Negotiated Rates for Surgical Cancer Care in the Era of Price Transparency—Prices Reflect Market Competition

Abstract

Objective:

To measure commercial price variation for cancer surgery within and across hospitals.

Summary Background Data:

Surgical care for solid organ tumors is costly and negotiated commercial rates have been hidden from public view. The Hospital Price Transparency Rule, enacted in 2021, requires all hospitals list their negotiated rates on their website, thus opening the door for an examination of pricing for cancer surgery.

Methods:

This was a cross-sectional study using 2021 negotiated price data disclosed US hospitals for the ten most common cancers treated with surgery. Price variation was measured using within- and across-hospital ratios. Commercial rates relative to cancer center designation and the Herfindahl-Hirschman Index at the facility level were evaluated with mixed effects linear regression with random intercepts per procedural code.

Results:

495,200 unique commercial rates from 2,232 hospitals resulted for the ten most common solid-organ tumor cancers. Gynecologic cancer operations had the highest median rates at $6,035.8/operation compared to bladder cancer surgery at $3,431.0/operation. Compared to competitive markets, moderately and highly concentrated markets were associated with significantly higher rates (HHI 1501–2500, coefficient $513.6, 95% CI, $295.5 - $731.7; HHI >2500, coefficient $1,115.5, 95% CI, $913.7 - $1,317.2). National Cancer Institute designation was associated with higher rates, coefficient $3,451.9 (95% CI, $2,853.2 – $4,050.7).

Conclusions:

Commercial payer-negotiated prices for the surgical management of 10 common, solid-tumor malignancies varied widely both within and across hospitals. Higher rates were observed in less competitive markets. Future efforts should facilitate price competition and limit health market concentration.

Mini-abstract

In this evaluation of 2,232 hospitals and 495,200 unique commercial rates, there was substantial variation both within and across hospitals for surgical procedures typically employed for cancer care. Higher commercial rates were observed in more consolidated (less competitive) healthcare markets. Policies to encourage competition may lower prices.

Introduction

Cancer care has been in the spotlight for the past few decades due to rising prices, higher patient cost-sharing, and disproportionate rates of bankruptcy among cancer patients.^1–7 Understanding the exponential growth in the cost of cancer care and the associated “financial toxicity”⁸ requires not only identifying key drivers of this cost, but also evaluating characteristics associated with cost variation. For the surgical management of privately-insured patients with cancer, this includes characterizing factors associated with varying commercial insurance payments to hospitals for oncologic surgery.⁹ While it is favorable for insurers to incentivize and reward hospitals that provide higher value care with greater payments,^10,11 indiscriminate price variation represents an area to target for cost containment. Moreover, elevated charges by hospitals in some scenarios may reflect anticompetitive market behavior.^12,13

A recent executive order for price transparency requires hospitals to post their negotiated rates. This offers insight into pricing that was previously infeasible due to concealed rate information. Historically, the price of cancer care in the United States has approximated a “menu without prices,” as both physicians and patients engage in price-naïve medical transactions.¹⁴ Following the implementation of the Hospital Price Transparency Rule in January 2021, hospitals are required to disclose commercial payer-negotiated rates for their shoppable services.¹⁵ Theoretically, price transparency laws should lead to a reduction in the variation across rates, though the long-term effect on the average rate is unknown.¹⁶

The purpose of this study was to utilize newly accessible hospital price data to characterize national price variation in the surgical management of cancer based on regional market competitiveness and National Cancer Institute (NCI) Cancer Center designation. Based on the widespread nature of healthcare market imperfections, we hypothesized that oncologic operations would be characterized by significant price variation without performance-based correlations.

Methods

Data Sources

We performed a cross-sectional study of 2021 pricing data from Turquoise Health, a data service platform that mines hospital data warehouses for price disclosures.¹⁷ Turquoise Health data are drawn from 5,700 US hospitals (94% of 6,090 total hospitals registered with a unique American Hospital Association identification number), and have been previously validated and employed for scientific study.^18–25 A total of 162 Current Procedural Terminology (CPT) codes were utilized to query the Turquoise Health platform and extract pricing data for the facility fee associated with surgical treatment of 10 cancer types: bladder, breast, colorectal, esophageal, gynecologic (including uterine, ovarian, and cervical), lung, pancreatic, prostate, renal, and thyroid (see Supplemental Table 1 for full list of CPT codes). These cancer types were selected for being among the leading types of new solid tumor malignancies for men and women,²⁶ with specific CPT codes for surgical management. CPT codes represented procedures typically employed for cancer care but were not limited to the care of cancer diagnoses (i.e. some CPT codes are used for benign and malignant disease). All pricing data were normalized to account for variable nationwide input costs using the Geographic Adjustment Factor (GAF),²⁷ with the exception of Medicare rates which are intrinsically price-adjusted. Pricing data from Maryland were excluded given the state’s global budget program that controls prices for acute care hospital services.²⁸

Pricing data were merged at the hospital level with data from the 2021 Lown Institute Hospitals Index, a publicly-available dataset containing facility-level data and performance metrics for 3,709 general acute care US hospitals. The methodology for construction of the Index has been previously described.²⁹ Briefly, a safety-net hospital was defined as the 20% of hospitals with the highest proportion of patients eligible for Medicare and Medicaid. Non-profit status was identified from facility tax returns, and teaching hospital designation is nationally reported.

Pricing data were combined with 2020 Herfindahl-Hirschman Index (Herfindahl-Hirschman Index) data for 186 metro areas, complied by the Health Care Cost Institute.³⁰ These data were merged at the metro area level to yield estimates of healthcare market competition, whereby a larger Herfindahl-Hirschman Index indicates a more concentrated (i.e. less competitive) market.^31–33 The Herfindahl-Hirschman Index was categorized as indicative of markets with low concentration (≤1500), moderate concentration (1501–2500), or high concentration (>2500), based on US Department of Justice standards.³⁴ See Supplemental Figure 1 for an illustration of dataset construction.

Variables

Pricing data included commercial payer-negotiated rates at the plan level, public payer (Medicare and Medicaid) rates, self-pay rates, and list prices (i.e. chargemaster rates). Hospital-level variables included: designation as a NCI-Designated Cancer Center, which is a title given to 71 NCI-funded centers in the US based on excellence in research related to preventing, diagnosing, and treating cancer;³⁵ hospital size [small (6–99 beds), medium (100–199 beds), and large (200+ beds)]; safety-net status; profit status (for-profit hospital vs. non-profit hospital); and teaching status (major or minor teaching hospital vs. non-teaching hospital). In addition to Herfindahl-Hirschman Index, local geographic variables included geographic division based on US census definitions.³⁶

Statistical Analysis

Within-hospital ratios and across-hospital ratios of commercial rates per CPT code were calculated to quantify price variation.^37–39 For each CPT code at a given hospital, the within-hospital ratio (WHR) was calculated as the maximum commercial rate divided by the minimum commercial rate; these values were averaged to obtain a median WHR per CPT code, and then further averaged across all CPT codes for a given cancer type to obtain a median WHR per cancer type. The across-hospital ratio for a CPT code was calculated as the 90^th percentile GAF-normalized commercial rate divided by the 10^th percentile GAF-normalized rate across all hospitals. These values were similarly averaged across all CPT codes for a given cancer type to obtain a median AHR per cancer type. Two-sided t-tests were used to compare rates between payers and cancer types.

A mixed effects linear regression with random intercepts by CPT code was utilized to model the association between median GAF-normalized commercial rate and hospital-level factors: NCI status, hospital size, safety-net status, profit status, teaching status, location (US Census division), and HHI. Additional mixed effects linear regression models examined the association of median GAF-normalized commercial rate with Medicare and GAF-normalized Medicaid, self-pay, and list prices. A random intercept for hospital was included to adjust for inter-hospital variation. The likelihood ratio (LR) test was used to evaluate mixed effects model fitness in comparison to a simple linear regression model. All models employed complete case analysis (i.e., listwise deletion).

Analyses were conducted using Stata/SE Version 18.0 (StataCorp LLC). P-values of less than 0.05 were considered statistically significant. Model p-values were adjusted using the Bonferroni correction. Geographic mapping of GAF-normalized commercial rates across continental US Hospital Referral Regions (HRRs) was performed with Tableau (Tableau Software LLC).

Results

There were 2,232 unique hospitals with at least one commercial rate (88.0%, N=1,965), Medicare rate (63.0%, N=1,406), Medicaid rate (43.3%, N=967), self-pay rate (63.9%, N=1,427), or list price (65.0%, N=1,450). In terms of hospital-level characteristics, 1.1% (N=25) were National Cancer Institute Cancer Centers, 16.6% (N=278) were safety-net, 78.0% (N=1,307) were non-profit, and 39.3% (N=658) were teaching hospitals. The majority of hospitals were in urban areas (N=67.9%). The percentage of hospitals in high concentration markets (Herfindahl-Hirschman Index>2500, 35.4%) was similar to that of hospitals in low concentration markets (HHI≤1500, 38.7%). Additional hospital-level characteristics are summarized in Supplemental Table 2.

Figure 1 illustrates the median GAF-normalized rate per payer category (Medicaid, Medicare, commercial, self-pay, and list price) by cancer type, and for the most common CPT code for each cancer type (see Table 1 for most common CPT codes). For all cancer types and most common CPT codes, average commercial rates were significantly higher than average Medicaid rates (p<0.001) and Medicare rates (p<0.001), except for the most common CPT codes for esophageal cancer (41322) and pancreatic cancer (48140), in which the commercial rate did not significantly differ from the Medicare rate (Supplemental Table 3).

Figure 1. Median rates per payer for each cancer type (top) and most common CPT code per cancer type (bottom).

All rates except for Medicare rates are normalized by Geographic Adjustment Factor (GAF). 51030, cystostomy/cystectomy with destruction of lesion. 19301, partial mastectomy. 45171, transanal excision of rectal tumor. 43122, partial esophagectomy. 58551, laparoscopic hysterectomy (250g or less) with removal of tube(s) and/or ovary(s). 32666, unilateral thoracoscopy and wedge resection. 48140, distal subtotal pancreatectomy. 55866, laparoscopic radical prostatectomy. 50543, laparoscopic partial nephrectomy. 60220, unilateral total thyroid lobectomy. See Supplemental Table 1 or full descriptions of CPT codes. Order of codes in bottom panel correspond to order of disease categories in top panel.

Table 1.

CPT codes with most commonly disclosed commercial rates

Cancer Type	CPT Code	Abbreviated Descriptiona
Bladder	51030	Cystotomy with cryosurgical destruction of lesion
Breast	19301	Lumpectomy
Colorectal	45171	Transanal excision of rectal tumor
Esophageal	43122	Partial esophagectomy
Gynecologic	58571	Laparoscopic hysterectomy with salpingo-oophorectomy
Lung	32666	Thorascopic wedge resection
Pancreatic	48140	Subtotal pancreatectomy
Prostate	55866	Laparoscopic prostatectomy
Renal	50543	Laparoscopic partial nephrectomy
Thyroid	60220	Thyroid lobectomy

^aSee Supplemental Table 1 for full description

Within-hospital ratios and across-hospital ratios for GAF-normalized commercial rates by cancer type and most common CPT code are enumerated in Table 2 and illustrated in Figure 2, in descending order by rate (highest: gynecologic, median $6,035.8, interquartile range [IQR] $2,737.5 - $12,127.7; lowest: bladder, median $3,431.0, IQR $2,078.3 - $7,234.2). For cancer type, median within-hospital ratios ranged from 1.0 (IQR 1.0 – 1.4) for esophageal cancer to 2.3 (IQR 1.0 – 5.2) for breast cancer. Median across-hospital ratios ranged from 7.0 (IQR 6.1 – 7.2) for colorectal cancer to 10.0 (IQR 8.5 – 10.0) for breast cancer. For the most common CPT code by cancer type, median within-hospital ratios ranged from 1.2 (IQR 1.0 – 3.9) for CPT 43122: partial esophagectomy, to 3.0 (IQR 1.0 – 7.1) for CPT 60220: unilateral total thyroid lobectomy. Median across-hospital ratios ranged from 6.1 for CPT 45171: transanal excision of rectal tumor, to 10.8 for CPT 50543: laparoscopic partial nephrectomy.

Table 2.

Descriptive statistics for commercial payer-negotiated rates for cancer operations by cancer type (top) and most common CPT code per cancer operation (bottom)

	No. hospitals	No. CPT codesa	No. ratesb	GAF-normalized payer-negotiated rate, $c			Within-hospital ratio (IQR)d	Across-hospital ratio (IQR)e
				Median	p10	p90
Cancer type
Bladder	721	13	27,045	3,431.0	1,262.5	12,883.9	1.2 (1.0 – 4.0)	7.2 (6.8 – 7.4)
Breast	1,618	14	41,738	5,045.0	1,239.3	17,659.5	2.3 (1.0 – 5.2)	10.0 (8.5 – 10.0)
Colorectal	1,291	37	98,147	3,666.0	1,451.7	12,257.2	1.5 (1.0 – 3.9)	7.0 (6.1 – 7.2)
Esophageal	457	16	27,971	4,992.1	1,517.6	14,309.8	1.0 (1.0 – 4.0)	6.4 (6.3 – 6.6)
Gynecologic	1,640	28	114,920	6,035.8	1,477.7	22,174.1	2.0 (1.0 – 5.1)	9.6 (8.4 – 10.9)
Lung	601	14	29,407	3,792.0	1,464.0	14,235.1	1.2 (1.0 – 3.6)	8.2 (7.5 – 8.5)
Pancreatic	499	9	17,246	4,141.3	1,585.9	13,094.2	1.2 (1.0 – 3.8)	7.2 (6.3 – 7.2)
Prostate	1,001	10	24,487	3,938.9	1,325.3	19.073.0	1.4 (1.0 – 4.5)	8.8 (7.7 – 10.5)
Renal	965	11	30,486	4,225.8	1,359.9	18,187.8	1.5 (1.0 – 4.6)	9.0 (7.3 – 10.8)
Thyroid	1,364	9	47,753	5,947.3	1,436.6	19,693.6	2.1 (1.0 – 5.5)	9.3 (9.1 – 9.9)
Most common CPT
51030	640	-	3,520	4,080.6	756.7	12,927.7	1.7 (1.0 – 4.7)	6.8
19301	1,507	-	11,883	4,023.8	1,064.7	14,004.1	2.9 (1.3 – 5.9)	8.5
45171	1,037	-	6,943	3,181.8	942.3	9,607.3	2.3 (1.0 – 5.1)	6.1
43122	437	-	1,979	4,350.9	1,655.0	13,628.6	1.2 (1.0 – 3.9)	6.5
58571	1,330	-	8,899	8,619.2	1,628.9	27,393.2	2.7 (1.1 – 6.5)	10.1
32666	525	-	2,719	2,790.9	935.2	11,346.4	1.7 (1.0 – 4.1)	9.3
48140	456	-	2,188	3,309.3	1,491.9	12,506.8	1.5 (1.0 – 4.5)	7.2
55866	982	-	6,166	9,426.5	2,056.4	33,651.3	2.4 (1.0 – 6.8)	10.5
50543	848	-	5,143	9,241.3	2,106.4	29,730.9	2.0 (1.0 – 5.4)	10.8
60220	1,253	-	9,058	6,247.6	1,233.1	20,898.7	3.0 (1.0 – 7.1)	10.1

CPT, Current Procedural Terminology. IQR, interquartile range. GAF, Geographic Adjustment Factor. p90, 90^th percentile median rate. p10, 10^th percentile median rate.

^aNumber of CPT codes totals 161 out of 162 since one code for bladder cancer did not have any commercial rates in database.

^bThese values serve as the sample size for the corresponding percentile and ratio calculations.

^cCalculated based on the median payer-negotiated prices at each hospital across all codes for cancer type.

^dCalculated as the median of the median of the maximum payer-negotiated rate divided by the minimum payer-negotiated rate per code for each hospital.

^eCalculated as the median of the 90^th percentile median GAF-normalized rate divided by the 10^h percentile median GAF-normalized rate per code across all hospitals.

Figure 2. Within-hospital ratios and across-hospital ratios for commercial rate by cancer type (top) and most common CPT code per cancer type (bottom).

Listed in descending order by median GAF-normalized commercial rate (i.e gynecologic cancer has highest median commercial rate, bladder cancer has lowest). WHR, within-hospital ratio. AHR, across-hospital ratio. Excludes outliers. Order of codes in bottom panel correspond to order of disease categories in top panel.

In the adjusted analysis, higher median GAF-normalized commercial rates were significantly associated with NCI designation (β $3,451.9, 95% CI $2,853.2 - $4,050.7, p<0.001) and greater market concentration (compared to HHI≤1500: β $513.6 and 95% CI $295.5 - $731.7 for HHI 1501–2500, β $1,115.5 and 95% CI $913.7 - $1,317.2 for HHI >2500, p<0.001; Table 3). Commercial rates were negatively associated with medium (β −$1,858.7, 95% CI −$2,150.3 - −$1,567.0) and large (β −$2,049.8, 95% CI −$2,333.4 - −$1,766.2) bed size compared to small (p<0.001), non-profit hospitals (β −$526.6, 95% CI −$686.1 - −$367.0, p<0.001), and teaching hospitals (β −$851.3, 95% CI −$1,006.9 - −$695.7, p<0.001). There was no significant effect observed based on whether the hospital was designated as safety-net. Geographic variation is further illustrated in Figure 3 and Supplemental Figure 2.

Table 3.

Mixed effects linear regression model, with random intercept by CPT code, for GAF-normalized commercial rate (N=34,415)

	Coefficient/β ($)	95% Confidence Interval	p-value*
NCI-CC
No	Ref	-	-
Yes	3,451.9	2,853.2 – 4,050.7	<0.001
Size (bed count)
Small (6–99)	Ref	-	-
Medium (100–199)	−1,858.7	−2,150.3 – −1,567.0	<0.001
Large (200+)	−2,049.8	−2,333.4 – −1,766.2	<0.001
Safety-net hospital
No	Ref	-	-
Yes	−61.1	−291.7 – 169.4	1.000
Non-profit hospital
No	Ref	-	-
Yes	−526.6	−686.1 – -367.0	<0.001
Teaching hospital
No	Ref	-	-
Yes	−851.3	−1,006.9 – −695.7	<0.001
US Census division
New England	Ref	-	-
Middle Atlantic	−1,199.7	−1,887.1 – −512.3	0.007
East North Central	880.7	227.6 – 1,533.8	0.056
West North Central	326.9	−331.9 – 985.6	1.000
South Atlantic	3,353.4	2,724.4 – 3,982.4	<0.001
East South Central	−2,806.6	−3,465.3 – −2,147.9	<0.001
West South Central	−2,738.1	−3,373.1 – −2,103.0	<0.001
Mountain	−2,111.2	−2,755.1 – −1,467.3	<0.001
Pacific	1,396.3	745.1 – 2,047.5	<0.001
HHI
≤1500	Ref	-	-
1501–2500	513.6	295.5 – 731.7	<0.001
>2500	1,115.5	913.7 – 1,317.2	<0.001

^*Bonferroni correction applied.

GAF, Geographic Adjustment Factor. NCI-CC, National Cancer Institutes-Designated Cancer Center. HHI, Herfindahl-Hirschman Index. Ref, reference.

Figure 3. GAF-Normalized Commercial Rate by Hospital Referral Region (HRR).

GAF, Geographic Adjustment Factor. Grey areas reflect lack of disclosure (e.g. Oregon) or lack of hospitals (e.g. Wyoming). Images created with Tableau (Tableau Software LLC).

Mixed effects regression at the hospital level demonstrated positive associations of GAF-normalized commercial rates with Medicare rates (β $0.44, 95% confidence interval [CI] $0.41 – $0.46, p<0.001), GAF-normalized Medicaid rates (β $0.26, 95% CI $0.24 – $0.28, p<0.001), GAF-normalized self-pay rates (β $0.20, 95% CI $0.17 – $0.23, p<0.001), and GAF-normalized list price (β $0.10, 95% CI $0.08 – $0.12, p<0.001; Supplemental Table 4).

Discussion

Commercial payer-negotiated rates for surgery typically performed for cancer varied substantially both across and within hospitals. For the most common renal surgical procedure (laparoscopic partial nephrectomy), commercial rates across hospitals varied by a factor of nearly 11. For the most common thyroid surgical procedure (unilateral total thyroid lobectomy), contracted rates varied by a median factor of three for payers within the same hospital. In other words, these data suggest that a given commercial insurer is paying three times the rate that another insurer is paying for the same service at the same hospital. This may be problematic because higher commercial rates have been shown to translate into higher OOP spending for both commercially-insured patients and self-insured employers.⁴⁰

Prior studies on cancer and pricing confirm substantial variation yet have been limited in scope. Xiao et al. examined prices for thyroid cancer operations across 52 NCI-Designated Cancer Centers, finding across-hospital ratios of 6.4–7.2 for partial or total thyroidectomy.³⁸ These authors also examined parenteral cancer therapies at 27 NCI-Designated Cancer Centers, detecting within-hospital ratios of 1.8–2.5 and across-hospital ratios of 2.2–15.8.⁴¹ Chino et al. found large variability in negotiated rates for colonoscopy and radiotherapy among NCI-Designated Cancer Centers, with some centers pricing these services at up to eight times the Medicare maximum allowable rate.⁴² In addition, Agarwal et al. evaluated chargemasters (i.e. list prices) for prostate cancer radiotherapy at 52 NCI-Designated Cancer Centers, finding a more than 20-fold difference in list prices.⁴ These authors acknowledge that list prices are not the actual payments for services, yet cite prior work by Batty and Ippolito suggested that list prices were positively correlated with prices paid by patients and insurers.⁴³ Our study findings confirm this correlation, as GAF-normalized commercial rates and list prices were correlated with a coefficient of 0.10 (p<0.001). However, this correlation is of small magnitude and cannot match the accuracy of utilizing actual commercial prices to estimate variation as performed in this study.

Our findings support the multifactorial origins of price variation, namely that prices vary with market power and inconsistently mirror the value of care. In terms of market power, hospitals that were larger or in more concentrated healthcare markets (i.e. higher Herfindahl-Hirschman Index) were associated with higher commercial payer-negotiated rates on multivariable regressions. This supports existing literature that consolidation leads to price increases and that market power affects hospitals’ abilities to negotiate prices with commercial payers.^9,44 In addition, it is possible that price variation has a hyper-contextual component whereby specific hospitals and payers negotiate rates outside the lens of variables that we can measure on a national scale. In our study, commercial rates for the surgical management of breast cancer varied by a factor of 2.3 within hospitals, suggesting that factors beyond the quality of care and regional market power contribute to price variation. The geographic variation further confirms this suspicion. Detroit and San Francisco Bay Area metropolitan areas had significantly higher rates for all cancer types, while rates were lower for most of Texas. Other markets such as Atlanta metropolitan area have variable pricing based on the cancer type. At least for the San Francisco Bay Area, antitrust lawsuits as recently as 2021 have penalized a large health system that had engaged in unfair pricing. These types of behaviors should be captured by HHI; however, there are likely nuances within markets such as single health system exhibiting monopolistic behavior that can lead to additional price fluctuations.

NCI Cancer Center designation could translate to higher rates through two mechanisms. First, it could be a symbol of quality, thereby allowing hospitals to differentiate themselves from competitors and charge higher rates. Second, because obtaining the designation has a cost, this cost could translate to higher rates. Our results suggests that this special designation may provide increased leverage in contract negotiations. Cancer centers go through a rigorous vetting and verification process to achieve NCI designation. Seemingly, this would signal quality to payers who would recognize the services and processes of these cancer centers. However, our data are comprised of rates from only 35% (25 out of 72) of NCI-Designated Cancer Centers, and thus we hesitate to draw broad conclusions about quality based on pricing data from a minority of cancer centers. Looking at other performance metrics such as equity, our data suggest that higher rates do not correlate with higher equity; there was no significant correlation between commercial rates and safety-net hospitals, while non-profit and teaching hospitals were significantly associated with lower rates. Future studies should continue to explore the relationship between commercial rates and hospital performance metrics.

Limitations of this study include non-compliance with disclosure requirements specified in the Hospital Price Transparency Rule. An estimated 50–83% of US hospitals were noncompliant at the end of 2021,^38,45 including approximately 80% of NCI-Designated Cancer Centers. Compliance with disclosure and availability of pricing data is likely to improve given the recent increase in penalties for nondisclosure, and the implementation of the Transparency in Coverage Rule in July 2022 that requires health plans to disclose prices for covered services.⁴⁶ The operations that we have included in our analysis are performed in the surgical management of the respective cancers but are not limited to cancer treatment. Lastly, due to the nature of our merged dataset, some analyses had missing data that were managed with listwise deletion. Specifically, in the case of HHI, we limited analyses to markets where information was available as this was more appropriate than other methods (e.g. multiple imputation) in this context.

Proposed solutions to the high cost of cancer care have included global budget payment models, implementation of backstop prices (i.e. commercial price caps), and greater popularization of price transparency.^4,41,44 The Hospital Price Transparency Rule does not specify that price data should be released in a consumer-friendly means. Instead, the data are released in the form of machine-readable files.²³ We cannot estimate exact OOP expenditures based on commercial price, since this varies based on the patient’s health plan.⁴¹ If the ultimate goal is to enable patients to budget for expenses, compare prices across hospitals, and engage in informed decision-making with their oncology team, then pricing needs to be transparent to both payers and patients. As currently structured, the Hospital Price Transparency Rule may favor insurers, and more work is needed to get these data to patients in the appropriate prospective fashion. The data in this study are from 2021, which is the year of implementation of disclosure requirements for hospitals, and prior to the implementation of analogous requirements for health plans. Thus, it will be prudent to re-examine cancer care prices in a few years once prices have had time to respond to market forces. In the meantime, since higher prices are associated with more concentrated markets, competition should be encouraged between health systems and attempts for consolidation should be tempered. The latter solution includes policy evolution and legal action (e.g. antitrust lawsuits from the Department of Justice) to thoroughly vet mergers of health systems and consider splitting large health systems in markets with monopolistic behavior.

Conclusion

The commercial payer-negotiated price of operations performed for the management of 10 common, solid tumor malignancies varied widely both within and across hospitals. Higher rates were also associated with factors indicating monopolistic hospital pricing behavior (i.e., higher Herfindahl-Hirschman Index). To slow and potentially reverse the rising cost of cancer care, future efforts should facilitate price transparency and limit health market concentration.