,053 | 85 | 118 | 9.5 | < .001 | 91 | 7.3 | < .001 | 1,242 |
| Race/ethnicity | |||||||||
| White | 5,592 | 67 | 2,251 | 27 | 903 | 11 | 8,330 | ||
| Black/African-American | 167 | 64 | 65 | 25 | 48 | 18 | 261 | ||
| American Indian/Alaska Native | 67 | 70 | 22 | 23 | 13 | 14 | 96 | ||
| Asian | 273 | 66 | 117 | 28 | 46 | 11 | 415 | ||
| Native Hawaiian/Other Pacific Islander | 10 | 83 | 2 | 17 | .74 | 0 | 0 | .01 | 12 |
| Other/unknown | 53 | 65 | 24 | 29 | 11 | 13 | 82 | ||
| Rural/urban home address (census tract) | |||||||||
| Rural | 614 | 74 | 154 | 18 | 95 | 11 | 834 | ||
| Urban | 5,538 | 66 | 2,322 | 28 | < .001 | 925 | 11 | .40 | 8,347 |
| Unknown | 10 | 67 | 5 | 33 | 1 | 6.7 | 15 | ||
| Health plan type | |||||||||
| Medicare | 3,214 | 73 | 891 | 20 | 433 | 9.9 | 4,389 | ||
| Commercial A | 703 | 54 | 520 | 40 | 173 | 13 | 1,307 | ||
| Commercial B | 238 | 39 | 343 | 56 | 109 | 18 | 608 | ||
| Commercial C | 134 | 35 | 237 | 62 | 69 | 18 | 383 | ||
| Managed Care | 1,471 | 80 | 287 | 16 | 135 | 7.3 | 1,848 | ||
| Medicaid | 162 | 61 | 67 | 25 | 56 | 21 | 266 | ||
| Multiple | 240 | 61 | 136 | 34 | < .001 | 46 | 12 | < .001 | 395 |
| Health plan subtype | |||||||||
| Prescription drug coverage (v not) | 3,653 | 62 | 1,910 | 32 | < .001 | 722 | 12 | < .001 | 5,912 |
| High deductible (v not) | 15 | 48 | 13 | 42 | .04 | 3 | 9.7 | .76 | 31 |
| Income ($, median census tract) | |||||||||
| < 41,100 | 1,640 | 73 | 474 | 21 | 240 | 11 | 2,254 | ||
| 41,101-50,400 | 1,602 | 70 | 558 | 24 | 233 | 10 | 2,293 | ||
| 50,401-61,700 | 1,458 | 65 | 649 | 29 | 261 | 12 | 2,254 | ||
| ≥ 61,701 | 1,462 | 61 | 800 | 33 | < .001 | 287 | 12 | .003 | 2,395 |
| Klabunde24 comorbidity score† | |||||||||
| 0 | 4,680 | 65 | 2,127 | 30 | 807 | 11 | 7,211 | ||
| 1 | 1,151 | 73 | 303 | 19 | 175 | 11 | 1,574 | ||
| > 1 | 331 | 81 | 51 | 12 | < .001 | 39 | 9.5 | .06 | 411 |
Includes computed tomography (CT) of chest, abdomen, and pelvis and positron emission tomography (PET)/PET-CT (468 women had both breast magnetic resonance imaging [MRI] and some other advanced imaging in the peri-diagnosis period).
Claims observed in the 11 months before the month preceding diagnosis.
Patients insured by managed care, Medicare, and Medicaid were less likely to have received breast MRI compared with patients insured by commercial types of health plans. Similarly, women insured by Medicare and managed care health plans were less likely to have received other advanced imaging compared with women insured by commercial plans, but women who had prescription drug plans were more likely to have received both breast MRI and other advanced imaging compared with those without prescription drug coverage. Women who had high-deductible plans (n = 31) were more likely to have received breast MRI compared with women who did not, but no difference was seen in terms of receipt of other advanced imaging. Women were more likely to have received breast MRI and other advanced imaging if they resided in census tracts with higher annual incomes. Finally, women with any comorbidities were significantly less likely to have received breast MRI.
Frequencies for tumor characteristics are presented in Table 2. Women diagnosed in more recent years were significantly more likely to receive breast MRI and/or other advanced imaging. Women who received breast MRI and women who received other forms of advanced imaging were more likely to have longer times between their dates of biopsy and the beginning of treatment. Although we could not discern when stage was determined in relation to imaging, women with AJCC stages I-III, but not stage IV, cancer were more likely to have received breast MRI compared with women classified as stage 0. Women with stages I-IV breast cancer were more likely to have received other advanced imaging compared to women with stage 0 cancer. Compared to women with ductal histologies, women with lobular or ductal plus lobular histologies were more likely to have received breast MRI and other advanced imaging. Women with negative hormone receptor tests were more likely to have received breast MRI and other advanced imaging compared to women with positive hormone receptor tests.
Table 2.
Tumor Characteristics, Stratified by Receipt of Breast MRI and Other Advanced Imaging
| Characteristic | No Advanced Imaging (n = 6,162) |
Breast MRI (n = 2,481) |
P: No Advanced Imaging v Breast MRI | Other Advanced Imaging* (n = 1,021) |
P: No Advanced Imaging v Other Advanced Imaging | Total (N = 9,196) | |||
|---|---|---|---|---|---|---|---|---|---|
| No. | % (row) | No. | % (row) | No. | % (row) | ||||
| Year of diagnosis | |||||||||
| 2002 | 811 | 91 | 37 | 4.2 | 50 | 5.6 | 889 | ||
| 2003 | 881 | 87 | 78 | 7.8 | 63 | 6.2 | 1,007 | ||
| 2004 | 794 | 79 | 150 | 15 | 89 | 8.7 | 1,009 | ||
| 2005 | 875 | 70 | 276 | 22 | 155 | 12 | 1,245 | ||
| 2006 | 830 | 66 | 340 | 27 | 149 | 12 | 1,256 | ||
| 2007 | 749 | 58 | 461 | 35 | 190 | 15 | 1,300 | ||
| 2008 | 667 | 49 | 620 | 46 | 180 | 13 | 1,361 | ||
| 2009 | 555 | 49 | 519 | 46 | < .001 | 145 | 13 | < .001 | 1,129 |
| Time from biopsy to treatment, days | |||||||||
| 1-14 | 1,585 | 86 | 177 | 9.6 | 111 | 6.0 | 1,845 | ||
| 15-30 | 2,662 | 71 | 880 | 23 | 382 | 10 | 3,765 | ||
| 31-60 | 1,285 | 52 | 1,051 | 42 | 354 | 14 | 2,493 | ||
| 61-120 | 630 | 58 | 373 | 34 | < .001 | 174 | 16 | < .001 | 1,093 |
| AJCC stage | |||||||||
| 0 | 1,120 | 74 | 374 | 25 | 37 | 2.5 | 1,511 | ||
| I | 2,767 | 71 | 1,004 | 26 | 213 | 5.5 | 3,900 | ||
| II | 1,701 | 64 | 775 | 29 | 387 | 15 | 2,666 | ||
| III | 324 | 46 | 255 | 36 | 260 | 37 | 711 | ||
| IV | 134 | 54 | 40 | 16 | 101 | 41 | 249 | ||
| Unknown | 116 | 73 | 33 | 21 | < .001 | 23 | 14 | < .001 | 159 |
| Grade | |||||||||
| I | 1,395 | 70 | 498 | 25 | 132 | 6.7 | 1,981 | ||
| II | 2,282 | 65 | 1020 | 29 | 364 | 10 | 3,487 | ||
| III | 1,850 | 63 | 821 | 28 | 482 | 16 | 2,925 | ||
| IV | 343 | 80 | 81 | 19 | < .001 | 16 | 3.7 | < .001 | 431 |
| Not determined | 292 | 78 | 61 | 16 | 27 | 7.3 | 372 | ||
| Histology category | |||||||||
| Ductal | 4,935 | 68 | 1,904 | 26 | 796 | 11 | 7,266 | ||
| Lobular | 510 | 59 | 298 | 34 | 100 | 12 | 868 | ||
| Ductal plus lobular | 441 | 62 | 229 | 32 | 93 | 13 | 711 | ||
| Other† | 268 | 79 | 47 | 14 | < .001 | 30 | 8.9 | .02 | 339 |
| ER status | |||||||||
| Positive/borderline | 4,701 | 67 | 1,970 | 28 | 711 | 10 | 7,060 | ||
| Negative | 899 | 61 | 426 | 29 | 284 | 19 | 1,470 | ||
| No results/missing/unknown | 562 | 84 | 85 | 13 | < .001 | 26 | 3.9 | < .001 | 666 |
| PR status | |||||||||
| Positive/borderline | 4,091 | 67 | 1,699 | 28 | 604 | 9.9 | 6,112 | ||
| Negative | 1,400 | 63 | 626 | 28 | 386 | 17 | 2,236 | ||
| No results/missing/unknown | 671 | 79 | 156 | 18 | < .001 | 31 | 3.7 | < .001 | 848 |
Abbreviations: AJCC, American Joint Committee on Cancer; ER, estrogen receptor; PR, progesterone receptor.
Includes computed tomography (CT) of chest, abdomen, and pelvis and positron emission tomography (PET)/PET-CT (468 women had both breast magnetic resonance imaging [MRI] and some other advanced imaging in the peri-diagnosis period).
Includes medullary, mucinous, tubular, papillary, and inflammatory histologies
MRI and advanced imaging receipt was more common among women younger than 50 years even after adjustment for other covariates (Table 3). Compared with women in Medicare, women enrolled in managed care or Medicaid were less likely to have received breast MRI, but women enrolled in two of the commercial plans were more likely to have received breast MRI. Urban residence and higher income were also associated with increased use of MRI and advanced imaging. In contrast to univariate findings, there was no difference in MRI receipt for women with prescription drug coverage or high-deductible plans in the multivariable model. Excepting stage IV, MRI receipt increased within increasing stage.
Table 3.
Multivariable Regression for Odds of Receiving Imaging
| Variable | MRI v No Advanced Imaging |
Other Advanced Imaging v No Advanced Imaging |
||
|---|---|---|---|---|
| OR | 95% CI | OR | 95% CI | |
| Age at diagnosis , years | ||||
| 18-40 | Referent | Referent | ||
| 41-50 | 0.7 | 0.5 to 1.1 | 0.7 | 0.5 to 1.2 |
| 51-64 | 0.5 | 0.3 to 0.7 | 0.6 | 0.4 to 1.0 |
| 65-70 | 0.3 | 0.2 to 0.5 | 0.6 | 0.3 to 1.0 |
| 71-80 | 0.3 | 0.2 to 0.4 | 0.5 | 0.3 to 0.9 |
| ≥ 81 | 0.1 | 0.1 to 0.2 | 0.3 | 0.2 to 0.5 |
| Race/ethnicity | ||||
| White | —* | Referent | ||
| Black/African American | 1.4 | 0.9 to 2.2 | ||
| American Indian/Alaska Native | 0.6 | 0.3 to 1.2 | ||
| Asian | 1.1 | 0.7 to 1.6 | ||
| Native Hawaiian/Other Pacific Islander | NA | |||
| Urban (v rural) | 1.5 | 1.2 to 1.9 | —* | |
| Median income category, $ | ||||
| < 41,100 | Referent | Referent | ||
| 41,100-50,400 | 1.2 | 1.0 to 1.4 | 1.1 | 0.9 to 1.4 |
| 50,401-61,700 | 1.3 | 1.1 to 1.6 | 1.4 | 1.1 to 1.7 |
| ≥ 61,701 | 1.6 | 1.3 to 1.9 | 1.4 | 1.1 to 1.8 |
| Klabunde comorbidity score | ||||
| 0 | Referent | Referent | ||
| > 0-1 | 0.7 | 0.6 to 0.8 | 1.1 | 0.9 to 1.3 |
| > 1 | 0.5 | 0.4 to 0.7 | 0.9 | 0.6 to 1.4 |
| Plan type | ||||
| Medicare | Referent | Referent | ||
| Commercial A | 0.8 | 0.6 to 1.0 | 1.1 | 0.8 to 1.5 |
| Commercial B | 1.6 | 1.2 to 2.1 | 2.1 | 1.5 to 3.2 |
| Commercial C | 1.7 | 1.3 to 2.4 | 2.3 | 1.4 to 3.4 |
| Managed care | 0.3 | 0.3 to 0.4 | 0.4 | 0.3 to 0.6 |
| Medicaid | 0.6 | 0.4 to 0.8 | 1.4 | 0.9 to 2.3 |
| Prescription plan (v no) | 1.1 | 1.0 to 1.3 | 1.0 | 0.8 to 1.2 |
| High-deductible plan (v no) | 0.8 | 0.3 to 1.9 | —3 | |
| Diagnosis year | ||||
| 2002 | Referent | Referent | ||
| 2003 | 1.8 | 1.2 to 2.8 | 1.3 | 0.8 to 2.1 |
| 2004 | 3.2 | 2.1 to 4.9 | 1.8 | 1.2 to 2.9 |
| 2005 | 4.8 | 3.3 to 7.1 | 2.3 | 1.5 to 3.4 |
| 2006 | 6.3 | 4.2 to 9.3 | 2.6 | 1.7 to 4.0 |
| 2007 | 9.8 | 6.6 to 15 | 3.5 | 2.4 to 5.3 |
| 2008 | 16 | 11 to 24 | 4.0 | 2.7 to 6.1 |
| 2009 | 17 | 11 to 25 | 4.0 | 2.6 to 6.1 |
| AJCC stage | ||||
| 0 | Referent | Referent | ||
| I | 1.5 | 1.2 to 1.8 | 3.6 | 2.1 to 6.1 |
| II | 1.9 | 1.5 to 2.4 | 9.5 | 5.6 to 16 |
| III | 2.7 | 2.0 to 3.6 | 29 | 17 to 49 |
| IV | 1.3 | 0.8 to 2.1 | 34 | 19 to 62 |
| Grade | ||||
| I | Referent | Referent | ||
| II | 1.2 | 1.0 to 1.3 | 1.2 | 0.9 to 1.5 |
| III | 1.0 | 0.8 to 1.2 | 1.4 | 1.1 to 1.9 |
| IV | 1.0 | 0.7 to 1.5 | 1.0 | 0.5 to 2.5 |
| Histology category | ||||
| Ductal | Referent | Referent | ||
| Lobular | 2.4 | 2.0 to 3.0 | 1.2 | 0.9 to 1.6 |
| Ductal plus lobular | 1.8 | 1.5 to 2.3 | 1.3 | 1.0 to 1.8 |
| Other† | 0.7 | 0.5 to 1.1 | 1.0 | 0.6 to 1.6 |
| ER status | ||||
| Positive/borderline | Referent | Referent | ||
| Negative | 1.1 | 0.9 to 1.4 | 1.1 | 0.8 to 1.4 |
| PR status | ||||
| Positive/borderline | Referent | Referent | ||
| Negative | 1.1 | 0.9 to 1.4 | 1.3 | 1.0 to 1.6 |
Abbreviations: AJCC, American Joint Committee on Cancer; ER, estrogen receptor; MRI, magnetic resonance imaging; OR, odds ratio; PR, progesterone receptor.
Variable left out because not significant (P > .10) in univariate analyses.
Includes medullary, mucinous, tubular, papillary, and inflammatory histologies.
The multivariable analysis of receipt of CT of the chest, abdomen, and pelvis; PET; or PET-CT showed an association between receipt of these imaging modalities and black/African American race. However, having no or few comorbidities and having a prescription drug health plan were not associated with receipt of other advanced imaging. Also, although women covered by Medicaid were less likely to have received breast MRI, these women were more likely (though not statistically significantly so) to have received other types of advanced imaging compared with women covered by Medicare. The strongest association with advanced imaging was by increasing stage, with a 34-fold increase (95% CI 18.8 to 62.1) in receipt for women with stage IV cancer relative to stage 0.
Finally, we investigated whether an interaction existed between year of diagnosis and the health plan that women were enrolled in at the time of diagnosis (data not shown). For receipt of both breast MRI and other advanced imaging as outcomes, after adjusting for other factors, we did not find any significant interactions between year of diagnosis and type of health plan.
Discussion
We conducted the first (to our knowledge) population-based, cross–health plan study to examine advanced imaging during the diagnostic process by demographic, tumor, and health plan characteristics. We found that women who were diagnosed in the later years of our study, had two of the commercial health insurance plans, were younger, and resided in areas that were classified as having higher median incomes were more likely to have received breast MRI and other advanced imaging. We also found that women with fewer comorbidities and with prescription drug plans were more likely to receive breast MRI, but not other advanced imaging, compared to women with more comorbidities and without prescription drug plans. In addition, we found that women with higher cancer stages (except stage IV) were more likely to have received breast MRI and other advanced imaging relative to women with in situ disease. We did not find any evidence that receipt of breast MRI or other advanced imaging increased at different rates over the years of our study on the basis of whether women were insured by Medicare, managed care, or commercial health plans.
We found that the frequency of use of advanced breast imaging increased dramatically over the time period of our study, especially for breast MRI. The expense of these imaging techniques is significant: the CMS reimbursement rates at independent outpatient testing facilities in 2009 were $1,244 for bilateral breast MRI,25 $896 for CT of the thorax,26,27 and $1,114 for PET-CT26,27 (by comparison, the 2009 reimbursement rate for a bilateral diagnostic digital mammogram was $15328 and for breast ultrasound was $11526,27). These large cost outputs lead to the question of whether receipt of advanced diagnostic imaging results in beneficial outcomes. Some studies have shown that breast MRI has apparent benefits compared with conventional imaging, including the ability to detect occult cancers8 and to target treatment appropriately.29 Research that has indicated that breast MRI may be helpful in identifying the extent of disease in certain subgroups, such as women at high risk of breast cancer30,31 and patients younger than 60 years.33,34 However, improvements in outcomes related to use of breast MRI, even in women with these factors, have not been shown.35–37 Two recent prospective randomized control trials showed no evidence that use of presurgical MRI reduced re-excision rates.38,39 Some argue that even though cancerous breasts may have additional foci of cancer, the rate of local recurrence is still low because the occult tumors are eradicated by systemic therapy.35 Receipt of advanced imaging has been associated with longer intervals from diagnosis to treatment4,40 and our results also indicate this. Furthermore, the high number of false-positive breast MRI scans may contribute to increased costs, psychological harm, and potentially avoidable mastectomies.3,5,41,42
Use of claims data has drawbacks compared with more resource-intensive methods such as medical record review, including the absence of any information on test results. Consistent with other studies using SEER, we were unable to disentangle whether imaging results affected ultimate tumor classification (eg, stage, histology). We also could not be certain that the imaging conducted was specifically for the purpose of diagnosing and staging breast cancer. We could not rule out the possibility that some women who were classified as not having received advanced imaging could have paid for the imaging out of pocket and been misclassified. We also could not rule out that women may have received imaging or benefits such as prescription drug coverage from health plans that did not participate in our study. We were also unable to evaluate the influence of peridiagnostic imaging on breast cancer recurrence and mortality rates. Because specific benefits and copays varied within insurance plans and changed over the time period of our study, we were not able to analyze receipt of imaging with regard to particular facets of each health plan. Finally, our results may not be generalizable to the entire population of patients with breast cancer because our claims data included only women from western Washington who were insured by one of the health plans participating in the ADVICE project. In addition, in order to perform our analyses, our inclusion criteria stipulated that women be enrolled in the participating health plans for 1 year before and 6 months after diagnosis, so our results may not be generalizable to the population of patients with breast cancer who did not meet this criterion.
In conclusion, using a large data set from several types of health insurance plans, we identified novel factors, including having two of the commercial health insurance plans, having a prescription drug plan, and having few or no comorbidities, related to receipt of breast MRI and other advanced imaging that had not previously been described. We found several subgroups with high frequencies of advanced imaging. Given the high costs of advanced imaging, it is critical that we determine whether this increasingly common practice among patients newly diagnosed with breast cancer is providing benefits in terms of treatment and survival.
Supplementary Material
Acknowledgment
We thank the ADVICE research staff for valuable assistance with data preparation and research administration: Neil Abernethy, Lydia Andris, Susan Brandzel, Alexis Drum, Holly James, Greg Klein, Rachel Hunter-Merrill, Catherine Fedorenko, Karma L. Kreizenbeck, David Mummy, Arvind Ramaprasan, Steven Zeliadt. The content is solely the responsibility of the authors and does not necessarily represent the official views of the participating health plans, the National Cancer Institute, or the National Institutes of Health.
Supported by National Cancer Institute Grant No. CA148433 and in part through collaborations with CA148577. The collection of cancer incidence data used in this study was supported by the Cancer Surveillance System of the Fred Hutchinson Cancer Research Center, which is funded by Contracts No. N01-CN-67009 and N01-PC-35142 from the SEER Program of the National Cancer Institute, with additional support from the Fred Hutchinson Cancer Research Center and the State of Washington.
Appendix
Figure A1.
Exclusion criteria.
Author's Disclosures of Potential Conflicts of Interest
The author(s) indicated no potential conflicts of interest.
Author Contributions
Conception and design: Laura S. Gold, Diana Buist, Elizabeth T. Loggers, Ruth B. Etzioni, Larry Kessler, Scott D. Ramsey, Sean D. Sullivan
Financial support: Larry Kessler, Sean D. Sullivan
Administrative support: Laura S. Gold, Diana Buist, Sean D. Sullivan
Collection and assembly of data: Laura S. Gold, Diana Buist, Sean D. Sullivan
Data analysis and interpretation: Laura S. Gold, Diana Buist, Elizabeth T. Loggers, Ruth B. Etzioni, Larry Kessler, Scott D. Ramsey, Sean D. Sullivan
Manuscript writing: Laura S. Gold, Diana Buist, Elizabeth T. Loggers, Ruth B. Etzioni, Larry Kessler, Sean D. Sullivan
Final approval of manuscript: Laura S. Gold, Diana Buist, Elizabeth T. Loggers, Ruth B. Etzioni, Larry Kessler, Scott D. Ramsey, Sean D. Sullivan
References
- 1.National Cancer Institute. Cancer Trends Progress Report: 2011/2012 Update. Bethesda, MD: National Cancer Institute; 2012. [Google Scholar]
- 2.Dinan MA, Curtis LH, Hammill BG, et al. Changes in the use and costs of diagnostic imaging among Medicare beneficiaries with cancer, 1999-2006. JAMA. 2010;303:1625–1631. doi: 10.1001/jama.2010.460. [DOI] [PubMed] [Google Scholar]
- 3.Katipamula R, Degnim AC, Hoskin T, et al. Trends in mastectomy rates at the Mayo Clinic Rochester: Effect of surgical year and preoperative magnetic resonance imaging. J Clin Oncol. 2009;27:4082–4088. doi: 10.1200/JCO.2008.19.4225. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Bleicher RJ, Ciocca RM, Egleston BL, et al. Association of routine pretreatment magnetic resonance imaging with time to surgery, mastectomy rate, and margin status. J Am Coll Surg. 2009;209:180–187. doi: 10.1016/j.jamcollsurg.2009.04.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Pettit K, Swatske ME, Gao F, et al. The impact of breast MRI on surgical decision-making: Are patients at risk for mastectomy? J Surg Oncol. 2009;100:553–558. doi: 10.1002/jso.21406. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Smith-Bindman R, Miglioretti DL, Larson EB. Rising use of diagnostic medical imaging in a large integrated health system. Health Aff (Millwood) 2008;27:1491–1502. doi: 10.1377/hlthaff.27.6.1491. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Dang CM, Zaghiyan K, Karlan SR, et al. Increased use of MRI for breast cancer surveillance and staging is not associated with increased rate of mastectomy. Am Surg. 2009;75:937–940. [PubMed] [Google Scholar]
- 8.Lehman CD, DeMartini W, Anderson BO, et al. Indications for breast MRI in the patient with newly diagnosed breast cancer. J Natl Compr Canc Netw. 2009;7:193–201. doi: 10.6004/jnccn.2009.0013. [DOI] [PubMed] [Google Scholar]
- 9.Sommer CA, Stitzenberg KB, Tolleson-Rinehart S, et al. Breast MRI utilization in older patients with newly diagnosed breast cancer. J Surg Res. 2011;170:77–83. doi: 10.1016/j.jss.2011.04.038. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Gold LS, Klein G, Carr L, et al. The emergence of diagnostic imaging technologies in breast cancer: Discovery, regulatory approval, reimbursement, and adoption in clinical guidelines. Cancer Imaging. 2012;12:13–24. doi: 10.1102/1470-7330.2012.0003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.National Comprehensive Cancer Network I. Breast Cancer v. 1.2001. Fort Washington, PA: National Comprehensive Cancer Network; 2001. [Google Scholar]
- 12.National Comprehensive Cancer Network I. Breast Cancer v. 1.2004. Fort Washington, PA: National Comprehensive Cancer Network; 2001. [Google Scholar]
- 13.National Comprehensive Cancer Network Inc. Breast Cancer, v. 3.2012. Fort Washington, PA: National Comprehensive Cancer Network; 2012. [Google Scholar]
- 14.Carpenter SG, Stucky CC, Dueck AC, et al. Scientific Presentation Award: The impact of magnetic resonance imaging on surgical treatment of invasive breast cancer. Am J Surg. 2009;198:475–481. doi: 10.1016/j.amjsurg.2009.06.015. [DOI] [PubMed] [Google Scholar]
- 15.Miller BT, Abbott AM, Tuttle TM. The influence of preoperative MRI on breast cancer treatment. Ann Surg Oncol. 2011;19:536–540. doi: 10.1245/s10434-011-1932-8. [DOI] [PubMed] [Google Scholar]
- 16.Crivello ML, Ruth K, Sigurdson ER, et al. Advanced imaging modalities in early stage breast cancer: Preoperative use in the United States Medicare population. Ann Surg Oncol. 2012;20:102–110. doi: 10.1245/s10434-012-2571-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Houssami N, Ciatto S. Design-related bias in estimates of accuracy when comparing imaging tests: Examples from breast imaging research. Eur Radiol. 2010;20:2061–2066. doi: 10.1007/s00330-010-1779-6. [DOI] [PubMed] [Google Scholar]
- 18.Klein G, Gold LS, Sullivan SD, et al. Prioritizing comparative effectiveness research for cancer diagnostics using a regional stakeholder approach. J Comp Eff Res. 2012;1:241–255. doi: 10.2217/cer.12.16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Li X, King C, deGara C, et al. Validation of colorectal cancer surgery data from administrative data sources. BMC Med Res Methodol. 2012;12:97. doi: 10.1186/1471-2288-12-97. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Meguerditchian AN, Stewart A, Roistacher J, et al. Claims data linked to hospital registry data enhance evaluation of the quality of care of breast cancer. J Surg Oncol. 2010;101:593–599. doi: 10.1002/jso.21528. [DOI] [PubMed] [Google Scholar]
- 21.Rayson D, Chiasson D, Dewar R. Elapsed time from breast cancer detection to first adjuvant therapy in a Canadian province, 1999-2000. CMAJ. 2004;170:957–961. doi: 10.1503/cmaj.1020972. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Aetna. Clinical policy bulletin: Magnetic resonance imaging (MRI) of the breast. 2011. http://www.aetna.com/cpb/medical/data/100_199/0105.html.
- 23.Aetna. Clinical policy bulletin: Positron emission tomography (PET) http://www.aetna.com/cpb/medical/data/1_99/0071.html.
- 24.Klabunde CN, Legler JM, Warren JL, et al. A refined comorbidity measurement algorithm for claims-based studies of breast, prostate, colorectal, and lung cancer patients. Ann Epidemiol. 2007;17:584–590. doi: 10.1016/j.annepidem.2007.03.011. [DOI] [PubMed] [Google Scholar]
- 25.iCAD. Reimbursement guide. http://www.icadmed.com/support/reimbursement/documents/ReimbursementGuideforWebsite.pdf.
- 26.Centers for Medicare and Medicaid Services. Physician fee schedule, 2009. http://chfs.ky.gov/nr/rdonlyres/4fd6f67c-6c97-4061-a9fd-2441467201f2/0/mdfee2009rev4909r2.pdf. [PubMed]
- 27.Centers for Medicare and Medicaid Services. Medicare program: Changes to the hospital outpatient prospective payment system and cy 2009 payment rates. 2009. http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/HospitalOutpatientPPS/downloads/CMS-1404-FC.pdf.
- 28.GE Healthcare. Medicare reimbursement for mammography procedures, 2009. http://www.gehealthcare.com/usen/community/reimbursement/docs/mammography_cust_advisory_22609edit.pdf.
- 29.Hwang N, Schiller DE, Crystal P, et al. Magnetic resonance imaging in the planning of initial lumpectomy for invasive breast carcinoma: Its effect on ipsilateral breast tumor recurrence after breast-conservation therapy. Ann Surg Oncol. 2009;16:3000–3009. doi: 10.1245/s10434-009-0607-1. [DOI] [PubMed] [Google Scholar]
- 30.Kuhl CK, Schrading S, Leutner CC, et al. Mammography, breast ultrasound, and magnetic resonance imaging for surveillance of women at high familial risk for breast cancer. J Clin Oncol. 2005;23:8469–8476. doi: 10.1200/JCO.2004.00.4960. [DOI] [PubMed] [Google Scholar]
- 31.Sardanelli F, Podo F, D'Agnolo G, et al. Multicenter comparative multimodality surveillance of women at genetic-familial high risk for breast cancer (HIBCRIT study): interim results. Radiology. 2007;242:698–715. doi: 10.1148/radiol.2423051965. [DOI] [PubMed] [Google Scholar]
- 32.Mann RM, Hoogeveen YL, Blickman JG, et al. MRI compared to conventional diagnostic work-up in the detection and evaluation of invasive lobular carcinoma of the breast: A review of existing literature. Breast Cancer Res Treat. 2008;107:1–14. doi: 10.1007/s10549-007-9528-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Deurloo EE, Klein Zeggelink WF, Teertstra HJ, et al. Contrast-enhanced MRI in breast cancer patients eligible for breast-conserving therapy: Complementary value for subgroups of patients. Eur Radiol. 2006;16:692–701. doi: 10.1007/s00330-005-0043-y. [DOI] [PubMed] [Google Scholar]
- 34.Sardanelli F, Boetes C, Borisch B, et al. Magnetic resonance imaging of the breast: Recommendations from the EUSOMA working group. Eur J Cancer. 2010;46:1296–1316. doi: 10.1016/j.ejca.2010.02.015. [DOI] [PubMed] [Google Scholar]
- 35.Houssami N, Hayes DF. Review of preoperative magnetic resonance imaging (MRI) in breast cancer: Should MRI be performed on all women with newly diagnosed, early stage breast cancer? CA Cancer J Clin. 2009;59:290–302. doi: 10.3322/caac.20028. [DOI] [PubMed] [Google Scholar]
- 36.Kuhl C, Kuhn W, Braun M, et al. Pre-operative staging of breast cancer with breast MRI: One step forward, two steps back? Breast. 2007;16(suppl 2):S34–S44. doi: 10.1016/j.breast.2007.07.014. [DOI] [PubMed] [Google Scholar]
- 37.Solin LJ, Orel SG, Hwang WT, et al. Relationship of breast magnetic resonance imaging to outcome after breast-conservation treatment with radiation for women with early-stage invasive breast carcinoma or ductal carcinoma in situ. J Clin Oncol. 2008;26:386–391. doi: 10.1200/JCO.2006.09.5448. [DOI] [PubMed] [Google Scholar]
- 38.Turnbull L, Brown S, Harvey I, et al. Comparative effectiveness of MRI in breast cancer (COMICE) trial: A randomised controlled trial. Lancet. 2010;375:563–571. doi: 10.1016/S0140-6736(09)62070-5. [DOI] [PubMed] [Google Scholar]
- 39.Peters NH, van Esser S, van den Bosch MA, et al. Preoperative MRI and surgical management in patients with nonpalpable breast cancer: The MONET - randomised controlled trial. Eur J Cancer. 2011;47:879–886. doi: 10.1016/j.ejca.2010.11.035. [DOI] [PubMed] [Google Scholar]
- 40.Hulvat M, Sandalow N, Rademaker A, et al. Time from diagnosis to definitive operative treatment of operable breast cancer in the era of multimodal imaging. Surgery. 2010;148:746–750. doi: 10.1016/j.surg.2010.07.012. discussion 750-751. [DOI] [PubMed] [Google Scholar]
- 41.Houssami N, Ciatto S, Macaskill P, et al. Accuracy and surgical impact of magnetic resonance imaging in breast cancer staging: Systematic review and meta-analysis in detection of multifocal and multicentric cancer. J Clin Oncol. 2008;26:3248–3258. doi: 10.1200/JCO.2007.15.2108. [DOI] [PubMed] [Google Scholar]
- 42.Berg WA, Gutierrez L, NessAiver MS, et al. Diagnostic accuracy of mammography, clinical examination, US, and MR imaging in preoperative assessment of breast cancer. Radiology. 2004;233:830–849. doi: 10.1148/radiol.2333031484. [DOI] [PubMed] [Google Scholar]
- 43.Kriege M, Brekelmans C, Boetes C, et al. Efficacy of MRI and mammography for breast-cancer screening in women with a familial or genetic predisposition. N Engl J Med. 2004;351:427–437. doi: 10.1056/NEJMoa031759. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.