Breast cancer, the most common malignancy in women in the United States,1 is a serious disease associated with substantial medical and economic burden. Nearly 3 million US women were living with breast cancer in 2012, and it is estimated that 231,840 women would be newly diagnosed with the disease in 2015.1 Approximately 12.3% of women in the United States will receive a breast cancer diagnosis at some point in their life.1
In 2015, it is estimated that breast cancer claimed the lives of 40,290 women in the United States,1 and it is second only to lung cancer as the leading cause of cancer-related death in women.2 However, the 5-year survival rate after diagnosis of breast cancer has improved over time, increasing from 74.8% for women whose cancer was diagnosed in 1975-1977 to 90.7% for those diagnosed in 2004-2011.3 Treatment of breast cancer at earlier (vs later) stages of the disease is associated with better survival outcomes: the 5-year relative survival rate is 98.6% if the cancer is diagnosed at the local stage, 84.9% if diagnosed at the regional stage, and 25.9% if diagnosed at the distant stage.1
The recent improvement in clinical and survival outcome is largely attributable to radiographic screening for breast cancer coupled with advances in treatment. Breast cancer screening guidelines were first published in 1989 in response to modeling based on studies conducted in the 1980s.4 At present, approximately 70% of commercially insured women aged 50 to 74 years have had at least 1 screening mammogram in the previous 2 years.5
Early detection of breast cancer through screening is generally also associated with lower costs of treatment; however, the literature that supports this association predates the year 2009.6 Some studies have analyzed therapy costs in Medicare (not commercially insured) populations.7-9 Other cost studies have been performed in Europe and Canada,10-12 but their health systems are publicly funded and, therefore, are different from the US commercially insured population. Few claims-based studies (discussed later) of breast cancer treatment costs in the US Medicare and commercially insured populations have examined how costs relate to the cancer stage at diagnosis.
The purpose of the current retrospective claims data analysis was to quantify the stage-dependent average per capita cost of breast cancer treatment for a commercially insured population of women with newly diagnosed breast cancer.
We conducted a retrospective claims data analysis, by cancer stage at diagnosis, to identify per capita allowed medical costs for women who were newly diagnosed with breast cancer in 2010 in the United States. The study population was selected from the Truven Healthcare MarketScan (MarketScan) commercial claims database using 2010 as the index year, 2009 as a look-back year, and 2011 and 2012 as the 24-month look-forward period.
The MarketScan database contains all annual paid claims generated by more than 50 million commercially insured individuals, and includes member identification codes that allow the longitudinal evaluation of members in the database. The database also includes standard codes for diagnosis, procedure, and diagnosis-related group, along with site-of-service information, amounts paid by commercial insurers, and cost-sharing amounts paid by members. Drugs are reported in the database using the National Drug Code, a unique, 3-segment number that serves as a universal drug identifier.13
Study Sample Identification
The study population comprised commercially insured US women aged 18 to 64 years who were newly diagnosed with breast cancer in 2010. Health insurance coverage was required for ≥1 months in 2010 and all of 2009 to perform a 12-month look-back analysis from the 2010 index diagnosis date. Claims data from 2011 and 2012 were used to calculate costs for the 24 months after the date of breast cancer diagnosis (ie, index date). For women who reached age 65 years in 2011 or 2012, claims continued to be included for the months in which commercial insurance was their primary coverage.
Monthly enrollment in the study was based on member eligibility criteria. Inclusion criteria included having ≥2 claims in 2010 that were ≥30 days apart and were coded with 1 of 2 breast cancer International Classification of Diseases, Ninth Revision (ICD-9) diagnosis codes (174.xx, 233.0) in any position of the claim. The index diagnosis date was defined as the date of the first claim in 2010 that contained an ICD-9 diagnosis code for breast cancer.
We performed the 12-month look-back from the index diagnosis to identify women with newly diagnosed breast cancer and to exclude women who had other confounding cancers. Women who had ≥2 claims that were ≥15 days apart and were coded for any cancer other than nonmelanoma skin cancer (140.xx-172.xx, 174.xx-208.xx) were excluded from the study. Because a claim coded for metastatic cancer may have occurred before a claim coded for the primary cancer, women were not excluded if the look-back period contained only a claim coded for metastatic disease not related to breast cancer (ie, codes 196.0x-196.2x, 196.5x-196.9, 197.xx, or 198.xx).
Claim-Coding Methods for Cancer Staging
Women with newly diagnosed breast cancer were stratified based on the estimated cancer stage at diagnosis. ICD-9 coding does not specify breast cancer stage; therefore, the stage was inferred based on identification of stage-specific treatments recommended in the National Comprehensive Cancer Network (NCCN) treatment guidelines.
We developed a claims-based algorithm for staging breast cancer using the system of the American Joint Committee on Cancer (AJCC) and assigned a stage to each newly diagnosed case—stage 0, I/II, III, or IV. Stages I and II cases were combined, because the NCCN treatment recommendations are interchangeable for these stages, and, therefore, a distinct claims-based treatment identifier could not be used to differentiate them. The staging algorithm and codes are summarized in Table 1.
We performed a sensitivity analysis in which breast cancer stage was assigned using an alternative approach that identified the cancer as in situ, small tumor, large tumor, or metastatic. In situ and metastatic tumors were identified using the same logic as that of AJCC staging. Small tumors were designated as cases with partial mastectomy and no neoadjuvant chemotherapy or radiation therapy. Cases with an excisional breast procedure (Current Procedural Terminology [CPT] code 19120, 19125, or 19126, or ICD-9 procedure code 85.21) and no additional claim for a partial (CPT code 19301 or 19302, or ICD-9 procedure code 85.22, 85.23) or complete mastectomy (CPT code 19305-19307 or ICD-9 procedure code 85.43-85.48) were considered small tumors. Large tumors were designated as cases with complete mastectomy, with or without neoadjuvant chemotherapy or radiation therapy (treatment before surgery), as well as cases with partial mastectomy and neoadjuvant chemotherapy or radiation therapy. Analysis of data for this classification method showed results similar to those obtained using the AJCC staging. Because the AJCC staging is a standard method of assigning breast cancer stage, we present our results according to this approach.
Total medical costs were defined as all costs allowed by the insurance company for claims incurred from the index date (ie, diagnosis of breast cancer) through the subsequent 24 months. First-year costs reflected costs for all claims incurred during the 12 months after the index date, divided by the starting number of patients with index diagnosis. Second-year costs reflected costs for all claims incurred during months 13 to 24 after the index diagnosis, divided by the number of patients with index diagnosis and with continuing coverage in month 13.
The cost categories included inpatient breast cancer surgery, outpatient breast cancer surgery, all costs incurred on the day of infused chemotherapy, oral chemotherapy drugs, radiation therapy, other prescription drugs, and other medical costs (Table 2).
Costs related to inpatient breast surgery included professional and facility inpatient claims that contained a breast cancer surgery ICD-9 procedure code in the primary position of the claim. Costs pertaining to outpatient breast surgery included all costs incurred on the day of an outpatient claim containing a breast cancer surgery ICD-9 procedure code in the primary position of the claim.
Claims for oral chemotherapy were identified using National Drug Codes obtained from Medi-Span 2013 (a comprehensive drug database that contains information about over-the-counter and prescription drugs).14 The Healthcare Common Procedure Coding System (HCPCS) codes were used to identify injectable drugs that generally cannot be self-administered, chemotherapy, immunosuppressive drugs, inhalation solutions, and some oral drugs.15
Claims for infused chemotherapy were also identified by HCPCS codes. Costs for radiation therapy included all claims with radiation HCPCS codes or revenue codes. Other medical costs included costs for prescription drugs (excluding oral chemotherapy) and all other annual medical costs, all other medical care for patients with breast cancer, including, for example, imaging studies, inpatient stays not associated with breast surgery, and other services (eg, skilled nursing facility, hospice, laboratory).
The average medical cost in each study period was calculated based on the number of patients who remained covered at the beginning of the period. For each breast cancer stage cohort, average allowed costs were calculated for each service category, and the total allowed claims costs incurred during each cumulative 6-month interval were divided by the starting sample size. This calculation represents costs in a manner similar to an insurance reserve (ie, the expected cost of care for a given period can be estimated at the beginning of the period for a specific group of members, such as those with a known condition). Cost calculations were not adjusted for inflation or any other changes.
We identified 15,761 women aged 18 to 64 years with newly diagnosed breast cancer who had ≥2 breast cancer claims and did not have a cancer diagnosis in the 1-year look-back period (Table 3). Of these, 8360 women met the criteria for study inclusion (stage 0, N = 2300 [27.5%]; stage I/II, N = 4425 [52.9%]; stage III, N = 1134 [13.6%], and stage IV, N = 501 [6.0%]). Approximately 81% of the women remained enrolled at the end of the first 12 months after the index diagnosis date, and 72% of the 8360 remained enrolled throughout the 24 months. The others either lost eligibility or died.
The costs were higher for women with more advanced cancer stage at diagnosis for all cumulative 6-month periods (ie, 0-6 months, 0-12 months, 0-18 months, and 0-24 months; Table 4).
The average allowed costs per patient by tumor stage in the 12 months after diagnosis were $60,637, $82,121, $129,387, and $134,682 for stages 0, I/II, III, and IV, respectively; the costs for months 13 to 24 after diagnosis were $13,523, $18,514, $35,801, and $69,464, respectively (Table 5).
The average allowed costs per patient in the 24 months after the index date were $71,909, $97,066, $159,442, and $182,655 for stages 0, I/II, III, and IV, respectively.
During the initial 12 months after breast cancer diagnosis, the per-patient increase in costs between each of the first 3 cancer stages was significant (P <.01), whereas the increase in cost between stages III and IV was not significant (P = .3). For months 13 to 24 after diagnosis, the average costs by cancer stage were lower than in the first year, and the difference in costs between each disease stage was significant (P <.01). The average per-patient medical costs declined in the second year after diagnosis (P <.01). The costs from year 1 to year 2 after diagnosis declined by more than 72% in all cohorts, with the exception of patients with stage IV disease, for whom the decline was approximately 48%.
In the first year after diagnosis, treatment-related costs for breast cancer accounted for 55% to 60% of the total medical costs for women with stage 0, I/II, or III breast cancer, but for only 41% of the total for women with stage IV breast cancer. In the second year after diagnosis, the treatment-related costs for breast cancer accounted for 17% to 38% of the total costs for each disease stage, and the proportion increased with the disease stage at diagnosis.
Among the treatment-related costs in the first 12 months after diagnosis, surgery was responsible for the highest share of the cost in stage 0 disease and for the lowest share of cost for stage IV disease. Surgery costs were more than 2-fold higher for stage 0 disease ($16,909) than for stage IV disease ($7660), a difference that reflects the curative nature of surgery in earlier stages of breast cancer. In the first 12 months after diagnosis, the costs for radiation therapy were highest for patients with stage III breast cancer ($21,133) and similar for those with stage 0, I/II, or IV disease ($14,454, $14,910, and $12,015, respectively).
Chemotherapy contributed the highest percentage to the total claim costs for stage IV disease and the lowest for stage 0. The costs for chemotherapy (infused and oral therapy combined) in the first 12 months after the diagnosis were $5170, $13,818, $34,425, and $35,686 for stages 0, I/II, III, and IV, respectively. In months 13 to 24 after diagnosis, chemotherapy costs were the largest contributor to the treatment costs in every stage of the disease.
We performed a retrospective claims data analysis, by stage at diagnosis, to identify the per capita allowed medical costs for commercially insured women aged 18 to 64 years who were newly diagnosed with breast cancer in 2010.
Our analysis showed that breast cancer treatment costs and other medical costs increased by disease stage at diagnosis (ie, lower stages were associated with lower treatment costs). Specifically, the average allowed cost per patient in the 12 months after diagnosis increased the most (58%) between disease stage I/II ($82,121) and stage III ($129,387), and was largely driven by differences in chemotherapy costs.
Economic analyses of the costs of breast cancer treatment are heterogeneous and not easily generalizable.16 However, data from several published US cost-of-illness studies are consistent with our finding that the overall costs are higher for patients whose breast cancer is more advanced at the time of diagnosis. Two US studies, neither of which distinguished cost by breast cancer stage at diagnosis, warrant mention.17,18
In the first study, the total per-capita medical costs for patients with breast cancer in the 12 months after diagnosis were approximately $60,000 (in 2008 US dollars), including patients who were identified by diagnosis code but did not necessarily have breast surgery or subsequent breast cancer treatment.17 The incremental medical cost associated with patients with breast cancer was reportedly $40,000, mostly incurred in an outpatient setting. In our analysis, the average first-year medical costs were $85,772 per patient with breast cancer (all stages of diagnosis), with $47,452 directly attributable to treatment services for breast cancer.17
The second study, in which costs were analyzed for a managed care population, demonstrated that breast cancer therapy was associated with a per-patient monthly cost of $2896 (in 2004 US dollars), with hospitalization responsible for most of the cost.18 The projected annual medical costs in that study were $12,828 higher for women with breast cancer than for women without breast cancer.18 The patients were actively treated, but the coding algorithm differed from ours, and new treatments have since become available.
Moreover, unlike our study, the costs in the period immediately after a newly diagnosed breast cancer were not examined. Although we found that the average cost of breast cancer treatment services in the first year after diagnosis was $47,452, the cost decreased significantly (to $5635) by the beginning of the second year after diagnosis.18
In several recent European and Canadian studies, the cost of breast cancer treatment was analyzed by disease stage at diagnosis. For example, in a French study, researchers used claims data to analyze the cost trajectory of care for 57,552 patients with breast cancer.10 Patients with the lowest costs (6957 Euros) were those with in situ carcinoma, and patients with the highest costs (26,139 Euros) were those hospitalized for palliative care.10
A retrospective incidence-based cost-of-illness analysis among 20,439 women with breast cancer in the Flanders region of Belgium showed that the average per-patient cost during a 6-year period was higher for those with more advanced stages of breast cancer, ranging from 19,827 Euros for patients with stage I disease to 35,201 Euros for patients with stage IV disease.19
In an analysis of patient-level routine health system data for 223 patients with breast cancer in the United Kingdom, clinical disease stage was found to be the most important predictor of the cost of hospital-based care, most of which was incurred during the first 6 months after diagnosis.11
An analysis of publicly funded healthcare costs for 39,655 Canadian patients with breast cancer showed that, during the first 2 years after diagnosis, the mean cost increased by stage (stage I, $29,938; II, $46,893; III, $65,369; IV, $66,627), and the main cost drivers were cancer clinic visits, physician billings, and hospitalizations.12
This study has several limitations. First, we lacked sufficient clinical data to confirm the stage of breast cancer at diagnosis. Specifically, the algorithm we used to estimate disease stage at diagnosis provides an approximation based on claims representing treatment at various stages of breast cancer, as recommended by the NCCN guidelines. Therefore, cases grouped together based on treatment patterns and diagnosis codes may not have represented true clinical stages.
Second, using the index diagnosis date (as determined by our diagnosis coding algorithm) as the start of cancer treatment costs may have produced different results than if we had chosen a different starting point in the diagnosis and treatment course for breast cancer. However, we performed a sensitivity analysis using the biopsy date (when available) as the index date and examined the subsequent 24 months for each cohort. The cost pattern for each cohort was similar to that resulting from use of the diagnosis date as the index date.
In addition, our study population included women aged 18 through 64 years. Because it is possible that younger women (ie, those aged 18-39 years) may represent more complex or costly cases, we performed a sensitivity analysis in which the age range was limited to 40 to 64 years (ie, those eligible for routine screening within the commercially insured population); this showed that the cost trend observed (ie, increasing cost by disease stage) for the older patients did not differ from that of the overall study population.
Furthermore, our analysis of claims data represents a national average experience of a cross-section of patients with commercial insurance between 2010 and 2012. We acknowledge that regional variability may exist in terms of treatment costs, incidence of breast cancer, and distribution of stage at diagnosis, which could affect the results. Over time, revised guidelines for screening, diagnosis, and treatment, as well as the adoption of new technologies, could influence the results. In addition, our results may not reflect the cost experience of treating other populations of insured women with breast cancer, such as those with Medicare or Medicaid coverage.
Finally, the look-back period cannot capture breast cancer cases that might have been diagnosed at an earlier stage of the disease (before the 12-month look-back period). As a result, it is likely that not all metastatic cases were newly diagnosed breast cancer. Some metastatic cases might have been recurrent breast cancer that had been diagnosed and treated at an earlier disease stage (ie, >12 months before the index date in this study); this might have resulted in an overestimation of de novo stage IV cases in our study.
Our analysis of the current costs of treating breast cancer in a commercially insured population shows that incremental costs are significantly higher for advanced-stage disease than for early-stage disease. Our disease stage–specific cost findings are consistent with the results of other studies, including those conducted in the US Medicare population and in populations with publicly funded insurance in other developed countries.
Relevant data on medical costs, stratified by disease stage, may facilitate initiatives aimed at strengthening cancer management. Earlier detection of breast cancer by routine screening leads not only to reduced morbidity and mortality but also to lower costs for cancer treatment. Our findings strongly suggest that the costs of treating breast cancer could be meaningfully reduced by earlier diagnosis and treatment. Knowledge of the relevant stage-specific cost data provides support for strengthening programs such as breast cancer screening that are designed to shift breast cancer diagnosis to earlier stages.
General Electric Healthcare.
Author Disclosure Statement
Dr Blumen and Ms Fitch have received research support from General Electric Healthcare; Mr Polkus is an employee and stockholder of General Electric.
Dr Blumen is Principal and Healthcare Consultant, Milliman, Inc, New York, NY; Ms Fitch is Principal and Healthcare Consultant, Milliman, Inc, New York, NY; Mr Polkus is Director, Health Economics & Reimbursement, General Electric Healthcare, Wauwatosa, WI.
1. National Cancer Institute. SEER stat fact sheets: female breast cancer: statistics at a glance. http://seer.cancer.gov/statfacts/html/breast.html. Accessed December 23, 2015.
2. Centers for Disease Control and Prevention. Lung cancer. Updated September 16, 2015. www.cdc.gov/cancer/lung/index.htm. Accessed December 23, 2015.
3. National Cancer Institute. Table 4.13: Cancer of the female breast (invasive): 5-year relative and period survival by race, diagnosis year, age and stage at diagnosis. http://seer.cancer.gov/csr/1975_2012/browse_csr.php?sectionSEL=4&pageSEL=sect_04_table.13.html. Accessed December 23, 2015.
4. Eddy DM. Screening for breast cancer. Ann Intern Med. 1989;111:389-399.
5. National Committee for Quality Assurance. The state of health care quality 2014.http://store.ncqa.org/index.php/catalog/product/view/id/2341/s/2015-state-of-health-care-quality-report/. Accessed December 23, 2015.
6. Campbell JD, Ramsey SD. The costs of treating breast cancer in the US: a synthesis of published evidence. Pharmacoeconomics. 2009;27:199-209.
7. Gross CP, Long JB, Ross JS, et al. The cost of breast cancer screening in the Medicare population. JAMA Intern Med. 2013;173:220-226.
8. Rao S, Kubisiak J, Gilden D. Cost of illness associated with metastatic breast cancer. Breast Cancer Res Treat. 2004;83:25-32.
9. Warren JL, Yabroff KR, Meekins A, et al. Evaluation of trends in the cost of initial cancer treatment. J Natl Cancer Inst. 2008;100:888-897.
10. Jay N, Nuemi G, Gadreau M, Quantin C. A data mining approach for grouping and analyzing trajectories of care using claim data: the example of breast cancer. BMC Med Inform Decis Mak. 2013;13:130.
11. Hall PS, Hamilton P, Hulme CT, et al. Costs of cancer care for use in economic evaluation: a UK analysis of patient-level routine health system data. Br J Cancer. 2015;112:948-956.
12. Mittmann N, Porter JM, Rangrej J, et al. Health system costs for stage-specific breast cancer: a population-based approach. Curr Oncol. 2014;21:281-293.
13. US Food and Drug Administration. National Drug Code Directory. www.fda.gov/Drugs/InformationOnDrugs/ucm142438.htm. Accessed December 23, 2015.
14. Medi-Span. Medi-Span Electronic Drug File (MED-File) v2. www.wolterskluwercdi.com/medi-span-electronic-drug-file/. Accessed December 23, 2015.
15. ReimbursementCodes.com. www.j-codes.com/. Accessed December 23, 2015.
16. Radice D, Redaelli A. Breast cancer management: quality-of-life and cost considerations. Pharmacoeconomics. 2003;21:383-396.
17. Fu AZ, Jhaveri M. Healthcare cost attributable to recently-diagnosed breast cancer in a privately insured population in the United States. J Med Econ. 2012;15:688-694.
18. Barron JJ, Quimbo R, Nikam PT, Amonkar MM. Assessing the economic burden of breast cancer in a US managed care population. Breast Cancer Res Treat. 2008;109:367-377.
19. Broekx S, Den Hond E, Torfs R, et al. The costs of breast cancer prior to and following diagnosis. Eur J Health Econ. 2011;12:311-317.