• Center on Health Equity & Access
  • Clinical
  • Health Care Cost
  • Health Care Delivery
  • Insurance
  • Policy
  • Technology
  • Value-Based Care

Cross-Validation of Insurer and Hospital Price Transparency Data

Publication
Article
The American Journal of Managed CareAugust 2024
Volume 30
Issue 8
Pages: e247-e250

Real-world cross-validation of insurer and hospital price transparency data finds low overlap but high concordance between data sources.

ABSTRACT

Given recent congressional interest in codifying price transparency regulations, it is important to understand the extent to which newly available price transparency data capture true underlying procedure-level prices. To that end, we compared the prices for maternity services negotiated between a large payer and 26 hospitals in Mississippi across 2 separate price transparency data sources: payer and hospital. The degree of file overlap is low, with only 16.3% of hospital–billing code observations appearing in both data sources. However, for the observations that overlap, pricing concordance is high: Corresponding prices have a correlation coefficient of 0.975, 77.4% match to the penny, and 84.4% are within 10%. Exact price matching rates are greater than 90% for 3 of the 4 service lines included in this study. Taken together, these results suggest that although administrative misalignment exists between payers and hospitals, there is a measure of signal amid the price transparency noise.

Am J Manag Care. 2024;30(8):e247-e250. https://doi.org/10.37765/ajmc.2024.89594

_____

Takeaway Points

Real-world analyses of hospital and insurer price transparency data find low overlap but high concordance between the 2 data sources.

  • Findings suggest that despite compliance issues from insurers and hospitals, the data being put forth are valid.
  • Regulatory efforts to decrease data redundancy on the part of insurers would greatly improve the usability of insurer price transparency data.
  • Insurer and hospital price transparency data are promising free and public sources of previously confidential information on health care prices. However, real-world utility of these data sources is limited by noncompliance on the part of insurers and hospitals.

_____

There has been a growing and sustained bipartisan policy interest in health services price transparency. Since January 1, 2021, almost all hospitals in the US have been required by federal regulations to publicly post machine-readable data files consisting of their “standard charges.”1 These standard charge data files must contain, among other things, the payer-specific negotiated rates for all items and services the hospital provides. Additionally, since July 1, 2022, the Transparency in Coverage (TIC) federal rule has mandated most group health plans and insurance issuers to post machine-readable data files consisting of the in-network rates for all covered services and all providers with which the payer contracts.2

These rules are a large step forward in terms of price transparency of health services, and a growing body of research has investigated both compliance with the regulation and price variation apparent in the standard charge data files.3-8 Additionally, although the TIC data files are complex, researchers have begun to systematically investigate the extent to which TIC networks overlap across large payers and to which TIC pricing data overlap with claims data, as well as the variation contained in the pricing files.9-11

These files have been costly to produce. CMS estimated a first-year burden of $71,415,397 with a subsequent-year burden of $21,672,502 for hospital price transparency files,1 and a first-year burden of $2,024,117,160 with a second-year burden of $414,728,136 for the in-network files of the TIC data.2 Regulators and policy makers have high hopes for price transparency information, specifically that it will “enable health care consumers to make more informed decisions, increase market competition, and ultimately drive down the cost of health care services.”1 Given the costs entailed in operationalizing price transparency regulations, as well as stakeholder aspirations for these price transparency data, it is of the utmost importance to understand whether these data sources are accurate.

In this study, we assess the accuracy of these data sources by examining their overlap and concordance for maternal health services. Although each source of price transparency data covers different aspects of the health services delivery system, the data sets overlap in 1 key way: They must both contain the negotiated rates between providers and payers. We compared TIC prices and hospital standard charge prices for maternity services for 26 hospitals in Mississippi and Blue Cross & Blue Shield of Mississippi (BCBS MS), the state’s dominant commercial insurer. Although prior research has assessed concordance between telephone-based and internet-based price estimates from hospitals,12 this is the first study to systematically assess price overlap and concordance between these 2 new sources of price transparency data.

STUDY DATA AND METHODS

Addressing the maternal health crisis is a top priority of HHS13; thus, we focused on pricing data for maternity-related services. Specifically, we used Current Procedural Terminology (CPT) codes 59000 to 59899, as well as codes for maternity case ultrasounds, fetal echocardiography, prenatal genetic and chromosomal metabolic testing, the analysis of fetal cells and amniotic fluid, and other newborn care. Additional detail is listed in eAppendix A (eAppendices available at ajmc.com).

We downloaded and processed the October 2023 BCBS MS TIC file and extracted the prices for the CPT billing codes presented in eAppendix A. Additional information on the source of the TIC data is presented in eAppendix B. Because public payers are not required to comply with TIC regulations, we restricted attention to the BCBS MS commercial in-network file.2 Although the TIC data contain both institutional prices and professional prices, we retained only institutional prices for this study to maximize the comparability of the 2 price transparency data sources. Within this file, billing codes for maternity services are provided at 52 hospitals. We sampled 26 of these hospitals and collected and processed the hospital-level standard charge data in October to November 2023. Hospital characteristics are derived from the 2023 quarter 3 CMS provider of service files.14 As with the TIC data, we retained only institutional prices for the hospital pricing data. Additionally, although the hospital pricing files contain negotiated rates for many different plans, we retained only negotiated rates for BCBS MS commercial plans. Additional detail on study hospitals is in eAppendix C and on data processing in eAppendix D.

We compared these 2 pricing data sources in 2 ways. First, we assessed the extent to which the files overlap in terms of the hospital–billing code combinations they contain. Then, for the observations that overlapped across the 2 sources, we assessed pricing concordance; that is, the degree to which the prices matched across data sources. Standard charge data can contain several negotiated prices for a given hospital–billing code combination, so we formed all pairwise combinations of hospital price and TIC price data for each hospital–billing code combination. For both sample overlap and price concordance, we examined prices overall and decomposed to service line. This study was exempt from review by the University of Maryland, Baltimore County Institutional Review Board.

RESULTS

Our TIC pricing sample consisted of 832 unique hospital–billing code observations: 32 codes for each of the 26 study hospitals. In contrast, our hospital pricing sample consisted of 1027 hospital–billing code observations, comprising 595 unique hospital–billing code combinations of 91 billing codes at the 26 study hospitals.

We found that 200 hospital–billing code combinations appear in both data sources, for an overall overlap rate of 16.3%. Only 24% of the TIC prices appear in the hospital data, and 33.6% of the hospital prices appear in the TIC data. Across all services included in this study, there is only 1 source of pricing data for 83.7% of observations.

The degree of overlap varied significantly by service line. Billing codes in the 59000 to 59899 and 90000 to 99999 ranges had little overlap: 8.0% and 5.7% of observations in these service lines appear in both data sources, respectively. However, billing codes in the 80000 to 89999 range appear in both data sources for 25.4% of observations, and billing codes in the 70000 to 79999 range appear in both data sources for 28.4% of observations. Notably, there is total overlap for hospital price transparency prices for the 80000 to 89999 service line: All the observations from the hospital pricing data also appear in the TIC data. Results are presented in Table 1.

For the hospital–billing code combinations that match, the price concordance is high. Results are displayed in Table 2. The correlation coefficient across hospital prices and TIC prices is 0.975, 77.4% of the pricing dyads match to the penny, and 84.4% match within 10%. As above, there is significant heterogeneity within service line. For service lines 59000 to 59899, 80000 to 89999, and 90000 to 99999, exact match rates are 93.2%, 92.8%, and 100%, respectively. For service line 70000 to 79999, however, the exact match rate is only 54.4%.

DISCUSSION

In this study, we sought to cross-validate the 2 primary sources of health services price transparency data—TIC data and hospital standard charge data—for a sample of maternity services. We document low levels of overlap: Only 16.3% of maternal health services hospital–billing code combinations appeared in both data sources. This is consistent with other research on the TIC files documenting high levels of data redundancy and “zombie” codes: There are many prices present in the TIC files corresponding to procedures that would never actually occur due to licensure or privileges—for example, a cardiologist having a negotiated price for childbirth services.9

The low overlap rates between the 2 sources of price transparency data are striking. In these instances, one data source records a negotiated rate between BCBS MS and a particular hospital, whereas another does not. For example, for Anderson Regional Medical Center - South, the TIC data contain a price for billing code 76827, whereas the price transparency data from that hospital do not; however, the price transparency data from that hospital contain a price for billing code 76819, whereas the TIC data do not. Both data sources—TIC and hospital—are intended to convey pricing information that reflect underlying, mutually agreed-upon contractual arrangements. Thus, the relatively low rates of file overlap potentially indicate substantial administrative misalignment between payers and hospitals. We note that of the overall pool of maternity services used in this study, 5 billing codes appear in the list of CMS “shoppable services.”1 Of these 5 billing codes, 4 appear in both the hospital price transparency and TIC data sources. This suggests that required reporting from CMS may mitigate the extent of administrative misalignment.

The higher rates of overlap in certain code segments may reflect additional pricing dynamics. For example, the relatively high rates of overlap for both imaging and laboratory services (28.4% and 25.4%, respectively) may reflect that these types of standardized services may be negotiated nationally (eg, as a percentage of the Medicare fee schedule), whereas other services may be negotiated more on a case-by-case basis. As a result, it may be more straightforward for both payers and hospitals to encode the underlying prices for imaging and laboratory services in price transparency files than it is for other service lines. An analysis of the drivers of such misalignment is beyond the scope of this study, but differing interpretation of underlying contracts could be a potential cause.

Limitations

This study has 2 primary limitations. First, by design, we collected prices from one payer, and for a sample of hospitals, and for a sample of procedure codes. Thus, we acknowledge that these results may have limited generalizability to other states, payers, or procedures.

We believe that, even as a case study, these results are meaningful given the health services landscape in Mississippi. BCBS MS is the dominant commercial payer in Mississippi, with 279,112 covered lives across the individual, small-group, and large-group markets; it also provides administrative services for 632,744 individuals with self-funded coverage.15 This is large relative to other commercial payers in the state: The Herfindahl-Hirschman Index for BCBS MS is 3555, indicating high concentration, and, as of 2021, it had 55% market share; UnitedHealth Group, the second-largest insurer, had 17% market share.16 Additionally, this payer’s negotiated rates have recently been the subject of both contracting disputes and state legislation: BCBS MS and the University of Mississippi Medical Center—the state’s sole academic medical center—spent much of 2022 in a contracting dispute, which was resolved in December 2022, and the state legislature subsequently passed SB2224 in 2023, which would have empowered the state insurance commissioner to “examine and address any inequalities regarding provider reimbursement rates paid by an insurer.”17,18 The governor vetoed the bill.18

Moreover, we collected pricing data for 26 hospitals in Mississippi, which represent 50% of the hospitals for which a price is reported in the BCBS MS TIC file. The hospitals used in this study are all short-term or critical access hospitals, and they account for 46.0% of beds across all of Mississippi’s active short-term or critical access hospitals. Additionally, although we focused only on billing codes related to maternal health, this is a policy-relevant set of procedure codes. Maternal mortality in Mississippi was second highest in the nation from 2018 to 2021,19 and its infant mortality rate was highest in the nation in 2020.20

More broadly, we hope other researchers and regulators conduct similar analyses using other states, payers, hospitals, and sets of procedures using the methodology outlined in this study. Recent research has investigated the high-level concordance among TIC pricing data, hospital pricing data, and commercial claims, but these are not matched to the payer-provider level; thus, more research is needed to investigate the accuracy of these new and promising price transparency sources.21

The second limitation relates to the nature of hospital standard charge data. These data are currently relatively unstandardized and, as such, require substantial parsing before they are analytically tractable. We list the details of our cleaning and filtering steps in eAppendix D; however, we recognize that limitations of the hospital price transparency data render direct comparisons challenging. Upcoming changes to the hospital price transparency data standards should make future comparisons more straightforward.22 Additionally, although third-party data vendors play an important role in the research ecosystem, we believe researchers should, whenever possible, work directly with the raw, underlying price transparency data to mitigate the possibility of unnecessary data filtering or standardization.

CONCLUSIONS

As the policy drive for price transparency continues, it is important to assess the usability and validity of the novel data being generated by these efforts. Policy goals can be realized only if, among other things, the pricing data issued by insurers and hospitals are accurate. Although there is low overlap between insurer and hospital data, the high degree of price concordance is encouraging: When the files contain the same hospital–billing code combinations, the prices tend to match. This suggests that a measure of signal exists within a relatively large pool of noise in the price transparency data writ large. As the research community takes stock of the tremendous amount of new, freely available price transparency data, we encourage future researchers and regulators to continue to assess the overlap and concordance of these 2 promising new data sources.

Author Affiliations: The Hilltop Institute, University of Maryland, Baltimore County (MAH, MCM), Baltimore, MD.

Source of Funding: This work was funded by Arnold Ventures.

Author Disclosures: The authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (MAH, MCM); acquisition of data (MAH, MCM); analysis and interpretation of data (MAH, MCM); drafting of the manuscript (MAH, MCM); critical revision of the manuscript for important intellectual content (MAH, MCM); and obtaining funding (MAH, MCM).

Address Correspondence to: Morgane C. Mouslim, DVM, ScM, The Hilltop Institute, University of Maryland, Baltimore County, 1000 Hilltop Circle, Sondheim Hall, 3rd Floor, Baltimore, MD 21250. Email: mmouslim@hilltop.umbc.edu.

REFERENCES

1. CMS, HHS. Medicare and Medicaid programs: CY 2020 hospital outpatient PPS policy changes and payment rates and ambulatory surgical center payment system policy changes and payment rates: price transparency requirements for hospitals to make standard charges public. Fed Regist. 2019;84(229):65524-65606.

2. Internal Revenue Service, Department of the Treasury; Employee Benefits Security Administration, Department of Labor; CMS, HHS. Transparency in Coverage. Fed Regist. 2020;85(219):72158-72310.

3. Henderson MA, Mouslim MC. Low compliance from big hospitals on CMS’s hospital price transparency rule. Health Affairs Forefront. March 16, 2021. Accessed November 15, 2023. https://www.healthaffairs.org/content/forefront/low-compliance-big-hospitals-cms-s-hospital-price-transparency-rule

4. Henderson M, Mouslim M. Assessing compliance with hospital price transparency over time. J Gen Intern Med. 2023;38(9):2218-2219. doi:10.1007/s11606-022-08020-3

5. Jiang JX, Polsky D, Littlejohn J, Wang Y, Zare H, Bai G. Factors associated with compliance to the hospital price transparency final rule: a national landscape study. J Gen Intern Med. 2022;37(14):3577-3584. doi:10.1007/s11606-021-07237-y

6. Henderson MA, Mouslim MC. Hospital and regional characteristics associated with emergency department facility fee cash pricing. Health Aff (Millwood). 2022;41(7):1029-1035. doi:10.1377/hlthaff.2022.00045

7. Wang Y, Meiselbach MK, Cox JS, Anderson GF, Bai G. The relationships among cash prices, negotiated rates, and chargemaster prices for shoppable hospital services. Health Aff (Millwood). 2023;42(4):516-525. doi:10.1377/hlthaff.2022.00977

8. Jiang J, Forman HP, Gupta S, Bai G. Price variability for common radiology services within US hospitals. Radiology. 2023;306(3):e221815. doi:10.1148/radiol.221815

9. Muhlestein D. Commercial insurer price transparency: a comparison of four national payers. Health Affairs Forefront. May 25, 2023. Accessed November 15, 2023. https://www.healthaffairs.org/content/forefront/commercial-insurer-price-transparency-comparison-four-national-payers

10. Wilson A, de Brantes F. Real world validation of payer pricing files: policy implications. Health Affairs Forefront. September 27, 2023. Accessed November 15, 2023. https://www.healthaffairs.org/content/forefront/real-world-validation-payer-pricing-files-policy-implications

11. Chartock BL, Simon K, Whaley CM. Transparency in Coverage data and variation in prices for common health care services. JAMA Health Forum. 2023;4(10):e233663. doi:10.1001/jamahealthforum.2023.3663

12. Thomas M, Flaherty J, Wang J, et al. Comparison of hospital online price and telephone price for shoppable services. JAMA Intern Med. 2023;183(11):1214-1220. doi:10.1001/jamainternmed.2023.4753

13. Strengthening maternal health. HHS. Updated May 14, 2024. Accessed November 15, 2023. https://www.hhs.gov/healthcare/maternal-health/index.html

14. Provider of services file – hospital & non-hospital facilities. CMS. Updated April 11, 2024. Accessed November 15, 2023. https://data.cms.gov/provider-characteristics/hospitals-and-other-facilities/provider-of-services-file-hospital-non-hospital-facilities

15. Medical loss data ratio and system resources: 2021 reporting year. Updated February 23, 2024. Accessed November 15, 2023. https://www.cms.gov/marketplace/resources/data/medical-loss-ratio-data-systems-resources

16. Competition in Health Insurance: A Comprehensive Study of U.S. Markets. American Medical Association; 2022. Accessed November 15, 2023. https://web.archive.org/web/20230913041301/https://www.ama-assn.org/system/files/competition-health-insurance-us-markets.pdf

17. Emerson J. BCBS Mississippi, UMMC end bitter dispute. Becker’s Payer Issues. December 19, 2022. Accessed November 15, 2023. https://www.beckerspayer.com/contracting/bcbs-mississippi-ummc-end-bitter-dispute.html

18. Bose D. Reeves vetoes health insurance bills that experts, watchdogs say would help consumers. Mississippi Today. March 22, 2023. Accessed November 15, 2023. https://mississippitoday.org/2023/03/22/reeves-vetoes-health-insurance-bills/

19. State health facts: maternal deaths and mortality rates per 100,000 live births. KFF. Accessed November 15, 2023. https://www.kff.org/other/state-indicator/maternal-deaths-and-mortality-rates-per-100000-live-births/

20. State health facts: total infant deaths. KFF. Accessed November 15, 2023. https://www.kff.org/other/state-indicator/infant-death-rate/

21. Wang Y, Meiselbach M, Anderson GF, Bai G. Hospital pricing information consistent between Transparency-in-Coverage data and other commercial data sources. Health Affairs Forefront. November 9, 2023. Accessed November 15, 2023. https://www.healthaffairs.org/content/forefront/hospital-pricing-information-consistent-between-transparency-coverage-data-and-other

22. CMS makes hospital prices more transparent and expands access to behavioral health care. News release. HHS; November 2, 2023. Accessed November 15, 2023. https://www.hhs.gov/about/news/2023/11/02/cms-makes-hospital-prices-more-transparent-expands-access-behavioral-health-care.html

Related Videos
Matias Sanchez, MD
James Chambers, PhD
dr carol regueiro
dr carol regueiro
Screenshot of Adam Colborn, JD during an interview
dr carol regueiro
Screenshot of an interview with Adam Colborn, JD
Screenshot of an interview with James Chambers, PhD
Related Content
© 2024 MJH Life Sciences
AJMC®
All rights reserved.