Risk management and increased efficiency are two best practices that can help healthcare payers optimize revenue. To accomplish this, many are looking to artificial intelligence (AI) and machine learning (ML) solutions — particularly solutions they can scale up in the future as the applications grow in sophistication.
The COVID-19 pandemic has led to an increase in payer adoption of technology and innovation. In fact, data from the 2021 State of Healthcare report from HIMSS suggests that 53 percent of health insurance executives have accelerated the adoption of AI and ML practices within their operations.
Meanwhile, private equity firms who are actively investing in AI innovations designed for the healthcare provider space — such as clinical decision support — also should consider the potential these tools have for boosting revenue projections on the payer side. The $692 billion payer market has a significant cache of capital to spend on AI where it promises a good return.
For example, an analysis in 2018 by Accenture ballparked the savings insurers can realize from AI in core administrative functions at $7 billion within 18 months. More than $2 billion of that would come from AI handling in customer interactions and another $1.1 billion from automated claims processing and review functions. But there’s more to be had from AI for payers.
Reduce readmissions and transform risk adjustment
Payers are leveraging AI to help reduce hospital readmissions among their covered populations — estimated to cost the U.S. healthcare system in excess of $26 billion. Last year, Blue Cross Blue Shield of North Carolina, which serves more than 3.8 million members in the state, developed predictive analytical models to identify target populations at risk for hospital readmissions.
The approach applies a readmission risk score to members currently undergoing inpatient procedures, with members further prioritized by probability of readmission, low engagement with their primary care physicians, and prescriptions for eight or more medications. Thanks to this AI solution, the Blue Cross Blue Shield of North Carolina care management team’s engagement success rate rose within a year from 12 percent to 57 percent, with the expectation of more hospital readmissions averted as a result.
Natural language processing (NLP), along with ML is also helping transform the hidden information in member data to create a more complete picture of each member’s health, including risk. Uncovering missing ICD-10 codes or diagnoses allows for better coordination of care and more accurate risk scores, which leads to more favorable payment for Medicare Advantage plans. Utilizing both of these AI technologies, plans can refine risk adjustment with the clinical documentation to support it. AI improves effectiveness in retrospective chart reviews as well as prospective diagnosis code capture.
Boost member engagement
Payers are also starting to use AI technologies to improve and personalize the member experience, all while creating greater value and driving operational efficiencies.
Blue Health Intelligence, for example, which provides clinical data expertise for Blue Cross-affiliated health insurance plans across the United States, examined health insurance claims from 48 million patients, augmented them with U.S. census data, and identified 6,000 variables across clinical and demographic categories. It then applied machine learning to identify high-cost claimants.
The resulting information has allowed the Blues to reach out to patients and physicians, collaborating to address those who might have gaps in care, limited mobility, or financial barriers preventing prescription fills. These efforts can also help payers identify potential enrollees for health management program participation.
Among other major health insurers, Cigna and Humana have been using chatbots to communicate with their members through real-time, automated message responses. Chatbots can reduce the call volume and increase efficiency for call centers.
For example, Cigna’s Answers chatbot uses NLP to understand and respond to more than 150 common questions with personalized benefits information. When combined with the firm’s digital member service platform, Cigna found that these technologies helped increase customer satisfaction by 20 percent.
Detect and prevent fraud
In fiscal year 2020, the Department of Justice obtained more than $1.8 billion in settlements and judgments from civil cases involving fraud and false claims in the healthcare industry. As insurers move from the old “pay and chase” model to a proactive model that detects and prevents fraud before the claim is paid, they can leverage the real-time response of AI solutions.
The National Health Care Anti-Fraud Association offers a conservative estimate, calculating fraud at 3 percent of total healthcare spend, while some government and law enforcement agencies figure the loss as high as 10 percent, or more than $300 billion.
AI technologies can be an invaluable tool for payers when it comes to analyzing claims, looking for patterns and red flags that could indicate the presence of healthcare fraud — or even repeated, honest mistakes that happen to benefit the provider financially.
Deep analyses could include any of the following:
- Suspected upcoding
- Suspected unbundling of bundled services
- Questionable overutilization
- Possible medical identity theft of providers or patients
- Obvious variation in care services delivered
While outright intentional fraud accounts for a significant slice of healthcare spending, the honest mistakes are much harder to manage. An occasional transposition of numbers, for example, can lead to a claim rejection, and the whole claims filing process begins again at a cost to payer and provider. About 10 percent of claims filed are rejected. But the sophistication of AI has advanced to the point where small mistakes can be corrected in the background more often, saving time and money.
AI’s potential is still emerging for payers. However, payers have an incentive to adopt the technology to gain efficiencies that pay off.
Investors would be wise to invest in established solution providers that can help payers get a handle on AI capabilities. Canton & Company can help identify those targets as well as provide the market intelligence you need to craft a value-creation strategy. Contact us today!