The Dollars in the Details

Selecting insurance carriers is akin to navigating complex terrain. Employers must consider numerous factors, ranging from network disruption to carrier service to financial impacts. As healthcare costs continue to rise, financial considerations remain a crucial landmark in this intricate landscape.

Employers’ fundamental challenge is lack of visibility into what they are paying providers in their network and the partners they rely on to procure healthcare. However, more employers today are contracting third-party audits and reevaluating their vendor relationships to address longstanding concerns about patient outcomes and experience and provider cost and quality. These efforts are being driven by inflationary pressures, employer lawsuits on health costs, federal rules regarding price transparency data, and fiduciary stewardship of plan assets.

This means the climate is shifting yet again for brokers and consultants. Winners and losers in the self-insured market will be distinguished by their approach to new data and more granular insights for cost reduction strategies. While there are many ways to control healthcare costs for employers, maintaining the status quo will only ensure they keep rising.

For decades, brokers and consultants have relied on discount analysis as their primary compass to chart the financial impact of different networks. This method uses standardized discounts to compare carriers and develop financial impact models for employers in network selection.

As the healthcare landscape evolves, though, we’ve come to understand that discount analysis may not be mapping the entire terrain. Enter more transparent healthcare pricing data—which must be made publicly available under the 2020 federal Transparency in Coverage (TiC) rule—a new instrument in our navigational tool kit. This emerging data source promises to reveal actual negotiated rates between providers and insurers, potentially offering a more detailed and accurate view of the true healthcare cost landscape.

TiC data isn’t without its challenges. In its current state in machine-readable files, it is filled with inconsistencies and gaps and demands immense computing resources, detailed subject matter expertise, and the ability to untangle the nuances between carriers.

Can we use TiC data, despite its flaws, to enhance our understanding of the healthcare cost landscape? And, ultimately, could it revolutionize how we approach network selection and design once these initial hurdles are overcome?

Familiar Territory: Discount Analysis

The tried-and-true approach of discount analysis involves the major carriers contributing to a universal discount database, which is validated and actuarially certified. Brokers and consultants use this data to estimate relative price differences between carrier networks, which is often pivotal in RFPs and contract negotiations.

The appeal of discount analysis stems from three key attributes. It offers conformity by ensuring standardized practices across carriers, creating a common language for network assessment. It is comprehensive, covering all services within a carrier’s book of business, providing a complete view across all medical procedures and negotiated arrangements. And it is convenient: over the years, brokers and consultants have developed well-established processes and tools for working with this data, making evaluation efficient and familiar.

That standardized nature allows for consistent comparisons across different networks, while the comprehensive coverage of discount analysis provides a broad view of a carrier’s pricing structure. The efficiency and familiarity have enabled quick and reliable network evaluations, streamlining the decision-making process for employers and brokers alike.

Navigating the Gaps in Discount Analysis

The limitations of discount analysis are well-known and multifaceted. The most glaring issue is the price versus discount dilemma. Discounts are calculated from hypothetical billed charges, which can vary dramatically between providers and bear little relation to the actual cost that a plan or member pays. This disconnect can lead to misleading comparisons where a higher discount percentage doesn’t necessarily translate to a lower final price. For instance, a 30% discount for a $1,000 charge ($700) is still more expensive than a 20% discount for an $800 charge ($640). Discounts are a function of altitude: if prices are higher, then the discount doesn’t necessarily matter as much.

Beyond this obvious flaw, discount analysis suffers from several hidden limitations that can undermine the accuracy of network comparisons. These stem from the way data is aggregated as part of the process. Discount analysis often assumes a standard mix of providers and patient utilization patterns, which may reflect something other than an employer’s specific workforce needs or geographic realities. Additionally, applying a single discount figure for a region can mask substantial variations between providers or different types of facilities. Moreover, high-level discount averages can obscure variations in pricing for specific procedures, potentially hiding poor pricing for high-cost or frequently used services behind excellent overall discounts.

Fundamentally, discount data is incomplete and fails to capture crucial aspects of healthcare value. It provides no information about the quality of care, potentially leading employers to prioritize cost savings over health outcomes. Moreover, discount analysis struggles to reflect the total cost of care over time. It doesn’t account for a carrier’s effectiveness in care management, preventive services, or long-term cost reduction programs. A carrier with lower upfront discounts but superior care management could deliver better value.

These factors significantly impact an employer’s healthcare costs and employee health outcomes but are entirely missed when focusing solely on discounts.

Charting New Waters with Transparency in Coverage Data

Unlike discount analysis, transparency data provides actual negotiated rates for specific services across different providers and carriers. This offers a clearer map of real healthcare expenses, enabling more informed decision-making. The key advantages include the following.

1. Actual Costs, Not Hypotheticals: Transparency data shows real negotiated rates, cutting through the fog of varying list prices and discounts to show the true cost of services.

2. Provider-Level Comparisons: Transparency data allows for detailed comparisons between different providers and provider types (not just hospitals), aiding in both selecting the best network and, once selected, steering to the most cost-effective providers within that network.

3. Insights into Variability: Transparency data exposes the spread of pricing within a network, highlighting carriers with more consistent pricing versus those with wide variations.

4. Potential for Quality Assessment: While not directly solving quality measurement, the provider-level detail within the transparency data opens new possibilities for linking price and quality in ways that are impossible with discount analysis.

It’s clear that transparency data has the potential to redraw our maps for healthcare cost analysis, carrier selection, and new forms of network design. The mission now lies in effectively leveraging its insights to plot better courses for employers and their teams.

Navigating Transparency in Coverage Data Challenges

Transparency in Coverage presents significant navigational hazards beyond the obvious vastness of the data itself.

1. Data Quality: Files often contain illogical rate combinations and filler data. Extensive cleaning is crucial. This goes beyond mere data scrubbing—it requires deep healthcare knowledge to effectively identify and remove outliers and illogical rates. For instance, by distinguishing between a genuinely high-cost procedure at a given provider and a data error.

2. Reporting Inconsistencies: Carriers report data differently, and medical coding modifiers and billing arrangements vary widely. Modifiers indicate additional information about a medical procedure or service without changing the meaning of the code. Normalizing this data calls for a nuanced understanding of carrier reporting and healthcare billing practices.

3. Complex Payment Arrangements: Some intricate pricing structures must be better captured. The data lacks the detail needed to untangle certain arrangements fully, particularly for value-based, capitated, and bundled payment models.

The result is a mix of noise and valuable insights. Some service comparisons may be invalid, and data for certain networks may be limited.

This demands approaching the data critically and understanding its limitations while leveraging its strengths. As reporting practices evolve, we can expect this data’s value to grow, gradually providing a more comprehensive view of the healthcare cost landscape. The key lies in combining this new tool with traditional methods and contextual information to enhance our understanding of those costs and chart a more informed course in carrier selection and network management.

Case Study: Network Comparison Analysis

This analysis considers a hypothetical employer in the Atlanta area and compares the networks offered by two actual health insurers—Carrier A and Carrier B. For Carrier A, the evaluation includes both a PPO and a local HMO. For Carrier B, the network included is a point of service (POS) plan that has a broad PPO-like network but covers out-of-network care at a higher cost to plan members.

A representative “service basket” is essential to facilitate network comparisons for the employer. Given the limitations in machine-readable files and the law of diminishing returns when attempting to capture all services, standard practice is to target coverage of approximately 70-80% of claims when establishing a service basket for comparison purposes. We constructed a sample service basket for this simplified example. A combination of typical employer experience, data completeness, and credibility within the machine-readable files guided the selection criteria. The sample service basket encompasses the following:

• Inpatient procedures (50 procedures): Procedures related to maternity and delivery, cardiac conditions, mental health/substance abuse, musculoskeletal conditions, and other conditions
• Outpatient procedures (200 procedures): Surgical and interventional procedures across multiple medical specialties, including orthopedics, general surgery, cardiology, urology, gynecology, neurosurgery, and ophthalmology
• Professional services (200 procedures): Surgical procedures, diagnostic tests, preventive care, therapeutic interventions, and medication administration across various medical specialties.

Each service within the basket is weighted to represent the distribution of services that an actual employer might experience. Subsequently, inpatient, outpatient, and professional services are weighted to determine an overall cost estimate across different networks.

The analysis must also consider which providers to include. There are two primary selection approaches.

1. The first method comprises all providers within each network, weighted by market-level use or modified use assumptions based on employer-specific considerations. This approach provides a broad overview of the entire network.

2. The second method focuses on providers common across all networks and then incorporates providers unique to individual comparison networks. While more complex, this allows for a detailed examination of network changes and potential shifts in employer use. It enables analysis of three key factors: rate comparisons for common providers that employers already use across networks, relative rates and impact of potential new providers, and the impact on rates for providers that would no longer be in-network under a given change.

The granularity afforded by the Transparency in Coverage data makes this approach feasible. It allows for a more comprehensive understanding of how network changes might affect an employer’s healthcare costs and employee access to care. The analysis can provide insights into the potential financial and use impacts of selecting different networks by examining common providers first, then layering in network-specific providers.

This analysis employs the second approach, but limits its scope to comparing rates for the set of providers common across all networks. This serves as a foundation for understanding the core differences between networks.

Rates are based on the negotiated rates reported in each carrier’s machine-readable files while removing outliers and filler rates across carriers. We also adjusted for differences in modifier treatment. We focused on analyzing valid procedure types and place-of-service combinations to remove much of the noise that is otherwise present in the machine-readable files.

An overall comparison of rates across the three networks, based on this selected service basket and provider mix, relative to average rates, is illustrated in Exhibit 1.

As indicated:

• Carrier A PPO costs are the highest, 4.4% above average
• Carrier B POS is 2.7% above average (a 1.6% savings relative to the Carrier A PPO)
• As expected, the local HMO costs from Carrier A are the lowest, 7.1% below average.

Relative averages are shown based on the selected service basket and for providers common across all networks. The chart assumes consistent utilization across all carrier networks, with no adjustments for any potential carrier case management programs.

Based on this initial overview, the analysis reveals that when selecting between broad networks, Carrier B would offer modest savings for the employer compared to Carrier A. However, these high-level comparisons only scratch the surface of network evaluation. Beyond this general average, the variation in prices seems to form certain observable patterns, from the setting of care (i.e., inpatient versus outpatient), to lines of services (e.g., orthopedics), to acute care hospitals overall. These differences provide insights that could lead an employer to making a different decision on networks or providers than if sticking with the overall average difference.

For example, this network pricing comparison shows an inverse relationship between inpatient and outpatient contracts. This is a contracting pattern we’ve often seen when analyzing machine-readable file data across different geographies. In this instance, Carrier B has more favorable pricing for an employer on inpatient and less favorable on outpatient, and vice versa with Carrier A. This is why it is essential to delve deeper and compare prices at the service line level and facility level. Price optimization requires much more nuance than just the network’s aggregate cost differential.

Extended Analysis: Cost Variations Across Care Settings and Service Categories

Further analysis to explore the more granular aspects of each network begins by examining inpatient services separately from the other categories. The average price per claim for the inpatient basket reveals significant differences between networks. The Carrier A PPO has the highest average price at $20,319, followed by Carrier B POS at $15,938 and Carrier A HMO at $14,740 (see Exhibit 2). While Carrier A PPO costs are generally only slightly higher than the Carrier B POS, this breakdown shows that inpatient costs are over 20% higher for the Carrier A PPO compared to the Carrier B POS. This stark difference in inpatient costs highlights the importance of drilling into specific care settings and service categories when evaluating networks.

The analysis confirms that the Carrier A PPO rates are consistently about 20% higher than the Carrier B POS rates across a wide range of inpatient services (see Exhibit 3). This is particularly relevant for employers with high use of inpatient services, potentially due to a less healthy overall workforce or a demographic distribution with high expected maternity spend. The extra cost of selecting the Carrier A PPO over the Carrier B POS for these employers will likely be much higher than the initial analysis indicated.

Conversely, the average cost per outpatient service for the Carrier B POS is 18% higher than the cost per service for the Carrier A PPO. This means that Carrier A PPO is 20% higher for inpatient services and 18% lower for outpatient services than the Carrier B POS.

Exhibit 4 illustrates that while patterns are similar, differences in outpatient costs show more variability between carriers than inpatient costs.

Exhibit 5 shows that when the prices are compared by clinical service line, the differences can be reduced or inverted in certain instances.

Given the increased cost variability among plans for these services, a complete analysis would more comprehensively examine differences across various clinical categories for outpatient rates. Additionally, our sample analysis reflects a constant set of providers across all three networks. In a complete analysis, we would look at the incremental impacts of providers added and removed due to changing networks to understand how shifts in employee steerage may impact overall plan cost.

Extended Analysis: Provider-Specific Comparisons

Provider-level analysis can offer valuable direction in network selection, but its true power often lies in informing employee steerage strategies and plan design after a network is chosen.

For example, let’s examine outpatient institutional rates for total knee arthroplasty across various hospitals and ambulatory surgical centers (ASCs) in the Carrier A PPO and Carrier B POS networks.

Despite Carrier B showing higher average outpatient facility costs—overall and for the orthopedics subset of our service basket (see Exhibit 4)—it offers significantly lower rates for knee replacements at several key facilities (see Exhibit 6). This pattern could extend to other high-cost orthopedic treatments, which could position Carrier B as a favorable option, especially considering its lower inpatient facility rates. However, the optimal choice will depend on the specific demographics and needs of an employer’s population.

When it comes to steering employees toward optimal providers, in this case within the Carrier B network, negotiated rates strongly favor sample hospitals over ASCs for knee replacements. The total value of choosing Carrier B, for both employer and employee, will hinge on successfully guiding employees to these cost-effective hospital options. On the other hand, within the Carrier A network, employees and employers may be better served by steering employees to certain ASCs.

While this illustration focuses on costs, a comprehensive analysis must also consider the quality of care in making informed network steerage decisions. The basic illustration provided here serves as a starting point, demonstrating the potential insights gained from granular, provider-level data analysis.

A New Dimension in Network Selection

Transparency in Coverage data presents a significant advancement in network evaluation and carrier selection methodologies as the healthcare cost management landscape evolves.

This information offers unprecedented granularity in understanding actual healthcare costs. However, like any emerging data source, its application requires careful consideration and expertise.

The following strategies can help employers and brokers effectively incorporate TiC data into their network selection process.

1. Complementary Analysis: Use TiC data to complement, rather than replace, traditional discount analysis. This integrated approach leverages the established reliability of discount analysis while benefiting from the detailed insights provided by TiC data.

2. Strategic Service Focus Concentrate analysis on services that constitute a significant portion of your organization’s year-over-year healthcare expenditure. This targeted approach yields meaningful insights while mitigating the risk of data overload inherent in comprehensive TiC data sets.

3. Discount Validation: Employ TiC data to corroborate discount analysis results. Discrepancies between these methodologies can highlight areas requiring further investigation, potentially revealing hidden cost implications or savings opportunities.

4. Price Variability Assessment: Leverage TiC data to evaluate pricing consistency within networks. Networks exhibiting high price variability may present less predictable costs, even if their average prices appear competitive.

5. Informed Steerage Strategies: Use the provider-level insights from TiC data to develop effective post-selection employee steerage strategies. As our example illustrated, significant savings can be achieved by strategically guiding employees to cost-effective providers within a chosen network.

6. Holistic Evaluation: While TiC data offers valuable cost insights, it’s imperative to consider additional factors such as care quality and network accessibility in decision-making. For example, provider-level TiC data can be linked to facility quality information to aid in this more holistic evaluation.

As we navigate this new era of healthcare cost transparency, it’s crucial to approach TiC data with enthusiasm and prudence. The current limitations of this data source necessitate the use of advanced business rules for data processing, external benchmark data to ensure rates are in the appropriate range, and careful analysis and interpretation. However, as data use increases, reporting practices evolve, and data quality improves, the value of TiC data is expected to increase, providing an increasingly comprehensive and accurate representation of the healthcare cost landscape.

By integrating the established methodologies of discount analysis with the granular insights offered by TiC data, this new analysis provides brokers and consultants additional tools to evaluate the discounts as well as the effectiveness of the network, and how to generate savings by pointing employees to higher-value facilities rather than just swapping carriers based on relative discounts. Stay tuned for our next case study, which will dive into care steerage.