Healthcare Price Disparities and Bargaining Power

Introduction:

Recent federal regulations, such as the Transparency in Coverage (TiC) rule, have introduced new requirements for healthcare providers and insurers to disclose negotiated prices. These regulations aim to enhance price transparency in the US healthcare system, a fragmented multi-payer environment characterized by significant variation in pricing across providers, specialties, and regions.

This study addresses two primary research questions:

1. Is there a wide variation in the prices healthcare providers are paid for treating patients?
2. What variables are associated with these differences in pricing among healthcare providers?
   
   

Initial findings confirm substantial price variability. Larger healthcare practices secure higher negotiated rates, leveraging market power to drive pricing. In contrast, higher proportions of poor health ratings in a population are associated with lower reimbursement rates. Providers offering higher quality services, as measured by CMS star ratings, are able to command premium prices, while insurers with dominant market share tend to pay lower prices, reflecting their bargaining power.

By analyzing key determinants such as practice size, population health, service quality, and payer market dynamics, this study provides critical insights into the factors driving price variability. These findings contribute to the broader goal of creating a more transparent and equitable healthcare pricing system.

Summary of Findings

Background

Evolution of the US Healthcare System

The US healthcare system has evolved into a fragmented multi-payer model through historical developments (Smith, 2023), including

• Early 20^th Century: The origins of health insurance began with employer-sponsored plans during World War II, incentivized by wage controls and tax benefits.
  
  
• 1965: Medicare and Medicaid were established, providing coverage for elderly and low-income individuals.
  
  
• 2010: The Affordable Care Act (ACA) introduced insurance marketplaces to expand coverage, yet 8% of the US population remains uninsured today.

US Healthcare Spending and Payment Models

The US spends over $4 trillion annually on healthcare, with funding sources split between private (50%) and public (50%) payers (AMA, 2023). Key contributors include:

• Private insurance through employers (28%)
• Out-of-pocket costs like copays and deductibles (10%)
• Public programs like Medicare (21%) and Medicaid (17%)
  
  

The dominant fee-for-service payment model compensates providers for each service performed. While there have been efforts to transition to value-based care – where payment is tied to patient outcomes and cost efficiency – progress has been limited. This fragmented structure of payers and payment models results in substantial variation in healthcare prices, driven by factors such as payer negotiations, geographic adjustments, and differences in provider characteristics.

Transparency in Coverage and Price Variability

The Transparency in Coverage (TiC) rule requires hospitals and insurers to disclose negotiated rates through publicly accessible, machine-readable files. While these regulations aim to empower consumers with pricing information, challenges persist. Lack of standardization and incomplete data make it difficult to compare prices across providers and insurers, and price transparency is often inaccessible to patients without digital proficiency, limiting practical impact.

The available data has revealed substantial variation in prices for identical services across different regions and providers, highlighting the fragmented nature of the system. For example, a knee replacement costs $58,193 in New York but $23,170 in Baltimore (Kirani et al., 2021). Discrepancies are not limited to complex procedures; routine lipid panel tests, used to measure cholesterol levels, range from $11 to $54 within the same city.

These disparities underscore the challenges consumers face when attempting to navigate healthcare pricing. Without standardized pricing, patients lack the tools to make informed decisions about the care. The lack of transparency and consistency not only impacts individual patients but also has broader implications for healthcare costs and policy.

Factors Influencing Price Variability

This study considers four key factors that influence the prices healthcare providers receive for their services:

1. Practice Size: Larger practices negotiate higher reimbursement rates due to greater market leverage.
2. Population Health: Providers in regions with poorer health indicators often receive lower payments.
3. Provider Service Quality: Higher CMS star ratings are associated with premium prices, incentivizing superior care.
4. Payer Market Share: Dominant insurers leverage their power to negotiate lower prices, with variations based on local market conditions.

These dynamics reflect the interplay of market power, patient demographics, and payer strategies, emphasizing the need for further research to align pricing models with quality and health outcomes.

Data and Analytical Framework

This study examines healthcare pricing by analyzing negotiated rates between providers and insurers using data from Transparency in Coverage (TiC) disclosures and Medicare’s physician fee schedule (PFS). TiC data, mandated by federal regulations, include rates from UnitedHealthcare, Aetna, and Cigna. Medicare prices, published by CMS, serve as benchmarks to validate TiC data and assess price fairness. National Provider Identifier (NPI) data links pricing with provider characteristics, focusing on organization-level negotiations. NPI data categorizes providers by specialization, enabling comparison across practice domains.

Correlates Data

Key correlates include practice size, population health ratings, quality of service ratings, and payer market share. Practice size data derives from CarePrecise’s CP Advanced database, while population health ratings are sourced from County Health Rankings. CMS Quality Star Ratings measure service quality, and payer market share reflects health plan enrollment proportions at the county level.

Focused Dataset

The initial dataset contains over 147 trillion records, narrowed to a Focused Dataset of 1.5 billion observations by excluding irrelevant codes, duplicate records, and zombie rates – contractual rates for services providers do not perform. The dataset emphasizes baseline procedures, including those performed in physician offices, hospitals, and surgical centers. Providers’ average quality ratings vary slightly by type, with specialists and surgeons scoring the highest (3.5/5). Surgeons command the highest average price ($1,384), while primary care providers average $341. These variations highlight the complexity of healthcare pricing.

Motivational Findings

Price Transparency and Variability

Analysis of coefficients of variation (CV) quantifies price dispersion among providers. Ambulatory service centers (ASCs) and hospitals show moderate to high variability (mean CV ~40%), while primary care providers exhibit slightly lower variability (~35%) but wider distribution. Variability is not uniform across specialties; cardiology, endocrinology, and radiology have higher CVs (~50% – 60%), revealing substantial price differences. Surgical procedures also show significant variability, particularly in “Nervous” and “Musculoskeletal” categories. Outliers indicate extreme pricing scenarios, underscoring the heterogenous nature of healthcare pricing.

Correlations Analysis

Preliminary correlation analyses suggest that larger practices receive higher negotiated rates (r = 0.053, p < 0.01), while providers in areas with poorer population health earn lower rates (r = -0.022, p < 0.01). Higher quality ratings correlate with higher prices (r = 0.030, p < 0.01).

Empirical Findings

Regression analyses confirm that practice size, population health, quality ratings, and insurer market share significantly influence pricing disparities. Larger practices leverage market power for higher rates, while insurers with greater market share negotiate lower prices. Population health ratings and quality of care also share reimbursement rates, highlighting the interplay between market dynamics and healthcare pricing.

Relationship Between Practice Size and Prices

This model evaluates how practice size influences prices received by providers. The regression equation is:

• Price_iptc: Negotiated rate for a procedure performed by provider p for payer i in county
• Group_Count_p: The number of doctors in the provider group.
• ∑ _{n-1 j-1} δ_jState_Locality_Num_j: Dummy variables for geographic regions, reflecting differences in pricing across CMS MAC localities.
• ∑ _{m-1 k-1} γ_kHCPCS_Code_Num_k: Dummy variables for procedure types, controlling for differences in services

The results demonstrate a positive relationship between practice size and prices, with larger practices generally receiving higher payments from payers. Most specialties show statistically significant increases in prices as practice size grows, though a few specialties do not exhibit significant effects. This suggests that practice size impacts negotiating power differently across specialties.

Relationship Between Population Health and Prices

This model examines how county-level health metrics affect provider prices:

• Perc_Poor_Health_c: Percentage of the county population reporting fair or poor health
  
  

The findings show a negative relationship between poor health ratings and prices, with providers in counties with higher poor health ratings receiving lower negotiated rates. This effect is particularly pronounced for primary care providers and hospitals. However, a positive relationship is observed for eye surgeons, and some categories are not statistically significant. These results suggest that population health may influence pricing strategies, particularly in regions where poorer health outcomes are prevalent.

Relationship Between Quality of Service and Prices

An ANOVA test and regression analysis assess the relationship between CMS Quality Star Ratings and prices. The regression model is:

• Quality_Rating_p: CMS star ratings (1-5), where higher values indicate better quality
  
  

The ANOVA results indicate significant differences in mean prices across the five quality-rating categories. Providers with higher ratings receive higher negotiated rates, with a clear upward trend as ratings increase from 1 to 5. Regression results confirm this positive relationship, demonstrating that better-rated providers command higher payments from payers.

Relationship Between Insurer Market Share and Prices

This model investigates the relationship between payer market share and prices:

• Payer_Market_Share_c: Proportion of total county health insurance enrollment held by a payer.
  
  

The results reveal negative relationships for Aetna and Cigna, where higher market shares correspond to lower prices. In contrast, UnitedHealthcare exhibits mixed results, with prices decreasing in approximately half of the specialties analyzed. These findings highlight how dominant payers often leverage their market position to negotiate lower rates, though the effect varies by payer and specialty.

Multivariate Regression Analysis

This model integrates key predictors, including practice size, population health, and payer market share, while controlling for geographic and procedure-specific effects:

• Payer_i: Categorical variables for Aetna, Cigna, and UnitedHealthcare, reflecting payer-specific effects.
• Mkt_Share_c: County-level market share of payers.
  
  

Key Findings:

1. Practice Size: Larger practices are associated with higher prices, reflecting economies of scale and stronger negotiating power.
   
   
2. Population Health: Higher poor health ratings correlate with lower prices, potentially reflecting payer cost strategies in high-need regions.
   
   
3. Provider Service Quality: Providers with higher ratings receive higher negotiated rates.
   
   
4. Payer Market Share: Dominant payers (Aetna and UnitedHealthcare) generally negotiate lower prices as their market share increases, while smaller payers may experience price increases with growing market presence.
   
   

These results emphasize the multifaceted drivers of healthcare pricing, including practice characteristics, population health, and payer dynamics.

Robustness Test

To ensure the validity of the study’s findings, a robustness test was conducted by re-estimating the multivariate regression model using a logarithmic transformation of the dependent variable, price. This transformation addresses potential issues such as skewness, heteroskedasticity, and outliers in the price data. By interpreting coefficients as percentage changes, the log-transformed model offers an alternative perspective on the relationships examined.

Methodology and Results

The robustness test replicated the original multivariate regression model, substituting log price as the dependent variable while maintaining all other specifications. The results demonstrate a high degree of consistency between the raw price model and the log-transformed model, strengthening confidence in the robustness of the findings.

Key relationships remained directionally consistent across both models:

• Practice Size: Larger practice groups consistently correspond to higher prices, likely reflecting economies of scale or stronger negotiating power.
  
  
• Population Health: Poorer health ratings are associated with lower prices, reaffirming that regional health metrics influence pricing strategies.
  
  
• Payer Market Share: Aetna and Cigna’s market shares continue to exhibit negative correlations with price, while UnitedHealthcare’s results show variation. Specifically, in counties with above-average poor health ratings, UnitedHealthcare’s market share correlates with lower prices when it is the market leader, a relationship not observed in the overall dataset.
  
  

These results confirm the reliability of the study’s conclusions across different model specifications and highlight the role of regional and market-specific dynamics in healthcare pricing.