Hospital-Administered Cancer Therapy Prices for Patients With Private Health Insurance

Key Points

Question  How much do US hospitals mark up the price of parenteral cancer therapies for patients with private health insurance?

Findings  This cross-sectional study examined private payer–specific prices for 25 commonly used parenteral cancer therapies at 61 National Cancer Institute–designated cancer centers. Median price markups across centers ranged from approximately 120% (sipuleucel-T) to 630% (leuprolide) of estimated hospital acquisition costs.

Meaning  The findings of this study suggest that, to reduce the financial burden of cancer treatment for patients, institution of public policies to discourage or prevent excessive hospital price markups on parenteral chemotherapeutics may be beneficial.

Importance  The federal Hospital Price Transparency final rule, which became effective in 2021, requires hospitals to publicly disclose payer-specific prices for drugs. However, little is known about hospital markup prices for parenterally administered therapies.

Objective  To assess the extent of price markup by hospitals on parenterally administered cancer therapies and price variation among hospitals and between payers at each hospital.

Design, Setting, and Participants  A cross-sectional analysis was conducted of private payer–specific negotiated prices for the top 25 parenteral (eg, injectable or infusible) cancer therapies by Medicare Part B spending in 2019 using publicly available hospital price transparency files. Sixty-one National Cancer Institute (NCI)–designated cancer centers providing clinical care to adults with cancer were included. The study was conducted from April 1 to October 15, 2021.

Exposures  Estimated hospital acquisition costs for each cancer therapy using participation data from the federal 340B Drug Pricing Program.

Main Outcomes and Measures  The primary outcome was hospital price markup for each cancer therapy in excess of estimated acquisition costs. Secondary outcomes were the extent of across-center price ratios, defined as the ratio between the 90th percentile and 10th percentile median prices across centers, and within-center price ratios, defined as the ratio between the 90th percentile and 10th percentile prices between payers at each center.

Results  Of 61 NCI-designated cancer centers, 27 (44.3%) disclosed private payer–specific prices for at least 1 top-selling cancer therapy as required by federal regulations. Median drug price markups across all centers and payers ranged between 118.4% (sipuleucel-T) and 633.6% (leuprolide). Across-center price ratios ranged between 2.2 (pertuzumab) and 15.8 (leuprolide). Negotiated prices also varied considerably between payers at the same center; median within-center price ratios for cancer therapies ranged from 1.8 (brentuximab) to 2.5 (bevacizumab).

Conclusions and Relevance  Most NCI-designated cancer centers did not publicly disclose payer-specific prices for cancer therapies as required by federal regulation. The findings of this cross-sectional study suggest that, to reduce the financial burden of cancer treatment for patients, institution of public policies to discourage or prevent excessive hospital price markups on parenteral chemotherapeutics might be beneficial.

The cost of cancer therapies is a substantial concern for patients, payers, and policy makers. Between 2015 and 2019, US spending on all cancer therapies increased from $39.1 billion to $67.5 billion.¹ Manufacturers routinely price novel anticancer therapies at more than $100 000 per year,^2,3 with subsequent price inflation by supply chain intermediaries (eg, wholesalers and group purchasing organizations) and the hospitals and physician offices where the drugs are administered.⁴ Some patients incur debt, reduce spending on necessities, and forgo recommended treatments.²

The increasing costs of cancer therapies are largely due to manufacturer and hospital market power.² Manufacturers benefit from time-limited monopolies during periods of patent and regulatory exclusivity.^3,5 Hospitals, in turn, can institute high prices for cancer therapies from private insurers. Although hospital markups are typically higher than markups in physician offices,^6,7 the provision of infused cancer therapy has shifted from physician offices to outpatient hospital departments.⁸ The historically proprietary nature of negotiated prices has limited understanding of hospital markup practices. Industry reports, however, estimate that hospitals, on average, mark up drug prices for patients with private insurance by approximately 140% to 280% in excess of acquisition costs.^9-11

On January 1, 2021, the Centers for Medicare & Medicaid Services implemented the Hospital Price Transparency final rule, which requires hospitals to publicly disclose payer-specific prices for items¹² and services, including drugs.¹³ The rule is intended to help patients make financially informed decisions when seeking care and stimulate cost-based competition among hospitals.

We analyzed price transparency files published by National Cancer Institute (NCI)–designated cancer centers to answer 3 questions. First, are the leading cancer centers disclosing negotiated prices for cancer therapies as required by federal regulation? Second, how much do these centers mark up cancer drug prices above acquisition costs? Third, what is the variation of negotiated prices for a cancer therapy across centers and between payers?

We performed a cross-sectional analysis of private payer–specific negotiated prices for cancer therapies at NCI-designated cancer centers. The study was conducted from April 1 to October 15, 2021. Because all data were obtained from publicly disclosed price transparency files from hospitals and other public sources, this study did not require exemption or approval by the Mass General Brigham Human Research Protection Program. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

The NCI recognizes 71 centers for their scientific and clinical leadership and provides them with funding to deliver innovative cancer treatments.¹⁴ We focused on these centers given their similar hospital characteristics, cancer-related services, and patient comorbidity burden.¹⁵ We restricted analysis to centers that provided clinical care to adults and excluded centers from Maryland, which regulates prices for hospital services.¹⁶ Using data from the Resources and Service Administration,¹⁷ we determined whether each center participated in the federal 340B Drug Pricing Program, which allows hospitals that serve financially vulnerable patients to purchase outpatient drugs at considerable discounts. We categorized each center as participating or not participating based on their registered status on January 1, 2021.

Using the publicly available Medicare Part B Drug Spending Dashboard,¹⁸ we identified the top 25 parenteral (eg, injectable or infusible) cancer therapies by Medicare Part B spending in 2019. Drugs were considered cancer therapies if their most recent drug label available within the US Food and Drug Administration (FDA) publicly available Drugs@FDA database included an oncologic therapeutic indication.¹⁹ We focused on parenteral cancer therapies because these drugs are typically administered in a facility setting and thus subject to hospital pricing. We characterized each included therapy using the publicly available FDA database (eTable 1 in the Supplement).¹⁹

Using previously described methods,^20-23 we systematically reviewed each cancer center’s website to determine whether the center publicly disclosed private payer–specific prices for items and services within a machine-readable file.¹³ For each center with a price transparency file, we then extracted available private payer–specific prices for each cancer therapy in the study cohort; we limited data to prices from plans administered by private payers and excluded prices listed for public payers (eg, Medicare) and international insurance plans. Data collection followed a standardized process for each cancer center and cancer therapy (eMethods in the Supplement). We confirmed that all price data were up to date as of October 15, 2021.

Centers disclose negotiated prices for cancer therapies per unit billed (eg, per milligram or milliliter). To facilitate analysis of price variation across centers, we converted all disclosed prices to the unit specified for each cancer therapy in the October 2021 update of the Medicare average sales price (ASP) Drug Pricing file.²⁴ If pricing for more than 1 formulation of a cancer therapy was listed, we used the formulation with the lowest price per unit.

Payers reimburse hospitals for covered drugs administered in the outpatient setting. The default Medicare Part B rate for each drug is statutorily calculated as manufacturer ASP plus 6%, which is intended to cover hospital costs to acquire drugs (ASP) and perform associated services (plus 6%), such as handling, storage, and other overhead.²⁵ ASP is defined as the volume-weighted mean price net of any manufacturer price concessions (eg, volume discounts or rebates), excluding concessions offered to federal programs.

Under the 340B Drug Pricing Program, the minimum allowed discount (ie, manufacturer price ceiling) is statutorily calculated based on several factors (eg, brand name and generic status). To reflect these lower drug acquisition costs, the Medicare Part B rate for each drug at hospitals participating in the 340B Drug Pricing Program equals ASP minus 22.5%.

We calculated price markups for each cancer therapy as the difference between the payer-specific prices and the estimated acquisition cost. To facilitate comparisons across cancer therapies, we expressed price markup as percent increase or decrease relative to estimated acquisition costs.

We estimated acquisition costs (eMethods in the Supplement) based on center participation in the 340B Drug Pricing Program. For nonparticipants, we used ASP as a proxy for the acquisition cost of each cancer therapy and extracted ASPs from the Medicare ASP Drug Pricing Program file.²⁴ For 340B Drug Pricing Program participants, we estimated the acquisition cost of each cancer therapy to be 65.3% of ASP in accordance with recent Centers for Medicare & Medicaid Services analysis, which used reported acquisition costs for 340B Drug Pricing Program drugs from a survey of more than 1400 participating hospitals.²⁵ The Health Resources and Service Administration estimates that hospitals participating in the 340B Drug Pricing Program are typically able to obtain discounts ranging between 25% and 50% of average wholesale drug prices.²⁵

We used descriptive statistics to characterize cancer therapies, disclosure of payer-specific cancer therapy pricing by cancer centers, and variation in disclosed cancer therapy prices. Using previously described methods,^20-23 we compared payer-specific prices between hospitals by performing a 4-step calculation to determine the across-center ratio for each drug. First, we determined the median price for each drug at each center that disclosed prices. Second, we adjusted the median price at each center for geographic differences in input costs using the local Medicare practice expense geographic practice cost index²⁶; hospitals that do not charge separate facility fees for outpatient chemotherapy administration visits may pass along input costs through drug prices. Third, we identified the 10th (lower) and 90th (higher) percentile geographic practice cost index–adjusted median prices among centers. We used 90th and 10th percentile prices to minimize variation owing to outliers. If prices did not exactly correspond to either percentile, we conservatively substituted for the next closest price to the median to avoid overstating the extent of price variation. Fourth, we divided the 90th percentile–adjusted median price by the 10th percentile–adjusted median price.²⁰

We compared payer-negotiated prices within each center (ie, between different payers contracting with the same center) by performing a 2-step calculation to determine the median within-center ratio for each drug. First, we calculated the within-center ratio for each drug at each center that disclosed prices as the quotient of the 90th and 10th percentile payer-specific prices. Second, we determined the median within-center ratio for each drug across these centers to represent payer-specific price variation.

Third, we determined the median price markup for each cancer therapy across centers. We identified the median payer-specific price at each center and calculated markup based on center participation in the 340B Drug Pricing Program. Using these representative values from each center, we then calculated the median price markup across centers.

To illustrate the potential magnitude of price variation over the entire course of cancer treatment, we scaled payer-specific prices to estimate the total costs that patients and payers would incur for cancer therapies prescribed for FDA-approved indications. For each cancer therapy, we identified all FDA-approved oncologic indications listed in the most recent available label (eMethods in the Supplement). We then estimated the total dosage of cancer therapy necessary for each indication using dosing instructions specified in the label.

We calculated the total costs for each indication as the product of the estimated total dose of cancer therapy and the negotiated prices per unit billed. For each indication, we used descriptive statistics to characterize price variation over the entire course of cancer treatment both across centers and within centers between payers. We performed across-center comparisons using median payer-specific prices at each center and performed within-center comparisons using 90th and 10th percentile prices. We conducted all analyses using R, version 4.1.1 (R Foundation for Statistical Computing).

We performed 4 sensitivity analyses (eFigures 3-5 in the Supplement). First, we calculated across-center price ratios using unadjusted prices to assess geographic differences in input costs and price variation. Second, we calculated price markups using 25th percentile payer-specific prices at each hospital because the dollar volume-weighted price of cancer therapies may be lower than the median payer-specific price if large insurers secure lower rates. Recent analysis, however, has reported that small health plans are sometimes able to secure lower rates.²⁷ Third, we calculated price markup for 340B Drug Pricing Program participants using more conservative estimates of acquisition costs in accordance with an analysis by the Medicare Payment Advisory Commission,²⁸ which estimated the price ceiling of cancer therapies without FDA-approved generic or biosimilar agents to be 76.9% of ASP and cancer therapies with FDA-approved generic and biosimilar agents to be 87.0% of ASP. Fourth, we calculated price markup using both 25th percentile payer-specific prices at each hospital and Medicare Payment Advisory Commission estimates of cancer therapy acquisition costs.

Among the top 25 parenteral cancer therapies by Medicare Part B spending (Table 1), most (17 [68.0%]) were biologics, did not have approved generic or biosimilar formulations (16 [64.0%]), and had been FDA approved as orphan drugs for at least 1 indication (20 [80.0%]). Ten drugs (40.0%) received initial FDA approval between 2013 and 2017, and 11 drugs (44.0%) received FDA breakthrough therapy designation for at least 1 approved indication. Cancer therapies were most often approved for the treatment of hematologic (11 [44.0%]) and pulmonary (9 [36.0%]) cancers.

Of the 71 NCI-designated cancer centers, we excluded 10 for the following reasons: basic laboratory research sites (n = 7), Maryland hospitals (n = 2), and exclusively pediatric hospital (n = 1). Sixty-one NCI-designated cancer centers were included in the analysis (eFigure 1 in the Supplement); among these, 68.9% (n = 42) disclosed payer-specific negotiated prices for at least 1 item or service, and 44.3% (n = 27) disclosed payer-specific prices for at least 1 of the 25 parenteral cancer therapies in our sample. Disclosure rates for cancer therapies ranged between 21.3% (n = 13) for rituximab with hyaluronidase and 42.6% (n = 26) for rituximab, bevacizumab, and leuprolide (Table 2). We excluded price data from 1 center that disclosed payer-specific prices for less than 5 private insurers to minimize variation due to outliers. Of the remaining 26 centers, 23 (88.5%) were 340B Drug Pricing Program participants. These 26 centers published prices for a median of 24 drugs (IQR, 22-25) and a median of 22 (IQR, 14-37) payer-specific prices per drug.

Payer-specific prices per unit for cancer therapies varied widely (Figure 1A). Across-center price ratios ranged from 2.2 (pertuzumab, 1 mg; 10th percentile–adjusted median price: $16 vs 90th percentile–adjusted median price: $36) to 15.8 (leuprolide 7.5 mg; 10th percentile–adjusted median price: $247 vs 90th percentile–adjusted median price: $3914). The median across-center ratio between hospitals for all drugs was 3.1 (rituximab with hyaluronidase; IQR, 2.8-3.8). Across-center price ratios were overall similar in sensitivity analysis using prices that were not adjusted for hospital locality–specific differences in practice input costs (eg, with Medicare geographic practice cost indices) (eFigure 2 in the Supplement).

Payer-specific prices for cancer therapies also varied widely between payers at each cancer center (Figure 1B). For instance, at one center, the 10th and 90th percentile–negotiated prices (n = 18) of leuprolide, 7.5 mg, ranged between $692 and $3284 per milligram. Among all centers, median within-center price ratios for cancer therapies ranged from 1.8 (brentuximab) to 2.5 (bevacizumab).

Median price markups for cancer therapies across centers ranged from 118.4% (sipuleucel-T) to 633.6% (leuprolide) in excess of estimated acquisition costs (Figure 2). Sensitivity analysis using both 25th percentile payer-specific prices and conservative estimates of acquisition costs yielded price markups ranging from 42.1% (sipuleucel-T) to 234.1% (leuprolide) (eFigure 3 in the Supplement).

The estimated price of cancer therapies over the entire course of treatment varied widely across and within cancer centers. Cost variation estimates for each indication both across centers and between payers at the same center are provided in eTable 2 in the Supplement. Figure 3 displays a subset of these estimates that includes the indication for each drug requiring the median total dosage amount relative to total dosage amounts for all FDA-approved indications for each drug. As an example of across-center price variation with pembrolizumab, the median total price across centers was $168 405 (IQR, $115 879-$225 111) for treatment of metastatic non–small cell lung cancer (total estimated dosage: 1607 mg during 24.1 weeks) (Figure 3A). As an example of within-center price variation with daratumumab, the median price difference within centers was $174 225 (IQR, $124 095-$213 186) for treatment as third-line therapy for multiple myeloma (total estimated dosage: 18 424 mg during 25.8 weeks) (Figure 3B).

We used price transparency files provided by NCI-designated cancer centers in accordance with the recent Hospital Price Transparency final rule to analyze the historically proprietary prices negotiated by hospitals and private insurers for cancer therapies. Most NCI-designated cancer centers did not publicly disclose payer-specific prices for cancer therapies as required by federal regulation. Many centers that disclosed this information substantially marked up the prices of top-selling parenteral cancer therapies administered to patients with private insurance in hospital outpatient departments. The extent of markup varied widely both across centers and within each center across payers. Our findings may have implications for further drug price transparency reforms and efforts to lower the burden of high drug prices on patients and payers.

A recent analysis of 10 high-expenditure drugs at 20 top-rated hospitals found that 45% were noncompliant with requirements to disclose payer-specific negotiated drug prices that, when disclosed, were typically much higher than Medicare payment limits.²⁹ Our study builds on this work by additionally estimating hospital markup relative to acquisition costs and the magnitude of price variation over the course of treatment in a larger and more clinically homogeneous cohort of NCI-designated cancer centers.

Some hospitals may lack the technical expertise or organizational resources necessary for compliance,³⁰ although this scenario is unlikely the case for leading cancer centers. Other hospitals may elect not to publish prices to protect bargaining power (relative to insurers with higher negotiated prices) or market share (among price-sensitive consumers). Hospitals may be more likely to adopt this strategy in the absence of pressure to disclose from competitors; research has found that hospital compliance rates are associated with those of their market peers.³⁰ To promote compliance, in January 2022, the Centers for Medicare & Medicaid Services increased penalties for nondisclosure from a fixed maximum of $300 per day to a maximum based on hospital bed size up to $5500 per day, amounting to more than $2 million per year. This policy change was intended to ensure that penalty levels were meaningful for larger hospitals.³¹ As of December 2021, the Centers for Medicare & Medicaid Services had issued warning letters to 335 noncompliant hospitals and requested 98 hospitals to provide corrective action plans.³² The agency, however, had not levied any monetary penalties on noncompliant hospitals.

Hospitals may earn greater revenue per unit from cancer therapies than the pharmaceutical companies that manufactured them. Our estimates of hospital markup exceed those of a prior claims-based, industry-sponsored analysis that did not account for drug purchasing discounts from the 340B Drug Pricing Program.⁹ Although there are disagreements about whether the resource-intensive process of drug development justifies high prices set by pharmaceutical companies,³³ hospitals that administer cancer drugs and inflate their prices do not create additional value. Cancer centers may rely on profits from administering cancer therapy to cross-subsidize other activities. For example, proponents of the 340B Drug Pricing Program contend that the profits earned through the administration of discounted drugs cross-subsidize the care of patients with low incomes. However, there is evidence that the program instead provides incentives for hospitals to increase market share among patients with private insurance.^34-36

Hospital price variation for cancer therapies cannot solely be explained by differences in input costs, such as the costs of drug acquisition, facilities, and clinician labor. We found that cancer therapy prices varied considerably between payers at the same hospital. This variation may reflect differences in hospitals’ ability to negotiate favorable prices with insurers based on relative market power.²⁰ To curb potential abuses of market power for drugs administered in outpatient settings, policy makers could consider implementing backstop prices, which are commercial price caps set relative to local market rates.³⁷ Alternatively, regulators and payers could offer incentives for hospitals to administer cost-effective cancer therapies by adopting global budget payment models. However, when revenues are fixed under a global budget, some hospitals (eg, community or rural institutions) may have less financial capacity to withstand increases in drug acquisition costs if pharmaceutical companies increase drug prices.^38,39

This study has limitations. First, the results may not be generalizable to nononcology drugs or other types of hospitals because NCI-designated cancer centers likely have greater market power than many other hospitals when negotiating with private payers. Second, we may have underestimated prices over the treatment course because the calculations assumed there was no waste from dose vials.⁴⁰ Third, the findings may underestimate the extent of price markups and variations if centers chose not to disclose any drug prices or selected drug prices to prevent public concern about price-gouging.⁴¹ Fourth, we were unable to ascertain the proportion of patients at each hospital covered by specific insurance plans; thus, the findings may not accurately reflect average prices weighted by payer and plan mix.²⁹ Fifth, we were unable to estimate the extent to which negotiated prices were associated with patient out-of-pocket costs, which depend on plan benefit design; for example, higher prices would not directly translate to higher costs for patients who have already reached out-of-pocket spending caps. Sixth, our analysis would be limited if the prices disclosed by cancer centers are inaccurate.

This cross-sectional study found that less than half of NCI-designated cancer centers publicly disclose payer-specific prices for at least 1 top-selling cancer therapy drug, and many substantially mark up the prices of parenteral cancer therapies administered to patients with private insurance. Negotiated prices varied widely across cancer centers and within centers between payers. These findings suggest that, to reduce the financial burden of cancer treatment for patients, public policies may be instituted to discourage or prevent excessive hospital price markups on parenteral chemotherapeutics.

Accepted for Publication: February 24, 2022.

Published Online: April 18, 2022. doi:10.1001/jamainternmed.2022.1022

Corresponding Author: Roy Xiao, MD, MS, Department of Otolaryngology–Head and Neck Surgery, Massachusetts Eye and Ear, 243 Charles St, Boston, MA 02114 (roy_xiao@meei.harvard.edu).

Author Contributions: Dr Xiao had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Xiao, Ross, Gross, McWilliams, Sethi, Rathi.

Acquisition, analysis, or interpretation of data: Xiao, Ross, Dusetzina, McWilliams, Rathi.

Drafting of the manuscript: Xiao, Rathi.

Critical revision of the manuscript for important intellectual content: Xiao, Ross, Gross, Dusetzina, McWilliams, Sethi.

Statistical analysis: Xiao, McWilliams.

Administrative, technical, or material support: Sethi.

Supervision: Sethi, Rathi.

Conflict of Interest Disclosures: Dr Ross reported receiving grants from the US Food and Drug Administration, Johnson & Johnson, Medical Devices Innovation Consortium, the Association for Healthcare Research and Quality, National Institutes of Health National Heart, Lung, and Blood Institute, and Laura and John Arnold Foundation outside the submitted work; and Dr Ross is an expert witness at the request of Relator's attorneys, the Greene Law Firm, in a qui tam suit alleging violations of the False Claims Act and Anti-Kickback Statute against Biogen Inc. Dr Gross reported receiving grants from the National Comprehensive Cancer Network (AstraZeneca), Johnson & Johnson support for new models of clinical trial data sharing, and personal fees from Genentech Support for conducting studies with their group of molecular profiling use in real-world practice outside the submitted work. Dr Dusetzina reported receiving grants from the Commonwealth Fund, The Leukemia & Lymphoma Society, Arnold Ventures, and Robert Wood Johnson Foundation; honorarium from West Health and Institute for Clinical and Economic Review for advisory roles; consulting fees from the National Academy of State Health Policy outside the submitted work; and compensation by the Medicare Payment Advisory Commission as a government employee. Dr McWilliams reported receiving grants from Arnold Ventures during the conduct of the study; consultant fees from Blue Cross Blue Shield of North Carolina; and served as a paid senior adviser to the Center for Medicare and Medicaid Innovation. Dr Rathi reported employment by F-Prime Capital to identify and qualify investment opportunities in early-stage life sciences companies. No other disclosures were reported.

Disclaimer: The views presented herein are those of the authors and do not reflect those of the Medicare Payment Advisory Commission, Center for Medicare and Medicaid Innovation, and Arnold Ventures.