Payment Accuracy: A Growing Concern.

Improper payments are a worsening problem in the US health care industry. It is estimated that overall waste tops $1 trillion per year, and improper payments contribute significantly to that figure. For Medicare and Medicaid in particular, improper payments increased from $64 billion in 2012 to $85 billion today.

Fraud is said to constitute 3% to 5% of total health care costs. These improper payments have ramifications beyond payers, leading to higher premiums for consumers and cutting into resources to improve the quality of care for members with legitimate expenses. Payers are working to address this problem, but they are confronted by departmental silos, outdated technology and inefficient approaches rife with redundancies, communication breakdowns and blind spots where unseen overpayments can thrive.

Fortunately, a new payment integrity paradigm is emerging. Under this paradigm, internal collaboration, process automation and better data support a holistic strategy and provide transparency across the organization. Payers can refine their selection of high-risk providers while preventing and correcting improper claims payments more quickly.

This process engages and educates providers, instead of creating headaches for them. Audits are fewer, more targeted and more effective, leaving both payers and providers more time and resources to spend on patient care.

New Payment Integrity Tools

While all plans take steps to address payment errors, the success rates depend on the maturity of those efforts and the tools used. Many companies are still using basic tools like Excel, and they have legacy systems that are “taped together internally” and cannot effectively communicate.

Information is often siloed, especially when health plans are acquired or combine, bringing together different systems that have different approaches to payment integrity. This means the payer’s business units are not sharing results with other units to make decisions, creating significant problems. For example, the special investigations unit may be building a case against a provider that is about to go to trial. If the data mining department collects on a minor claim during this process, it could hurt the SIU case.

Some payers have made strides with high-powered advanced analytics solutions and toolsets. But for many plans that use these tools, the failure rate is still high because they don’t operationalize the data and use it to inform decisions. “It’s like a farmer hitching a horse to a pickup truck instead of driving the truck.”

There are several reasons plans struggle to use the information gleaned from data mining. Data often isn’t cleaned up before analysis and is riddled with null values and blanks. Such data produces faulty insights. Organizations are also hamstrung by manual processes and one-off deliverable approaches that apply just one logic set, such as diagnosis-related group audits, to an entity.

Payers are using old methods to try to prevent new schemes, which results in a low ROI per entity audited. Instead, they need a risk-based approach that uses robotic process automation and universal logic/rules engines to make better audit selections. They must leverage historical data, trends and patterns to get new insights about providers - particularly as payer practices evolve.

This approach begins with data management. Proper data management starts with a data governance plan and includes infrastructure support such as security, the ability to pull information from data stores across disparate systems, data quality measures to ensure information is analytics-ready and enterprise data access to ensure everyone who needs the data can get to it. Creating a common data model with a shared data language allows health care organizations to glean insights from disparate data sets that were previously locked.

Robotic Process Automation & Artificial Intelligence

Good data management and analytics allow payers to better identify improper payments by focusing on the behavioral pattern of a medical provider, rather than a single questionable act.

To embrace this new approach, payers need to clean up their data, create integration links and add data quality measures. They can make these tasks easier with robotic process automation.

Using a computer to perform these functions reduces the potential for errors and more efficiently reconciles information - such as a patient’s name listed as both Johnny and John in records.

It also allows plans to tap additional sources, such as social media posts and peer-reviewed publications, and to leverage unstructured data from medical records.

Using artificial intelligence to conduct medical record reviews and capture data can increase auditor productivity by 25%

Automated Solutions Can Analyze Data Using:

Behavioral Analytics

statistical methods that find patterns, anomalies and outliers such as a provider seeing patients at a much higher rate than their peer group

Claims and/or Policy
Violation Business Rules

that search plan policies and rules, as well as any built-in exceptions

Clinical Targeting

a strategic look at very specific areas, such as a short hospital stay that was billed as an inpatient stay

With robotic process automation, payers can catch improper payments earlier, which is key to avoiding losses.


like “Dr. Bob,” who was seeing 11,000 Medicare beneficiaries and was certifying and recertifying patients for home health care. Data revealed that Dr. Bob was supposedly spending 37 hours a day certifying patients from home health care and that beneficiaries were being shared across 78 home health agencies.

It was later discovered that these patients were not receiving care at all, and a $375 million fraud scheme fell apart. Dr. Bob was successful for a time because he figured out how to exploit a lack of checks and balances within the Medicare system. A good data management can quickly and automatically flag atypical behavior so that it never reaches this a much higher rate than their peer group

Universal Logic/Rules Engine

Plans can also make better use of their data by feeding it into a logic/rules engine that includes coverage determination policies, claims logic, misuse logic, and other logic sets, applying all of them to each entity at once. These engines assess providers from a behavioral analytics perspective, identifying outliers by looking at how each provider measures up against their peer group. This helps spot anomalies such as a claim that is unusually high for a pediatrician.

Providers are scored for prioritization, with the most frequent, most costly offenders rising to the top. This ensures a higher ROI per entity audited and gives payers more confidence in pursuing highpriority leads. At the same time, false positives are reduced so provider abrasion goes down.

Profiles to Guide Selection

With these tools, payers get a clearer picture of providers and can evaluate all of a provider’s claims in near realtime to identify problematic behavior. This allows payers to assemble high-quality profiles to aid in provider selection. These profiles should include:

  • Claims analytics based on a library of queries that are continually updated
  • Provider analytics based on claims, member risk, demographics, quality and efficiency, and provider benchmarks
  • Member analytics based on conditions, utilization, cost, clinical and payment history, comorbidity, propensity, sociodemographics and lifestyle preferences
  • Payer analytics based on payment life cycle, bundled payments, quality and compliance, and payer benchmarks

Payers can then fine-tune selection sensitivity based on error likelihood and dollar amounts that meet their specific requirements, reducing unnecessary burden on providers.

Incremental Provider Behavior Modification

Provider selection is one part of a delicate process. Because of provider abrasion, many plans struggle with addressing providers that have payment integrity issues but are important to the plan, such as a large teaching hospital.

Payers need to be able to approach these providers under the assumption of an honest mistake, rather than jumping to the conclusion of fraud.

When attempting to influence provider behavior, payers should first use educational outreach and then move to incentives/penalties before pursuing a law enforcement referral. To ensure the most appropriate type of outreach is chosen, they must align internal stakeholders to de-conflict their efforts and route claims to the right department. After an outreach effort, it is important to monitor providers and measure outcomes to gauge success.

The New Payment Integrity Paradigm

It is time for a paradigm shift in payment integrity so that payers can leverage better data while supporting all of the stakeholders involved - including providers, internal departments and third-party vendors. New tools are allowing more targeted, more successful audits, ensuring plans make the best use of their resources to get members the care they actually need.

When payers automate and de-conflict their payment integrity efforts, they can move to a holistic strategy that eliminates payment errors and engages providers while continually improving the quality of care for members

Written by EXL Health Team

Contact US