Healthcare insurance providers are navigating new ground and capitalizing on new opportunities made possible through artificial intelligence. AI can be utilized to lower costs for payers and improve the member experience as well as the overall health and well-being of members, leading to greater member satisfaction and improving trust and reputation for the payer’s brand.
There are many ways to leverage machine learning in the healthcare insurance space. We’ll focus on three major areas: claims processing, including limiting fraudulent payments, member engagement, and streamlining operations to reduce costs.
Claims Processing Automation
Claims processing has long been one of the most labor-intensive areas for healthcare payers. Manual review of health insurance claims can be eliminated or significantly reduced through the use of AI and machine learning.
1. Claims Review and Processing:
AI is already being used in claims processing to minimize the need for human involvement. Through Natural Language Processing (NLP), technology can be used to extract the necessary information from submitted claims and medical records, to verify the accuracy and validity of claims. This “first pass” can automate approval for the majority of claims, increasing the speed of the claims review process (by 75% in one study)1 and improving member satisfaction.
2. Predictive Analytics:
A recent study of health insurance claims processed by a mid-sized payer (1.5 million members) found that 70% of all claims were flagged as unusual, requiring manual review2. Machine learning can take aggregated data from how such claims were handled and create a dynamic response, rather than relying on static rules that may not accurately assess each situation. An Accenture study found that by utilizing predictive analytics in claims processing, insurance payers could save $15 million per 100 full-time employees over 18 months3.
3. Fraud Detection:
Correctly detecting fraudulent claims is a top priority for insurance payers. Incorrectly rejecting a claim as fraudulent can result in delays in necessary treatment, create a financial burden for patients, and may result in a loss of trust and a damaged reputation for insurance payers. However, failing to flag and reject fraudulent claims means increased costs and diminished resources.
So far, one of the largest investment areas in AI technology among payers has been fraud detection1. For example, health insurance disruptor, Collective Health, boasts a 99% accuracy rate in claims processing thanks to its use of machine learning technology4 to detect fraud4.
Member Engagement
One of the most promising areas for growth with AI involvement is member engagement.
1. Member Service Chatbots:
Most member service inquiries involve policy information (questions about coverage, providers, etc.) and claims status. Answers to these types of questions can be difficult for the average consumer to find on their own; however, through NLP, chatbots can access a member’s policy and healthcare records to generate personalized answers to specific questions (resolving questions successfully 97% of the time, according to one case study)1. What’s more, these resources can be available to members 24/7 without any additional operating cost, resulting in improved access and greater member satisfaction.
As one example, Cigna introduced its “Cigna’s Answers” chatbot in 2017. According to the company, this chatbot, coupled with Cigna’s digital service platform, One Guide, resulted in a 20% increase in member satisfaction1.
2. Member Insights:
Harnessing AI technology can not only help address member questions and concerns in the moment but can also predict what interventions and guidance a member might need, or if a significant medical event might be imminent. This predictive ability can prevent the need for more costly medical treatment, and of course, provide a better experience for members.
Using aggregate health data, AI can identify potential gaps in care, resulting in a suggestion to members or healthcare providers to address that area. Payers could also access biometric data through wearable technology, obtained in exchange for reduced premium rates, for example. A study by Troubadour Research and Consulting found that nearly half of those surveyed said they’d be willing to share biometric data with their insurer to receive a discount on premiums5. Coupled with medical records and claims reports, this additional data could provide greater insight into necessary preventative care, as well as fine-tuning treatment standards.
One interesting case study in how AI can provide member insights is from technology-based health insurance provider, Oscar. In an experiment through the Oscar Medical Group, which is a group of 120+ medical providers offering virtual care to subscribers, AI-generated summaries of lab reports were offered for doctor review, before being relayed to patients in natural language. The study found that medical providers were able to approve AI-generated patient instructions with minimal changes in about 24% of cases, saving healthcare providers time and preventing errors. Oscar expects this rate to increase as the AI system has access to greater context and learns from doctor corrections. In this way, doctor-patient interactions are streamlined, and the data available to payers to improve member insights is expanded.
Operations and Cost Reduction
Along these same lines, AI can help insurance companies create personalized plans for their members, review medication usage (offering more cost-effective solutions), and optimize insurance workflow to reduce administrative burdens and costs.
1. Personalized Healthcare Plans:
Using AI and drawing on patient data and preferences, health insurance companies can offer members personalized plans that can best meet their needs while lowering premium costs. Consumers have come to expect an e-commerce-like level of personalization in their other digital interactions and will have greater satisfaction with their insurance provider when they perceive their plan as efficient and comprehensive to their needs.
2. Drug Utilization Review:
By aggregating data on prescriptions, treatment plans, and what pharmaceutical claims are submitted, AI can monitor medication usage and optimize the use of medications in client treatment. Utilizing AI for this review saves costs, provides greater insights, and can ensure the most appropriate treatment for members.
3. Workflow Optimization:
In addition to claims review and processing, AI can be implemented in other areas of payer workflow in order to create greater efficiency. For example, repetitive tasks in underwriting, billing, and compliance can be identified and automated through AI in order to reduce administrative overhead.
AI technology is a boon for both health insurance providers and their members, providing a host of cost-saving implementation options, which also have the potential to create a more user-friendly and personalized experience for consumers, and ultimately improve their healthcare experience.
Resources
[1] Can AI Cure What Ails Health Insurance? (forbes.com)
[2] Artificial intelligence in health insurance: Smart claims management with self-learning software | McKinsey
[3] Intelligent Health: A Guide to Targeting Your AI | Accenture
[4] Company Overview - Collective Health
[5] Consumers Ready to Give Insurers Biometric Data, For a Price | Digital Insurance (dig-in.com)
[6] AI Use Case: Electronic Lab Review | by Oscar Health | Oscar Tech | Oct, 2023 | Medium
