Both payers and hospitals are now publishing comprehensive negotiated rate data, an ongoing shift that opens up new possibilities for benchmarking, contracting, and network design. Unlike insurance claims data, which is lagged, has a high degree of noise (from various pieces of billing logic), and rarely identifies both payers and providers, price transparency machine-readable files (MRFs) offer a new, up-to-date, and unfiltered window into healthcare spending and costs. 

Just like any data source, however, price transparency data has its own challenges. Hundreds of payers and thousands of hospitals generate data independently. While a CMS schema places guardrails (e.g., standard columns, guidance on methodology) on what they can publish, performing apples-to-apples comparisons can still be tricky when using data from so many independent sources. 

KFF recently posted some of the core challenges here - for example, one payer may use 4-digit revenue codes to capture outpatient pricing whereas another payer may use 5-digit CPT codes.To make the most of this data, Serif Health combines MRFs with data science and normalization, claims data, and contracting logic to support accurate benchmarking and analysis.

Focus Area: Inpatient Hospital Pricing

In today’s blog, we’ll be discussing analytics we’ve performed on hospital inpatient data to help:

1. Can an employer build a better network?
   Employers build their own provider networks and directly contract with high-volume hospitals at more favorable terms than what a standard carrier network can offer

2. Is this hospital’s reimbursement competitive?
   Hospitals & payers both understand if they have fair reimbursement terms compared to peers in their market
   

3. How do plan options compare?
   Brokers and advisors can compare different plans on their overall affordability and individual hospital coverage

We’ll focus on a sample regional market with a mix of academic medical centers and regional hospital systems: highlighting how contracting and negotiated rates can vary substantially even within the same metro area.The first step here is to figure out the contracting methodology any hospital is using to generate their fee schedule. 

Benchmarking Inpatient Pricing at Northwestern Memorial Hospital

Let’s take a look at DRG 27 at negotiated rates reported for Northwestern Memorial Hospital. A given contracting methodology is often reported in  the hospital’s own price transparency disclosure and any notes, such as this one below, they include in their reporting:

From Northwestern Memorial Hospital’s disclosure:
MS - DRG: 27 - Craniotomy And Endovascular Intracranial Procedures Without Cc/McC
Plan-name: UHC PPO [158027]

For those who want this in lay person English, this breaks down Northwestern Memorial’s inpatient contract is based on:

1. A consistent contractual base rate
   (if you are unfamiliar with base rates, check out this blog)
2. That contractual rate = 16,633.00
3. The base weight used for DRG 27 = 2.4329
4. Contractual rate * base weight = $40,466.43 negotiated contracted allowed

Using these terms, we can create a clean comparison for which payer has the best inpatient contract with Northwestern Memorial:

Key insight: Looks like Cigna is approximately 15% less expensive than UHC for the same service at the same facility.

Now, this comparison worked because Northwestern Memorial used the same DRG-based contracting methodology across payers. Let’s say, however, instead of comparing 1 hospital across several payers you wanted to compare several hospitals against each other for a given network. 

This is a very common use case we encounter: for example, for employers wanting to tier their network for preferred low-cost, high-quality providers or navigate their members to their best facility. 

New York City Case Study:

For illustrative purposes, we can use New York City as our market of interest and look at the following hospitals for inpatient rates negotiated with Anthem / Empire’s specific plan for the 32BJ labor union:

1. New York Presbyterian
2. Mt. Sinai
3. NYU Langone Tisch
4. Lenox Hill Hospital

Again, to begin properly, we have to understand the contracting approach used by each facility:

For instance, within the payer files Anthem publishes, they used 3-digit MS-DRGs for 3 of the above facilities and 4-digit APR-DRGs for New York Presbyterian. Conversely, looking at the hospital files each of these facilities published separately reveals only MS-DRGs:

From this review alone, we know we will have to cross-walk APR-DRGs to MS-DRGs to compare New York Presbyterian successfully to Mt. Sinai, NYU Langone Tisch, and Lenox Hill inside the payer MRF data.

Luckily, we have a natural experiment in the data that allows us to do this: we can find which MS-DRGs match to which APR-DRGs by comparing what New York Presbyterian posted in their own hospital MRF to Anthem’s payer MRF -> the outcome of such work is a table of ‘best matches’ like this:

The way to interpret any given row in the above is:

• For MS-DRG 293 - Heart failure and shock without significant complications or co-morbidities (CC/MCC), New York Presbyterian posted they have a contracted allowed of $54,256.
• To figure out which APR-DRG this ‘293’ corresponds to in Anthem’s payer data, our data science team looked at possible APR-DRGs whose descriptions had a semantic match - e.g., 194-1, 194-2, 194-3, 194-4 which all correspond to ‘heart failure’ - and then found the right APR-DRG severity level (denoted by suffix ‘-1’, ‘-2’, …) to map to based on which APR-DRG had the closest $ value. 

(While exact matches between hospital and payer-MRF dollar values are possible, typically some % difference exists due to differences in timing of when hospitals and payers release their data.)

This kind of work shows the unique value in combining hospital and payer price transparency data to generate reliable and validated reimbursement insights. After normalizing APR-DRGs vs. MS-DRGs, we can now proceed with baselining each facility’s fee schedule to Medicare across all inpatient services to generate topline results:

While exact magnitudes between hospital and payer MRF data may differ for many reasons (e.g., timing of disclosure, implant carve-outs, negotiated terms), directionally the findings align well across both sources:

• New York Presbyterian = Most expensive
• NYU Langone Tisch = 2nd
• Lenox Hill Hospital = 3rd
• Mt. Sinai Hospital = 4th

Conclusions

The examples above highlight three common insights when working with inpatient price transparency data:

• Commercially negotiated prices can vary dramatically—even in the same market.
  Some hospitals receive 2–3x higher reimbursement than peers for the same DRG.
  
• Valid comparisons require alignment on contracting logic.
  In our New York example, crosswalking MS-DRGs and APR-DRGs was essential to normalize and compare across payer and hospital data.
  
• Preferred pricing may not always reflect clinical value.  
  ‍While some quality variation may justify higher reimbursement for certain systems, it's unlikely that any quality gap alone accounts for the 2–3x difference observed between, say, New York Presbyterian and other New York facilities.

From work like this, 

1. Plan administrators can create high-performance networks focused on the most cost-effective providers serving a region

2. Employers and other payers can provide incentives (e.g., $0 co-pay) for members to go to one facility vs. another or contract directly with hospitals

3. Hospitals and other providers can determine if their reimbursement terms are in line with peers

If you are interested in chatting more about price transparency analytics, please reach out to hello@serifhealth.com or book time here. Also feel free to check out our sample data on our web platform Signal.

Bill Pajerowski contributed research, editing, and review for this article.