Rising cost of healthcare claims is a major challenge facing the healthcare industry. Many existing payers are facing challenges with legacy claims adjudication platforms that do not offer the desired level of flexibility and digitized capabilities. There is a need for a digitized, configurable and intelligent solution that can deliver a superior experience for all stakeholders while lowering the cost of operations.
Modern experiences for modern users
Customers today prefer ease of use while making any product purchase, and this also applies to healthcare. Smartphone apps from auto insurers like State Farm allow payees to submit a paperless claim by transmitting an image of the vehicle, eliminating the mundane process of filling lengthy claims. Customers expect personalized rewards for their auto insurance policies where telematics tracking is used to assess member risk profile and safe driving is encouraged with additional discounts. This trend is not just limited to the end customers, but also influences the expectations of the employees of insurance organizations who are constantly looking for more insights and automation of the claims process.
Aetna has created an AI-based claims platform that blends Natural Language Processing, an unstructured text parsing methodology and special database software to identify payment attributes and construct additional data that can be automatically read by systems. Like the Aetna example, more payers are looking at transforming claims processes to meet the customer expectation and at the same time, improve their efficiencies.
Automated claims processing with AI and ML
Healthcare payers have traditionally been operating in a fee-for-service model. These traditional claim management processes require manual intervention for adjudication and audits. Additionally, this is inefficient and unsuitable while moving towards outcome-based models. Healthcare payers need to push for greater digitization of the entire claims process. Even a partial automation of the workflow can result in significant gains in the form of reduced cycle times, lower operational costs, and improved experience for members as well as providers.
An automated claims processing system can transfer claims in real time from the provider along with necessary electronic health records. The automated algorithms can process the claims and perform real-time validation of the eligibility, benefits, and provider contract along with the medical diagnostic data. Any follow-up requests for additional information to providers can also be electronically parsed. The approvals or denials can be communicated electronically to the providers as well as members while digitally processing payments.
What’s more, AI-based claims solutions offer analytic capabilities that can assess the effectiveness of care management by helping track medication errors, adherence to medication therapies, and adverse drug interactions. These problems can result in expensive hospitalizations, regulatory penalties, and increased morbidity, respectively.
A well-designed claim solution can improve the experience for members and providers. Analytics can help members with timely detection of anomalies and suggest personalized care interventions. Better understanding of the path of the illness can help payers and providers devise appropriate interventions and can reduce costs while delivering superior care outcomes. Further, these AI capabilities assist with studies across multiple cohorts, when it comes to comparing the effectiveness of the recommended treatments for a large group of providers. Providers can benefit from faster reimbursements and greater transparency in the digitized process.
Fraud, waste, and abuse
Healthcare fraud, waste, and abuse are serious problems and considerable efforts have been made by CMS and HHS to control them. CMS estimates that improper payments worth over USD 105 billion have been made in the FY19 alone for government-sponsored plans such as Medicare, Medicaid, and CHIP. An AI- based claims processing system can assist in leakage and fraud prevention by identifying abnormal price patterns among providers such as upcoding and overcharging for services.
Intelligent AI algorithms can help identify unusual claims while automatically clearing normal claims. It can also predict the potential success rate if a claim is challenged and provide guidance to auditors for claims that may have to be denied. For example, the system help identify the right set of claims to be reviewed or denied, by comparing the cost of reviews against the value of the claim itself.
Risks and challenges
Building a successful AI solution requires a robust data model, process restructuring, and training models with high quality data. Any bias in training data can result in biased and incorrect predictions. A successful AI solution may require integration with other sources such as lab results and EMRs. Appropriate de-identification techniques need to be adopted to anonymize data and ensure privacy concerns are addressed.
The future of claims
Driven by increased consumerization of healthcare and regulatory pressures to control costs, there is an increasing shift towards value-based models. The focus of the health insurance industry as a whole is shifting from episodic care to the health and wellbeing of the covered population. Intelligent claims solutions can help the entire healthcare ecosystem by reducing cost of operations and improving the quality of care delivered.