Health care prices in the United States have historically been hidden from the public hindering meaningful analysis of health care markets and prices. This matters because prices impact access to care and are a major driver of high health care spending [1,2].

The information landscape began to change with first price transparency rule, Hospital Price Transparency (HPT) in 2021 which required hospitals to disclose prices of a select set of services. Then, in 2022, the Transparency in Coverage (TiC) rule began to require insurers to release negotiated rates for all covered services.

These datasets have proven valuable in shedding light on variation in health care prices. HPT data has been used to demonstrate large variation in prices for urological procedures [3], and radiology services [4], among others. Recent studies have begun to exploit more comprehensive TiC data and show wide variation in prices for inpatient and outpatient services [5], within hospitals for commonly provided services [6], among others.

We assess the within-provider variation in prices for the top 20 most commonly services in an office setting and find that the average difference between the minimum and maximum rates is 40% of the national Medicare rate. Our results complement Wang et al (2024) study which examines within-hospital price variation for inpatient and outpatient services and finds that difference between minimum and maximum prices negotiated by hospitals to be 86% and 222% of Medicare rates, respectively [6].

We downloaded the machine-readable files for the largest national payers (Cigna, Aetna, United Healthcare, various Blue Cross and Blue Shield plans) for the third quarter of 2025.

These files contain information on negotiated rates for all covered services along with identifying information on provider tax identification number (TIN), payer name and plan type.

We removed duplicates and ‘ghost rates’—negotiated rates corresponding to a provider for a service they are unlikely to ever bill for,—which are common in TiC data following the established methodology from published literature [7].

This analysis focuses on the top twenty most commonly utilized HCPCS ((Healthcare Common Procedure Coding System) codes in an office setting (determined based on national Medicare data and Colorado and New Hampshire’s All-Payer Claims databases). To limit the influence of outliers, we exclude rates below the 1st percentile and above the 99th percentile for each billing code and restrict the sample to billing code–TIN combinations with at least two negotiated rates.

The analysis data consists of approximately 5.6 million negotiated rates for 20 HCPCS codes in an office place of service. Table 1 shows the distribution of negotiated rates by billing codes.

We first calculate the mean, median, minimum, and maximum rates at the provider TINs and billing code level. These represent the calculated measures of prices a provider is willing to accept across insurers for a billing code.

We then take the average of these estimates across provider TINs at the billing code level to calculate the mean, median, minimum, and maximum rate a provider is willing to accept for a billing code on average.

We define the range of rates as the difference between the average minimum and average maximum rates divided by the national Medicare rate for each billing code.

Table 2 shows the average mean, median, minimum, and maximum –negotiated rates for the top 20 billing codes. All four summary statistics are unweighted averages and represent the average of the respective summary statistics across all provider TINs at the billing code level. Out of the 20 billing codes, difference between the maximum and minimum rate is greater than or equal to 50% of the national Medicare rate for four billing codes - 80048 (Basic Metabolic Panel), 80061 (Lipid Panel), 85610 (Prothrombin Time), and 88305 (Surgical Pathology).

Across the 20 billing codes, we find the average variation between the minimum and maximum rates to be 40 % of the Medicare rate (unweighted).

Our assessment shows that providers negotiate a wide range of prices for the same service across insurers. Thus, the same provider can be low cost for one network but high cost for another network. Our findings that prices for commonly utilized services can vary by up to 40% of the Medicare rate in an Office setting compliments existing literature that documents substantial variation in prices across and within hospitals. This highlights the importance of the choice of network and benefit design in reducing health care spending. For employers, health plans, and policymakers, our findings demonstrate that choice of networks and providers can have a substantial impact on health care spending. Future work should aim to understand how navigating patients to lower cost providers can generate potential savings.