Designing and maintaining a high-performing provider network is difficult; not because there is a shortage of great providers practicing medicine in caring, cost-effective, and patient-centric ways. Rather, it’s a lack of clear visibility into provider performance and an ever-changing definition of “high value.”

Vast amounts of provider performance data exist, but it’s often disparate, siloed, and not available when we need it. Furthermore, the analyses are often not adjusted for factors unique to that provider, making it nearly impossible to compare one provider to another. It’s like trying to compare the Golden State Warriors to the Atlanta Braves. Both are great but they play completely different sports.

When creating or optimizing high-performing provider networks, it is critical to understand provider performance on cost, quality, and utilization from day one. Without this data, most payers are building networks blind – basing network design on how other competitors in the market have designed their networks. It can then take up to two years before the organization has enough data on providers to begin to assess performance.

However, data science combined with the use of machine learning has allowed previously inaccessible data to be unlocked, surfaced, and analyzed in near-real-time.

These capabilities have the ability to transform network design in the same way predictive analytics and logistics have transformed online retail.

Applying the same transformative thinking found in these industries to designing a network, we start with 3 critical considerations:

1. Provider cost
2. Provider quality and utilization
3. The extent to which those providers integrate with other providers in the network (referral performance)

What if you could make building a provider network as easy as shopping online?

Before any big purchase, you want to understand and consider a multitude of factors. Can you imagine buying an expensive item, such as a computer, without knowing anything about it? This is akin to adding providers to your network without reliable data. The same decision criteria used to make an expensive purchase online can also be applied to choosing the best providers to add to your provider network. Here are the 3 questions you should be answering with quality data and analytics:

Key Decision Criteria #1: How much does it cost?

The first thing to assess a provider on when adding them to your network is their provider performance on cost metrics. Would you buy a computer online if you didn’t know how much it was going to cost you? And how much provider performance improvement would you be willing to pay more for? Is that worth the price?

Now take that same lens to providers. Before adding a provider or group to your provider network it is important to assess critical cost metrics: Total PMPY cost, inpatient cost, outpatient cost, professional cost, and cost on key services and procedures. It is also important to look at the underlying metrics driving these costs, such as post-acute care usage and expensive drug prescribing patterns. Further, it is important to understand a provider’s reimbursement rates. This can help you have confident, evidenced-based rate negotiations that tie reimbursement to value and provider performance and ensure you don’t over or under pay.

Key Decision Criteria #2: Is it high-quality?

More expensive doesn’t always equate to higher quality – but how can we tell? As consumers, we turn to online reviews to validate quality and satisfaction. Online reviews and specific details (e.g., processing power) would be a crucial input into your decision to purchase one computer over another.

A more expensive provider may still be valuable to have in your provider network if they provide high quality care and are effectively utilizing services. Or even better, a provider is both low cost and high quality.

The Centers for Medicare and Medicaid Services (CMS), consumers, and employers are ever-increasingly beginning to rightfully care about the quality of care they receive from their providers. Leading payers and health insurers will continue to be creative in network design, focusing on whole-person care and thinking beyond just the care delivered in the provider’s office and rather about the patient’s or member’s overall health management.

Therefore, accurate and risk-adjusted metrics like readmission rate, mortality rate, claims-based HEDIS measures, HCAHPS scores, and clinical appropriateness of care measures are crucial. In considering utilization, it is important to understand visits per 1,000 across acute IP admissions, observations, ED visits, and referrals by specialty type.

Key Decision Criteria #3: When and how will it arrive?

Lastly, it is critical to understand the referral patterns within a network you are trying to build. When online shopping, you need to understand shipping and delivery time and rates. If a retailer won’t delivery to your home or charges you an extra fee to do so, you might consider shopping somewhere else. Similarly, with respect to a network of providers, it is important that the providers who are added to that network send a significant portion of their referrals between one another. Selecting a provider therefore requires not only understanding their provider performance but the performance of those they refer to. Referral patterns are hard to change, so it is important to know upfront.

Clarify makes building a high-quality provider network as simple shopping online

Using the Clarify platform, our customers assess providers on risk-adjusted Cost, Quality, and Utilization metrics using outside-in claims data, and then dynamically build and model new network scenarios. Our claims data allows us to see a provider’s performance across all the patients they’ve seen and to set strict inclusion and exclusion criteria to allow for fair evaluation. Quality is hard to measure, and while measures like STAR ratings are good first step, at Clarify we believe that detailed, claim-level insights, surfaced in the right manner, can help our customers identify truly high-quality providers.

Our customers can also review individual “Referral Scores” for each provider, as well as the provider network overall, based on the performance of the providers being referred to. Clarify’s solution allows our customers to go a level deeper and see to whom individual providers most often refer to and then add those specialists to their model networks on the fly. This ensures that referrals stay within your network and are made to high-quality providers.

At Clarify, we rely on a dataset of 220+ million claim lives across Medicare, Medicaid, and commercial populations. We combine this rich dataset with best-in-class data science and machine learning techniques to create “Expected Values” for each patient-provider interaction across a vast library of metrics. For every provider, provider group, hospital, health system, and downstream provider, we calculate observed performance on any given metric and compare it to how they were expected to perform based on factors unique to that provider. This allows for an even comparison across providers, incorporating over 300 different patient and clinical factors, from HCC risk scores to social determinants of health (SDoH). By giving customers transparency into cost, quality and utilization, and referral patterns from day one, Clarify makes building a new provider network as easy as shopping online, all while ensuring that your members receive the high-value clinical care they need.

To learn more about this topic, download our 2023 Networks Playbook for Health Plans. In this playbook, you’ll learn three plays to design and optimize provider networks in new & existing markets.