AI for Healthcare Costs Reduction

AI for Healthcare Costs Reduction– Claims Processing, Fraud Detection, and Care Quality

AI is now on the top of mind for healthcare providers, payers, practitioners, governments, and innovators. The focus is on applications impacting existing care delivery and how the evolving healthcare needs can disrupt them. Additionally, AI for healthcare can impact three significant areas – day-to-day operations across payers and providers, population health management, and care innovation.

How is AI adoption in healthcare? As per McKinsey, AI adoption in healthcare lags the other digitized sectors like retail. But AI is poised to transform healthcare operations, disease diagnosis, care quality, patient experience, and decisions in radiology.

The main objective of AI for healthcare is to reduce the administrative overhead, which accounts for 25% of the total US healthcare spending. Where does this spending go? The Prior Authorization for any treatment requires a lot of manual information exchange and reviews by the payer and provider’s clinical staff. Though it looks operational, the activity involves a lot of validation that can prevent fraud, revenue leakage, and mistrust.

How does revenue leakage happen?

As we know, there are different payment integrity vendors for healthcare payers – solutions offered to question the integrity of every claim, either before or after the claims processing. But, even after leveraging these solutions, around3-7% of claims are improperly paid. Also, experts estimate that around $200bn seems wasted on excessive or unnecessary medical testing and treatment.

The volume of medical claims became a challenge to detect fraud; Medicare processes around 4.5 million claims daily. Amid the pandemic, fraudulent transactions in healthcare increased drastically. Estimates suggest that US healthcare loses around $300bn annually, i.e., 10% of US healthcare spending to fraudulent claims. With relaxed telehealth circumstances, increased digitization of healthcare, and increased use of telehealth platforms has already mounted concerns for payers on the growing fraud problem.

What are the best ways to reduce this?

Though we know the sources of revenue leakage, it is often tough to spot them. The payers analyze every transaction to prevent losses from payment integrity, and fraud and assure satisfactory care quality. But these are ineffective in eliminating the revenue leakage as it involves manual rule-based decisions or eliminates specific data points from the analysis. Hence it is crucial to have AI and advanced analytics to define the outcomes as fraud or waste for the healthcare payers.

Pattern Recognition to Reduce Fraud and Waste

Many payment integrity and fraud assessment solutions available in the market use data analysis and automation to reduce the workload days. Still, they endure manual verification and rule-based decisions. As they analyze individual patient and provider data independently to identify fraud, misuse, and detect anomalies, the data is limited to medical records.

AI for healthcare can include data like imaging, pharma, behavioral, financial, and outcome-based data apart from claims and EHR data for each patient and collate it to detect patterns from historical data. As the claims data is increasing significantly and other information is also skyrocketing from different systems, it would be difficult to identify patterns for simple data analysis systems. Hence healthcare has a better opportunity to rebuild its systems with minimal revenue leakage by leveraging AI.

Suppose healthcare organizations – payers and providers can leverage AI correctly; in that case, it has the potential to become an efficient and reliable solution that can mitigate improper claims payments, improve care quality, and save billions of dollars every year. Most importantly, AI can detect patterns and anomalies in seconds and not hours, unlike the current payment integrity and fraud systems.

AI for Healthcare Payment Integrity

1.    Claims Processing

Healthcare providers and payers consider payment integrity a key focus area for compelling AI use cases in the industry. Ensuring the confidence of both payers and providers with accurate and relevant payment for the care plan is the primary focus area of payment integrity. AI is poised to transform claims detection, investigation, and recovery processes while the claims are being settled between payers and providers. Let us discuss the document intake, analysis, and decision-making in this section while figuring out fraud and care quality in upcoming sections.

Information and documents intake and analysis – Though it looks simple, the claims document usually has a lot of information that needs to be processed and validated. By leveraging NLP, AI, and ML, you can automatically process the unstructured data to a structured format and reduce medical coding errors. By classifying documents and standardizing data elements, you can minimize documentation problems. You can also include other data from different documents like claims history and patient records for comprehensive analysis.

Predict and Prescribe – Once the data is processed and aligned for analysis, AI can predict the investigation queue, prioritize the claims queue for immediate processing, or further drill down based on the learned attributes from the claims processing history. Prescribing solutions powered by decision intelligence adds more value and saves a lot of time in the processing. For example, a prescriptive solution can guide the subsequent best actions. For instance, if the confidence level for the claim is at 99.9%, you can automatically trigger auto payment without any manual intervention.

2.    Fraud Detection

As per BCG, you can have less than a 1% chance to recover funds once the claim gets paid. And McKinsey estimates that healthcare organizations can save around $30bn by implementing AI and ML in the payment integration programs. Fraud is significant in healthcare and consumes about 10% of US healthcare spending. The pandemic has also resulted in more claims and fraud with the changes bought by telehealth and digital tools.

Gartner says that addressing the root causes like data quality and earlier transaction surveillance will only solve the fraud challenges to some extent. A key trend, multi-dimensional analysis, powered by AI and ML, leveraging data from all disparate sources, will potentially determine the fraud activity for healthcare payers.

How to do a multi-dimensional analysis?

As the claims data increases for payers, other data sources also emerge for a holistic claim analysis to prevent fraud. The most important being – behavioral data, claims history, clinical and personal health profile, among many others. As we use AI for multi-dimensional analysis, it gets easier to detect the patterns from all this data to find anomalies in seconds. We have discussed fraud detection methods in detail in our earlier blog.

3.    Care Quality

Many experts say that around 40% of healthcare spending seems wasted on unnecessary procedures and diagnoses. Why? It is often difficult to spot these expensive and unnecessary treatment options.

As per McKinsey, payers can save 10-20% if they can identify the overpaid claims using AI and ML. The savings can increase to 30% if they leverage automation with AI and ML in the claims management process.

The treatment plans are getting complex with new tests and services for complicated diseases. It is getting harder for payers to analyze these claims to validate necessary tests or services. With AI, payers can identify the pattern recognition and review the payments related to the Diagnosis Related Group (DRG) to provide insights about the care quality. Additionally, you can recoup the expense by predicting the readmission risk, level of care, and quality of service for each code in the medical documentation.

How do we help you?

With over 20+ years of expertise in data, AI, and emerging technologies, we have created an end-to-end solution that leverages multiple data sources to avoid revenue leakage and combat fraud.

Our outcome-based services can power your AI-led transformation with customized pilots and AI services as per your need. Do you still have questions?

You can leverage our 4-week pilot program to experience the AI assessment, consulting, and decision intelligence approach. Contact us for more information.

Khalil Sheikh

Khalil Sheikh

Khalil Sheikh is the Executive Vice President of Solutions and strategy at Saxon. Under his leadership, Saxon is building ROI driven Data Science AI/ML and recommendation engines for its fortune 1000 customers. Khalil has extensive experience in the Software & IT services industry and in turning around businesses through actionable business intelligence leveraging AI. With more than 28 years of experience across the industry verticals, he has led several successful Data Science & digital transformation journeys for ISV’s and enterprises. He led CTO/CIO forums in the valley and is known for fostering creativity, collaboration, and diversity.