Abstract
BACKGROUND
Cancer treatment accounts for approximately 5% of national health expenditures. However, no state-level estimates of cancer treatment costs have been published.
METHODS
In analyses of data from the Medical Expenditure Panel Survey, the National Nursing Home Survey, the US Census Bureau, the Current Population Survey, and the Centers for Medicare & Medicaid Services, this study used regression modeling to estimate annual state-level cancer care costs during 2004 to 2008 for 4 categories of payers: all payers, Medicare, Medicaid, and private insurance.
RESULTS
State-level cancer care costs ranged from $227 million to $13.6 billion (median = $2.0 billion) in 2010 dollars. Medicare paid between 25.1% and 36.1% of these costs (median = 32.5%); private insurance paid between 36.0% and 49.6% (median = 43.3%); and Medicaid paid between 2.0% and 8.8% (median = 4.8%). Cancer treatment accounted for 3.8% to 8.7% of all state-level medical expenditures (median = 7.0%), 8.5% to 15.0% of state-level Medicare expenditures (median = 10.6%), 1.0% to 4.9% of state-level Medicaid expenditures (median = 2.2%), and 5.5% to 10.9% of state-level private insurance expenditures (median = 8.7%).
CONCLUSIONS
The costs of cancer treatment were substantial in all states and accounted for a sizable fraction of medical expenditures for all payers. The high cost of cancer treatment underscores the importance of preventing and controlling cancer as one approach to manage state-level medical costs.
Keywords: cost of illness, payer, cancer, state cancer cost, state cancer cost by payer
INTRODUCTION
Each year, approximately 1.6 million persons in the United States receive a cancer diagnosis.1 Cancer accounts for approximately 5% of overall national health care expenditures.2–4 Over the past 20 years, the cost of treating the most common cancers has nearly doubled nationally.2,5
Previous studies have quantified cancer costs at the national level and for specific payers.2,3,6–10 For example, in 2010, Tangka et al2 reported that 34% of US medical expenditures attributed to cancer treatment in 2001 to 2005 were financed by Medicare and Medicaid and 7% by other public sector health plans. However, there has not been an analysis of state-level cancer treatment costs to public sector payers even though these costs are among the justifications for publicly funded cancer prevention efforts at the state level. Although we are not aware of any estimates of state-level spending on cancer prevention, overall public health spending varies widely by state11 despite its beneficial effects.12
This study estimated annual state-level medical expenditures during 2004 to 2008 for cancer treatment by Medicare, by Medicaid, by private insurers, and by all payers combined. We also estimated the following by state and payer: the proportion of the population treated for cancer within the last year; the aggregate cost of cancer treatment in dollars and as a percentage of all medical costs; and the percentage of cancer treatment costs paid by Medicare, by Medicaid, and by private insurance.
Our cost estimates represent the dollars that could be saved if cancer were prevented, an important part of the decision to invest in cancer prevention. Prevention investments should also consider the costs of the prevention efforts13 and nonmonetary benefits provided by prevention.14,15 For example, if cancer treatment costs are low relative to other diseases, then investments in effective cancer prevention need to either be cheap or provide substantial nonmonetary (quality of life) benefits. Otherwise, prevention efforts should be targeted elsewhere. This study may be useful to state-level cancer prevention and control programs in allocating resources across insured populations and setting priorities for cost containment.
MATERIALS AND METHODS
Overview
We estimated state-level cancer treatment costs in 3 steps. First, we estimated, by state, the payer populations (ie, the number of people with health care coverage through each payer) for Medicare, Medicaid, private insurance and all payers combined. Second, we estimated “treated cancer prevalence” (ie, the percentage of members of each population who had been treated for cancer within the previous year) and average cost per person treated by adult age group (18–44, 45–64, or ≥65 years), sex (male or female), and payer (Medicare, Medicaid, private insurance, or all payers). Third, we generated the total costs of cancer treatment by multiplying population, treated cancer prevalence, and average costs for each demographic cell by state and payer.
Payer Population Estimates
Our payer enrollment estimates for all payers combined were based on 2008 US Census Bureau estimates, those for Medicaid on Medicaid Statistical Information System statistics for fiscal year 2008, those for Medicare on data from the Kaiser Family Foundation 2008 Medicare Health and Prescription Drug Plan Tracker, and those for private insurance on data from the 2008 Current Population Survey (CPS).16–19 Estimates of the number of Medicaid beneficiaries included all beneficiaries, including those enrolled in capitated Medicaid managed care, during 2008. Estimates of the number of Medicare beneficiaries included all beneficiaries with Medicare Parts A or B.
Because the Centers for Medicare & Medicaid Services does not report Medicaid or Medicare enrollment by sex or age, we estimated the sex and age distribution of the population covered by each payer on the basis of 2008 CPS data for private payers and 2007 through 2009 CPS data for Medicaid and Medicare. We then applied these estimated distributions to each of the state enrollment populations.
Estimates of the Treated Cancer Population
To estimate the percentage of people who had been treated for cancer during the interview year, we used data from the “Medical Condition files” of the 2004 to 2008 Medical Expenditure Panel Survey (MEPS),20 a nationally representative survey of the civilian noninstitutionalized population administered by the Agency for Healthcare Research and Quality. Respondents’ self-reported conditions were transcribed by professional coders using the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) and grouped into related categories using Clinical Classification Codes.21 We considered treatment for medical conditions reported with Clinical Classification Codes of 11 through 43 and 45 as having been treatment for cancer.
For the population covered by each category of health care payer, we used logistic regression models to estimate the probability of their receiving cancer treatment that adjusted for survey year and survey participants’ age, sex, and region of residence (Northeast, South, Midwest, or West). We used stepwise regressions to identify significant interactions among these variables to be included in the models. The significant interactions in the stepwise regressions represent age-by-sex-by-region categories with enough sample and power to detect differences in cancer treatment rates. For each category of payer, we used coefficients from the logistic regressions and MEPS sampling weights to produce nationally representative estimates of average annual cancer treatment rates during 2004 to 2008 by age, sex, and region. These estimates also reflected additional adjustments made to account for cancer treatment received by nursing home residents, on the basis of data from the 2004 National Nursing Home Survey.22
To estimate the number of people in each payer/age/ sex category who had been treated for cancer in 2008, we multiplied our estimates of treated cancer prevalence for each payer/age/sex category by our estimates of the total number of people in the corresponding category. To estimate the total number of people treated for cancer in each payer category, we then added the estimated numbers treated in each sex and age category within that payer category.
Estimates of Cancer Treatment Costs
MEPS measures total annual medical spending, which in addition to the payments by public and private insurers that we looked at in our study, also includes out-of-pocket payments by treatment recipients or other noninsurance payers in the form of copayments, deductibles, and payments for services not covered by insurance. The costs captured by MEPS represent actual payments to providers rather than the charges of providers, which may or may not ever be collected. MEPS spending data are obtained through a combination of self-reports of utilization by patients and validation of these reports by payers.
We used a 5-step process to estimate cancer treatment costs incurred by each category of payer. First, we estimated per-person cancer treatment costs by using a logistic regression model to predict the probability that a person would incur any medical costs and a generalized linear model with a gamma distribution and a log-link to estimate the total annual medical expenditures for people who incurred such costs. To choose among alternative nonlinear estimators, we used an algorithm recommended by Manning and Mullahy23 and found the generalized linear model was the most appropriate for the data. All regressions included the following variables: age; age squared; sex; race/ethnicity; education; family income; other sources of health insurance; year indicators; and indicators for cancer, arthritis, asthma, back problems, congestive heart failure, chronic obstructive pulmonary disease, coronary heart disease, depression, diabetes, dyslipidemia, HIV/AIDS, hypertension, injuries, other cardiovascular disease, other mental health/substance abuse, pneumonia, pregnancies, renal failure, skin disorders, and stroke.
Second, we calculated expenditures attributable to cancer by comparing predicted expenditures for people with each unique combination of diseases with predicted expenditures for people without that combination of diseases while holding all other variables constant. For example, cancer alone and cancer with hypertension were treated as 2 different combinations of diseases. We then divided the total expenditures attributable to the combinations of diseases back to the constituent diseases (ie, a share of all cancer with hypertension disease costs that are attributable to cancer). The shares attribute a greater share of the joint expenditures to the disease with the larger coefficient in the regression.24 We estimated per-person cancer treatment costs for each age/sex/region category on the basis of coefficients from the national model.
Third, we used data from the 2004 National Nursing Home Survey22 and National Health Accounts25 to adjust the per-person estimates of cancer treatment costs to account for the population of people with cancer that reside in nursing homes. Cost for nursing home residents is of 2 types: baseline, per-diem nursing home costs and any medical treatment received outside of the home (eg, for acute illnesses). For the latter, we assume cancer treatment costs received outside of the nursing home are equal to those for non-nursing home patients. This includes hospice costs, which are captured in MEPS for the noninstitutionalized population.
Fourth, we used restricted-access MEPS data to generate state-specific per-person cost estimates for residents of the 30 largest states. For other states, we used aggregated data to the Census division to which they belong. We regressed log (positive) medical expenditures on the variables in the model plus state/census division dummies. The coefficients on the dummies provided measures of the differences in average medical care costs across states that we used to scale the national estimates to make them state-specific.
Fifth, after adjusting MEPS cost data to 2010 values in accordance with Agency for Healthcare Research and Quality recommendations,26 we multiplied our per-person estimates of cancer treatment costs in each age/sex category by the estimated number of people whose cancer treatment was paid for by each category of payer in each state. We also estimated the share of all medical care costs attributable to cancer treatment (ie, the attributable fraction) by dividing our estimates of cancer treatment costs by national health account estimates of total medical expenditures by payer and state. All cost estimates are expressed as 2010-equivalent dollars.
Because of the large number of data sources that we used to produce our estimates, we could not generate standard errors for our estimates of cancer treatment costs by payer and state. However, because MEPS data were the primary source of sampling error, we generated standard errors for the logistic regression-treated prevalence estimates and the per-person medical cost estimates.
RESULTS
Our estimates of state-level average annual cancer prevalence rates during 2004 to 2008 ranged from 3.2% in Utah to 5.1% in Florida (median = 4.2%). By category of payer, the estimated prevalence rate was highest among Medicare beneficiaries (median = 16.1%) and lowest among Medicaid beneficiaries (median = 2.8%) (Table 1). Average annual cancer costs per person ranged from $9990 (Alaska) to $12,620 (Michigan) and were lower for Medicaid than Medicare or private insurance (Table 2). Total estimated state-level cancer treatment costs ranged from $227 million in Alaska to $13.6 billion in California (median = 2.0 billion; Table 3). Cancer treatment costs accounted for 3.8% (Alaska) to 8.7% (Florida) of total state medical expenditures (median = 7.0%; Table 4).
TABLE 1.
Estimates of Average Annual State-Level Cancer Prevalence Rates During 2004 to 2008, Overall and Among Residents Covered by Medicare, Medicaid, and Private Insurance
| State | All Residents | Medicare | Medicaid | Private Insurance |
|---|---|---|---|---|
| Alabama | 4.4 | 15.9 | 3.3 | 4.8 |
| Alaska | 3.3 | 15.4 | 2.2 | 4.1 |
| Arizona | 4.2 | 16.9 | 1.9 | 4.9 |
| Arkansas | 4.5 | 16.2 | 2.1 | 5.0 |
| California | 3.8 | 16.9 | 2.5 | 4.4 |
| Colorado | 3.8 | 16.9 | 2.7 | 4.4 |
| Connecticut | 4.5 | 16.3 | 2.6 | 4.9 |
| Delaware | 4.5 | 16.5 | 3.0 | 5.0 |
| District of Columbia | 4.1 | 16.1 | 3.4 | 4.5 |
| Florida | 5.1 | 17.2 | 4.0 | 5.5 |
| Georgia | 3.7 | 15.6 | 2.4 | 4.0 |
| Hawaii | 4.5 | 16.9 | 2.4 | 5.2 |
| Idaho | 4.0 | 17.0 | 2.0 | 4.7 |
| Illinois | 3.7 | 14.7 | 2.4 | 3.9 |
| Indiana | 3.8 | 14.9 | 2.2 | 4.1 |
| Iowa | 4.2 | 15.4 | 3.5 | 4.4 |
| Kansas | 3.9 | 15.1 | 2.5 | 4.1 |
| Kentucky | 4.4 | 15.9 | 3.3 | 4.8 |
| Louisiana | 4.2 | 15.8 | 2.9 | 4.3 |
| Maine | 4.9 | 15.9 | 4.0 | 5.5 |
| Maryland | 4.2 | 16.8 | 3.0 | 4.6 |
| Massachusetts | 4.5 | 16.1 | 3.4 | 4.9 |
| Michigan | 4.0 | 14.8 | 2.8 | 4.2 |
| Minnesota | 3.9 | 15.4 | 3.5 | 4.0 |
| Mississippi | 4.2 | 15.6 | 3.3 | 4.6 |
| Missouri | 4.0 | 14.5 | 3.5 | 3.8 |
| Montana | 4.6 | 16.5 | 2.2 | 5.5 |
| Nebraska | 3.9 | 15.4 | 3.3 | 4.2 |
| Nevada | 3.9 | 16.8 | 2.3 | 4.0 |
| New Hampshire | 4.5 | 16.2 | 2.3 | 4.7 |
| New Jersey | 4.4 | 16.2 | 3.1 | 4.9 |
| New Mexico | 4.2 | 16.5 | 2.2 | 5.2 |
| New York | 4.4 | 16.1 | 3.5 | 4.7 |
| North Carolina | 4.2 | 16.0 | 2.8 | 4.8 |
| North Dakota | 4.2 | 15.7 | 3.1 | 4.4 |
| Ohio | 4.0 | 14.8 | 2.8 | 4.0 |
| Oklahoma | 4.3 | 16.4 | 2.3 | 4.7 |
| Oregon | 4.4 | 17.2 | 2.1 | 5.2 |
| Pennsylvania | 4.8 | 16.4 | 3.1 | 5.2 |
| Rhode Island | 4.6 | 15.8 | 2.9 | 4.9 |
| South Carolina | 4.4 | 16.0 | 3.5 | 4.7 |
| South Dakota | 4.1 | 15.1 | 3.4 | 4.6 |
| Tennessee | 4.4 | 16.2 | 3.1 | 4.8 |
| Texas | 3.7 | 16.2 | 2.5 | 4.1 |
| Utah | 3.2 | 16.6 | 1.8 | 3.4 |
| Vermont | 4.7 | 16.2 | 3.5 | 5.0 |
| Virginia | 4.2 | 16.5 | 3.0 | 4.4 |
| Washington | 4.1 | 16.9 | 2.1 | 4.8 |
| West Virginia | 4.9 | 15.9 | 2.8 | 5.8 |
| Wisconsin | 4.0 | 15.0 | 3.4 | 4.0 |
| Wyoming | 4.2 | 17.1 | 2.0 | 4.7 |
| Median | 4.2 | 16.1 | 2.8 | 4.7 |
TABLE 2.
Estimates of Average Annual State-Level Cancer Cost per Person During 2004 to 2008, Overall and Among Residents Covered by Medicare, Medicaid, and Private Insurance
| State | All Payers | Medicare | Medicaid | Private Insurance |
|---|---|---|---|---|
| Alabama | 11,300 | 6190 | 4850 | 6340 |
| Alaska | 9990 | 6210 | 4530 | 6450 |
| Arizona | 9660 | 5810 | 3910 | 5650 |
| Arkansas | 10,010 | 5650 | 4350 | 5750 |
| California | 9680 | 5590 | 4150 | 5590 |
| Colorado | 11,760 | 6830 | 5070 | 6920 |
| Connecticut | 12,450 | 7180 | 4580 | 7020 |
| Delaware | 11,870 | 6550 | 4170 | 6590 |
| District of Columbia | 11,510 | 6870 | 4330 | 6540 |
| Florida | 11,810 | 6760 | 5390 | 6760 |
| Georgia | 10,140 | 6400 | 4750 | 6190 |
| Hawaii | 11,600 | 6540 | 4700 | 6400 |
| Idaho | 11,260 | 6110 | 4490 | 6290 |
| Illinois | 11,360 | 6560 | 4420 | 6640 |
| Indiana | 11,630 | 6580 | 4440 | 6710 |
| Iowa | 11,510 | 6020 | 4470 | 6210 |
| Kansas | 11,150 | 6280 | 4460 | 6360 |
| Kentucky | 11,200 | 6250 | 4560 | 6430 |
| Louisiana | 10,090 | 6030 | 4620 | 6020 |
| Maine | 10,790 | 5760 | 4310 | 5950 |
| Maryland | 10,380 | 5940 | 4350 | 5960 |
| Massachusetts | 11,410 | 6430 | 4020 | 6290 |
| Michigan | 12,620 | 6960 | 4640 | 7200 |
| Minnesota | 11,250 | 6200 | 4150 | 6350 |
| Mississippi | 9800 | 5660 | 4380 | 5680 |
| Missouri | 10,360 | 5970 | 4450 | 6030 |
| Montana | 11,470 | 6320 | 4560 | 6550 |
| Nebraska | 11,220 | 6140 | 4660 | 6240 |
| Nevada | 11,030 | 6950 | 5170 | 6660 |
| New Hampshire | 10,490 | 6130 | 4250 | 6030 |
| New Jersey | 11,820 | 6850 | 4850 | 6730 |
| New Mexico | 11,400 | 6590 | 4830 | 6770 |
| New York | 11,480 | 6800 | 4760 | 6580 |
| North Carolina | 10,990 | 6200 | 4400 | 6340 |
| North Dakota | 11,400 | 6110 | 4830 | 6260 |
| Ohio | 10,570 | 6030 | 4140 | 6110 |
| Oklahoma | 10,880 | 6210 | 4340 | 6190 |
| Oregon | 11,410 | 6220 | 4310 | 6390 |
| Pennsylvania | 11,540 | 6430 | 4280 | 6290 |
| Rhode Island | 10,800 | 6280 | 3990 | 6020 |
| South Carolina | 10,940 | 6220 | 4640 | 6330 |
| South Dakota | 11,360 | 5970 | 4770 | 6220 |
| Tennessee | 12,450 | 7120 | 5110 | 7240 |
| Texas | 10,540 | 6720 | 5150 | 6640 |
| Utah | 10,840 | 6570 | 4650 | 6320 |
| Vermont | 10,610 | 6020 | 4000 | 5940 |
| Virginia | 11,310 | 6590 | 4880 | 6650 |
| Washington | 11,180 | 6350 | 4380 | 6400 |
| West Virginia | 12,090 | 6270 | 4400 | 6580 |
| Wisconsin | 12,290 | 6860 | 4490 | 6810 |
| Wyoming | 11,200 | 6350 | 4480 | 6460 |
| Median | 11,140 | 6350 | 4530 | 6350 |
TABLE 3.
State-Level Estimates of Average Annual Payments for Cancer Treatment Made During 2004 to 2008 by All Payers Combined and by Medicare, Medicaid, and Private Insurance
| State | Payments by All Payersa | Medicare Payments (% of Total) | Medicaid Payments (% of Total) | Private Insurance Payments (% of Total) | |||
|---|---|---|---|---|---|---|---|
| Alabama | 2330 | 792 | (34.0) | 132 | (5.7) | 966 | (41.5) |
| Alaska | 227 | 57 | (25.1) | 12 | (5.2) | 113 | (49.6) |
| Arizona | 2607 | 835 | (32.0) | 106 | (4.1) | 1068 | (41.0) |
| Arkansas | 1282 | 462 | (36.1) | 77 | (6.0) | 476 | (37.1) |
| California | 13,614 | 4214 | (31.0) | 1082 | (7.9) | 5672 | (41.7) |
| Colorado | 2203 | 662 | (30.1) | 85 | (3.9) | 1037 | (47.1) |
| Connecticut | 1973 | 642 | (32.5) | 62 | (3.2) | 894 | (45.3) |
| Delaware | 466 | 151 | (32.5) | 23 | (4.8) | 208 | (44.5) |
| District of Columbia | 280 | 83 | (29.6) | 25 | (8.8) | 118 | (42.0) |
| Florida | 11,028 | 3702 | (33.6) | 620 | (5.6) | 4192 | (38.0) |
| Georgia | 3680 | 1145 | (31.1) | 191 | (5.2) | 1528 | (41.5) |
| Hawaii | 678 | 214 | (31.6) | 25 | (3.7) | 306 | (45.2) |
| Idaho | 683 | 221 | (32.4) | 21 | (3.1) | 319 | (46.7) |
| Illinois | 5482 | 1706 | (31.1) | 251 | (4.6) | 2412 | (44.0) |
| Indiana | 2853 | 938 | (32.9) | 109 | (3.8) | 1236 | (43.3) |
| Iowa | 1446 | 469 | (32.4) | 79 | (5.4) | 643 | (44.4) |
| Kansas | 1213 | 396 | (32.6) | 39 | (3.2) | 536 | (44.1) |
| Kentucky | 2092 | 718 | (34.3) | 132 | (6.3) | 836 | (40.0) |
| Louisiana | 1850 | 623 | (33.7) | 155 | (8.4) | 666 | (36.0) |
| Maine | 698 | 230 | (33.0) | 53 | (7.6) | 292 | (41.8) |
| Maryland | 2461 | 739 | (30.0) | 100 | (4.1) | 1164 | (47.3) |
| Massachusetts | 3318 | 1051 | (31.7) | 170 | (5.1) | 1518 | (45.7) |
| Michigan | 4994 | 1618 | (32.4) | 236 | (4.7) | 2192 | (43.9) |
| Minnesota | 2265 | 713 | (31.5) | 110 | (4.9) | 1029 | (45.4) |
| Mississippi | 1199 | 419 | (35.0) | 96 | (8.0) | 433 | (36.1) |
| Missouri | 2454 | 835 | (34.0) | 162 | (6.6) | 935 | (38.1) |
| Montana | 509 | 166 | (32.7) | 11 | (2.2) | 231 | (45.3) |
| Nebraska | 789 | 255 | (32.3) | 38 | (4.8) | 347 | (44.0) |
| Nevada | 1126 | 383 | (34.0) | 30 | (2.7) | 454 | (40.3) |
| New Hampshire | 622 | 202 | (32.5) | 13 | (2.0) | 290 | (46.5) |
| New Jersey | 4534 | 1415 | (31.2) | 160 | (3.5) | 2083 | (46.0) |
| New Mexico | 953 | 317 | (33.3) | 53 | (5.6) | 375 | (39.4) |
| New York | 9909 | 3141 | (31.7) | 808 | (8.2) | 3948 | (39.8) |
| North Carolina | 4240 | 1385 | (32.7) | 222 | (5.2) | 1822 | (43.0) |
| North Dakota | 306 | 102 | (33.3) | 11 | (3.6) | 135 | (44.3) |
| Ohio | 4896 | 1633 | (33.4) | 236 | (4.8) | 2015 | (41.2) |
| Oklahoma | 1720 | 588 | (34.2) | 76 | (4.4) | 707 | (41.1) |
| Oregon | 1888 | 620 | (32.8) | 43 | (2.3) | 878 | (46.5) |
| Pennsylvania | 6888 | 2326 | (33.8) | 285 | (4.1) | 3015 | (43.8) |
| Rhode Island | 519 | 176 | (33.9) | 23 | (4.5) | 220 | (42.3) |
| South Carolina | 2146 | 711 | (33.1) | 143 | (6.7) | 856 | (39.9) |
| South Dakota | 377 | 118 | (31.4) | 22 | (5.9) | 168 | (44.4) |
| Tennessee | 3371 | 1144 | (34.0) | 235 | (7.0) | 1312 | (38.9) |
| Texas | 9440 | 3017 | (32.0) | 507 | (5.4) | 3657 | (38.7) |
| Utah | 955 | 286 | (30.0) | 20 | (2.1) | 459 | (48.0) |
| Vermont | 312 | 102 | (32.6) | 23 | (7.2) | 131 | (42.0) |
| Virginia | 3669 | 1166 | (31.8) | 124 | (3.4) | 1634 | (44.5) |
| Washington | 3014 | 965 | (32.0) | 109 | (3.6) | 1373 | (45.5) |
| West Virginia | 1073 | 371 | (34.6) | 47 | (4.4) | 462 | (43.1) |
| Wisconsin | 2769 | 898 | (32.4) | 231 | (8.3) | 1135 | (41.0) |
| Wyoming | 251 | 82 | (32.7) | 6 | (2.5) | 114 | (45.4) |
| Median | 1973 | 642 | (32.5) | 96 | (4.8) | 856 | (43.3) |
Cost estimates in millions of 2010-equivalent US dollars.
TABLE 4.
Estimates of the Percentage of State-Level Medical Expenditures Attributable to Cancer Treatment During 2004 to 2008, Overall and by Category of Payer
| State | All Payers % | Medicare % | Medicaid % | Private Insurance % |
|---|---|---|---|---|
| Alabama | 8.2 | 10.4 | 3.4 | 9.8 |
| Alaska | 3.8 | 11.7 | 1.3 | 5.5 |
| Arizona | 7.5 | 10.6 | 1.6 | 8.9 |
| Arkansas | 7.5 | 10.5 | 2.5 | 8.0 |
| California | 6.2 | 8.9 | 3.1 | 7.4 |
| Colorado | 7.7 | 13.6 | 2.8 | 10.4 |
| Connecticut | 6.8 | 10.9 | 1.3 | 8.9 |
| Delaware | 6.6 | 10.6 | 2.2 | 8.4 |
| District of Columbia | 4.8 | 10.3 | 1.7 | 5.9 |
| Florida | 8.7 | 10.1 | 4.7 | 9.6 |
| Georgia | 7.1 | 10.5 | 2.8 | 8.5 |
| Hawaii | 8.0 | 15.0 | 2.3 | 10.4 |
| Idaho | 8.1 | 13.8 | 1.8 | 10.9 |
| Illinois | 6.5 | 9.5 | 2.2 | 8.3 |
| Indiana | 6.9 | 10.4 | 1.8 | 8.7 |
| Iowa | 7.3 | 11.5 | 2.9 | 9.4 |
| Kansas | 6.6 | 10.6 | 1.8 | 8.4 |
| Kentucky | 7.7 | 10.6 | 3.0 | 8.9 |
| Louisiana | 6.3 | 8.5 | 2.7 | 6.6 |
| Maine | 6.5 | 10.9 | 2.5 | 7.9 |
| Maryland | 6.0 | 8.9 | 1.9 | 8.2 |
| Massachusetts | 5.7 | 9.6 | 1.6 | 7.5 |
| Michigan | 7.9 | 9.7 | 2.6 | 9.9 |
| Minnesota | 6.1 | 11.0 | 1.8 | 8.0 |
| Mississippi | 6.4 | 8.6 | 2.8 | 6.7 |
| Missouri | 6.2 | 9.3 | 2.4 | 6.8 |
| Montana | 8.3 | 14.1 | 1.5 | 10.8 |
| Nebraska | 6.4 | 10.8 | 2.5 | 8.2 |
| Nevada | 7.7 | 12.6 | 2.4 | 8.9 |
| New Hampshire | 6.3 | 11.3 | 1.0 | 8.5 |
| New Jersey | 7.1 | 9.7 | 1.8 | 9.4 |
| New Mexico | 7.6 | 13.8 | 1.9 | 8.6 |
| New York | 6.4 | 9.8 | 1.8 | 7.3 |
| North Carolina | 7.4 | 10.6 | 2.2 | 9.2 |
| North Dakota | 6.5 | 12.2 | 2.1 | 8.3 |
| Ohio | 6.3 | 9.1 | 2.0 | 7.4 |
| Oklahoma | 7.5 | 10.5 | 2.2 | 8.9 |
| Oregon | 7.9 | 13.4 | 1.5 | 10.6 |
| Pennsylvania | 7.3 | 10.3 | 1.9 | 9.2 |
| Rhode Island | 6.1 | 10.3 | 1.4 | 7.4 |
| South Carolina | 7.8 | 10.6 | 3.3 | 8.9 |
| South Dakota | 7.0 | 11.6 | 3.5 | 9.0 |
| Tennessee | 8.7 | 11.9 | 3.7 | 9.7 |
| Texas | 6.8 | 9.8 | 2.5 | 7.6 |
| Utah | 7.2 | 13.7 | 1.4 | 10.0 |
| Vermont | 6.8 | 11.6 | 2.3 | 8.3 |
| Virginia | 7.8 | 12.9 | 2.5 | 10.0 |
| Washington | 7.1 | 13.2 | 1.9 | 9.4 |
| West Virginia | 8.1 | 11.0 | 2.2 | 10.1 |
| Wisconsin | 7.0 | 12.1 | 4.9 | 8.3 |
| Wyoming | 6.9 | 14.1 | 1.3 | 9.0 |
| Median | 7.0 | 10.6 | 2.2 | 8.7 |
State-level estimates of cancer treatment costs paid by Medicare ranged from $57 million in Alaska to nearly $4.2 billion in California (median = $642 million). Medicare paid for 25.1% (Alaska) to 36.1% (Arkansas) of state-level cancer treatment costs (median = 32.5%), and cancer treatment accounted for 8.5% (Louisiana) to 15.0% (Hawaii) of state-level Medicare expenditures (median = 10.6%).
State-level estimates of cancer treatment costs paid by Medicaid ranged from $6 million in Wyoming to $1.1 billion in California (median = $96 million). Medicaid paid for 2.0% (New Hampshire) to 8.8% (District of Columbia) of state-level cancer treatment costs (median = 4.8%), and cancer treatment accounted for 1.0% (Connecticut) to 4.9% (Wisconsin) of state-level Medicaid expenditures (median = 2.2%).
State-level estimates of cancer treatment costs paid by private insurance ranged from $113 million in Alaska to $5.7 billion in California (median = $856 million). Private insurance paid for 36.0% (Louisiana) to 49.6% (Alaska) of state-level cancer treatment costs (median = 43.3%), and cancer treatment accounted for 5.5% (Alaska) to 10.9% (Idaho) of state-level private insurance expenditures (median = 8.7%).
We estimated that the relative standard errors for the estimates of treated cancer prevalence based on MEPS data were 8% for all payers, 10% for private insurance, 11% for Medicare, and 15% for Medicaid. The relative standard errors for the estimates of per-person cancer treatment costs were 9% for Medicare, 10% for private insurance, 11% for all payers, and 26% for Medicaid.
DISCUSSION
This study shows the impact of cancer treatment costs on the total annual medical expenditures of individual states during 2004 to 2008, as well as the percentage of these costs that were financed by Medicare, Medicaid, and private insurance. Our estimates that cancer treatment accounted for 3.8% to 8.7% of total state-level medical costs were consistent with results from previous national estimates showing that cancer treatment costs accounted for approximately 5% of total medical expenditures.2–4 Our finding that 30.4% to 43% of state-level expenditures for cancer treatment were paid for by either Medicare or Medicaid indicates that the cost of treating cancer has a substantial impact on both the federal budget and state budgets.
Our analysis shows high costs of cancer treatment, with the large share of these costs financed by the public sector. For comparison with other conditions, cardiovascular disease accounts for approximately 17% of national health expenditures27 and obesity accounts for approximately 9%, 42% of which is financed by public insurance and includes some cancer costs.28 Several types of cancer screening (ie, breast, cervical, and colorectal) have been shown to be cost-effective.29,30 Furthermore, the higher the treatment costs of cancer, the more cost-effective cancer prevention and control efforts become.31 Thus, helping states quantify and understand the financial impact caused by cancer can inform state decisions on investments in cost-effective cancer prevention and disease management programs. In the future, it will be beneficial to document the costs and benefits of such efforts to identify affordable approaches for decision-makers to consider for preventing and controlling cancer in their states.
Limitations
The MEPS, our primary data source, has at least 5 notable limitations that may have affected our estimates: 1) we used cross-sectional files; 2) its results are subject to sampling error; 3) participants’ reports of their cancer status were not verified by chart review; 4) its small sample sizes precluded us from stratifying our estimates of cancer costs by type of cancer; and 5) the survey did not include cancer stage or sample people who were institutionalized (although we did adjust estimates to account for cancer treatment received by nursing home residents). Our estimates do not reflect the true overall costs of cancer, which, in addition to medical treatment costs, also include patients’ and caretakers’ time costs, the costs of lost productivity, and intangible costs associated with psychological pain and stress experienced by cancer patients and their families.32,33
Because we generated state estimates from a national model, differences in our estimates of state-level costs were primarily a reflection of differences in population size and the distribution of demographic characteristics, including insurance status, rather than differences in cancer prevalence and treatment or in how cancer treatment was paid for. We adjusted state-level estimates of treated prevalence rates to account for regional differences in these rates. We also adjusted estimates of per-person treatment costs to account for state-level differences in per-person medical expenditures while controlling for demographic factors and disease prevalence. However, despite these adjustments, our results do not represent differences in state-level treatment patterns among providers.34
Conclusions
In every state, Medicare covers the largest share of cancer treatment costs, although cancer treatment accounts for a sizable fraction of medical expenditures by all payers. As the US population ages, Medicare expenditures related to cancer treatment for older Americans are likely to continue to increase.10
State-level estimates of cancer treatment costs can complement state-level cancer incidence and prevalence estimates to provide a comprehensive picture of the impact of cancer in a state population. This comprehensive picture may help cancer prevention and management programs determine priorities. The evidence from this report clearly indicates that cancer treatment imposes high annual total and public sector medical costs across states.
Acknowledgments
FUNDING SOURCES
This study was supported by the Centers for Disease Control and Prevention (contract 200-2008-27958, task order 0015).
Footnotes
Disclaimer: The findings and conclusions in this article are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.
CONFLICT OF INTEREST DISCLOSURE
The authors made no disclosure.
References
- 1.Centers for Disease Control and Prevention. United States Cancer Statistics (USCS) [Accessed August 14, 2012]; http://www.cdc.gov/Features/CancerStatistics.
- 2.Tangka FK, Trogdon JG, Richardson LC, Howard D, Sabatino SA, Finkelstein EA. Cancer treatment cost in the United States: has the burden shifted over time? Cancer. 2010;116:3477–3484. doi: 10.1002/cncr.25150. [DOI] [PubMed] [Google Scholar]
- 3.Howard DH, Molinari NA, Thorpe KE. National estimates of medical costs incurred by nonelderly cancer patients. Cancer. 2004;100:883–891. doi: 10.1002/cncr.20063. [DOI] [PubMed] [Google Scholar]
- 4.Brown ML, Lipscomb J, Snyder C. The burden of illness of cancer: economic cost and quality of life. Annu Rev Public Health. 2001;22:91–113. doi: 10.1146/annurev.publhealth.22.1.91. [DOI] [PubMed] [Google Scholar]
- 5.Dittrich B, McKinnon JJ, Warren JE. Improvement of anisotropic displacement parameters from invariom-model refinements for three L-hydroxylysine structures. Acta Crystallogr B. 2008;64:750–759. doi: 10.1107/S0108768108032163. [DOI] [PubMed] [Google Scholar]
- 6.Mariotto AB, Yabroff KR, Shao Y, Feuer EJ, Brown ML. Projections of the cost of cancer care in the United States: 2010–2020. J Natl Cancer Inst. 2011;103:117–128. doi: 10.1093/jnci/djq495. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Warren JL, Yabroff KR, Meekins A, Topor M, Lamont EB, Brown ML. Evaluation of trends in the cost of initial cancer treatment. J Natl Cancer Inst. 2008;100:888–897. doi: 10.1093/jnci/djn175. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Yabroff KR, Lamont EB, Mariotto A, et al. Cost of care for elderly cancer patients in the United States. J Natl Cancer Inst. 2008;100:630–641. doi: 10.1093/jnci/djn103. [DOI] [PubMed] [Google Scholar]
- 9.Thorpe KE, Howard D. Health insurance and spending among cancer patients. Health Aff (Millwood) 2003;(Suppl Web Exclusives):W3-189–W3-198. doi: 10.1377/hlthaff.w3.189. [DOI] [PubMed] [Google Scholar]
- 10.Trogdon JG, Tangka FK, Ekwueme DU, Guy GP, Jr, Nwaise I, Orenstein D. State-level projections of cancer-related medical care costs: 2010 to 2020. Am J Manag Care. 2012;18:525–532. [PMC free article] [PubMed] [Google Scholar]
- 11.Trust for America’s Health. [Accessed December 8, 2011];Prevention and Public Health Fund Strategic Investments in America’s Health. http://healthyamericans.org.
- 12.Mays GP, Smith SA. Evidence links increases in public health spending to declines in preventable deaths. Health Aff (Millwood) 2011;30:1585–1593. doi: 10.1377/hlthaff.2011.0196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Subramanian S, Ekwueme D, Gardner JG, Trogdon JG. Developing and testing a cost-assessment tool for cancer screening programs. Am J Prev Med. 2009;37:242–247. doi: 10.1016/j.amepre.2009.06.002. [DOI] [PubMed] [Google Scholar]
- 14.Avis NE, Crawford S, Manuel J. Quality of life among younger women with breast cancer. J Clin Oncol. 2005;23:3322–3330. doi: 10.1200/JCO.2005.05.130. [DOI] [PubMed] [Google Scholar]
- 15.Lidgren M, Wilking N, Jonsson B, Rehnberg C. Health related quality of life in different states of breast cancer. Qual Life Res. 2007;16:1073–1081. doi: 10.1007/s11136-007-9202-8. [DOI] [PubMed] [Google Scholar]
- 16.US Census Bureau. [Accessed June 6, 2011];Population Estimates - People and Households. http://www.census.gov/popest/
- 17.KaiserFamily Foundation. [Accessed May 5, 2011];2008 Medicare Health and Prescription Drug Plan Tracker. http://healthplantracker.kff.org/
- 18.Centers for Medicare & Medicaid Services. [Accessed May 5, 2011];MSIS State Summary Fiscal Year 2008. http://www.cms.hhs.gov/MedicaidDataSourcesGe-nInfo/MSIS/list.asp.
- 19.US Census Bureau. [Accessed May 5, 2011];Current Population Survey. http://www.census.-gov/cps/data/cpstablecreator.html.
- 20.Cohen JW, Monheit AC, Beauregard KM, et al. The Medical Expenditure Panel Survey: a national health information resource. Inquiry. 1996;33:373–389. [PubMed] [Google Scholar]
- 21.Agency for Healthcare Research and Quality, Healthcare Cost and Utilization Project. [Accessed October 13, 2011];Clinical Classifications Software (CCS) for ICD-9-CM. http://www.hcup-us.ahrq.gov/toolssoftware/ccs/ccs.jsp.
- 22.Centers for Disease Control and Prevention. [Accessed March 16, 2009];National Nursing Home Survey. http://www.cdc.gov/nchs/nnhs.htm.
- 23.Manning WG, Mullahy J. Estimating log models: to transform or not to transform? J Health Econ. 2001;20:461–494. doi: 10.1016/s0167-6296(01)00086-8. [DOI] [PubMed] [Google Scholar]
- 24.Trogdon JG, Finkelstein EA, Hoerger TJ. Use of econometric models to estimate expenditure shares. Health Serv Res. 2008;43:1442–1452. doi: 10.1111/j.1475-6773.2007.00827.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Centers for Medicare & Medicaid Services. [Accessed March 17, 2009];National Health Expenditure Accounts. https://www.cms.gov/NationalHealthExpendData/
- 26.Agency for Healthcare Research and Quality. [Accessed October 13, 2011];Medical Expenditure Panel Survey. http://www.meps.ahrq.gov/mepsweb/about_meps/Price_Index.shtml.
- 27.Trogdon JG, Finkelstein EA, Nwaise I, Tangka FK, Orenstein D. The economic burden of cardiovascular disease for major insurers. Health Promot Pract. 2007;8:234–242. doi: 10.1177/1524839907303794. [DOI] [PubMed] [Google Scholar]
- 28.Finkelstein EA, Trogdon JG, Cohen JW, Dietz W. Annual medical spending attributable to obesity: Payer- and service-specific estimates. Health Aff (Millwood) 2009;28:w808–w817. doi: 10.1377/hlthaff.28.5.w822. [DOI] [PubMed] [Google Scholar]
- 29.Maciosek MV, Coffield AB, Edwards NM, Flottemesch TJ, Goodman MJ, Solberg LI. Priorities among effective clinical preventive services: results of a systematic review and analysis. Am J Prev Med. 2006;31:52–61. doi: 10.1016/j.amepre.2006.03.012. [DOI] [PubMed] [Google Scholar]
- 30.Greenberg D, Earle C, Fang CH, Eldar-Lissai A, Neumann PJ. When is cancer care cost-effective? A systematic overview of cost-utility analyses in oncology. J Natl Cancer Inst. 2010;102:82–88. doi: 10.1093/jnci/djp472. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Lansdorp-Vogelaar I, Knudsen AB, Brenner H. Cost-effectiveness of colorectal cancer screening. Epidemiol Rev. 2011;33:88–100. doi: 10.1093/epirev/mxr004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Ekwueme DU, Stroud LA, Chen Y. Cost analysis of screening for, diagnosing, and staging prostate cancer based on a systematic review of published studies. Prev Chronic Dis. 2007;4:A100. [PMC free article] [PubMed] [Google Scholar]
- 33.Yabroff KR, Davis WW, Lamont EB, et al. Patient time costs associated with cancer care. J Natl Cancer Inst. 2007;99:14–23. doi: 10.1093/jnci/djk001. [DOI] [PubMed] [Google Scholar]
- 34.The Dartmouth Atlas of Health Care. [Accessed April 4, 2012]; http://www.dartmouthatla-s.org/