By Navaneeth Nair, vice president of product, Infinx.
Recently, major U.S. health insurance companies have begun implementing prior authorization processes. These insurers do not possess the specialized knowledge needed to determine medical necessity, so they have to shift the responsibility to providers in order to minimize instances of fraud and abuse.
However, shifting prior authorization responsibility back to the practitioners can be troublesome. Many providers do not have the necessary staff headcount needed to handle the seemingly endless number of phone calls needed to obtain manual insurance verifications.
That may lead providers to seek completely automated solutions as a solution, but—despite all the advances in artificial intelligence (AI)—machines simply cannot do everything. The true purpose of AI is to help people make better decisions rather than completely automate tasks and remove the human element.
Decision-making in the healthcare industry absolutely requires the intelligence and empathy that can only come from people and not machines. As a result, it is more accurate for providers to look to “augmented intelligence,” which is the enhanced capabilities of human clinical decision-making coupled with AI’s computational methods and systems as defined by the American Medical Association.
Make no mistake—AI is absolutely necessary in healthcare because the industry is either swimming or drowning in data, depending upon whom you ask. As a result, the value proposition with the most potential is to implement a tool that can input the data, make sense of it, and present it back to the provider in a way that allows people to use their knowledge and empathy to make the best possible decisions.
The AI component is designed to continuously learn and improve from all data and interactions to provide prescriptive insights for decision making while providing an increasing transparency into the process. Ultimately, the AI’s job is to enable intelligent decisions by applying machine learning, natural language processing and understanding. When decision making is more deductive based on a learning system, it improves outcomes.
Predictive models are inherently simple to build, but difficult to maintain, because none of our healthcare processes remain stable enough to utilize the data and patterns that are produced. Those aspects are always changing, so practices need a solution that seamlessly integrates the process and constantly accesses the latest and most relevant data.
AI should be prescriptive and not just predictive since predicting the outcome without explain ability (black box AI) is limiting, particularly in healthcare where we need to better understand machine rationale before applying it.
Lastly, the human component is required to make the decisions that robotic automation and AI cannot. Only people with empathy and knowledge of each unique situation can tailor the patient experience to their individual context and needs.
Simply put, prior authorization needs to be a product of human determination with as much assistance as AI can provide. The best possible outcomes are the result of good intelligence and great execution.
The most significant value proposition for AI is the ability to take mundane administrative processes like prior authorization or revenue cycle management off the task list of staff members who have more important things to focus on without letting machines make critical healthcare decisions. The most valuable automation systems will continuously learn from the data gathered from previous experiences in order to help people make better decisions.
This “digital workforce” will not replace human workers, but instead will accelerate employee expertise, augment decision making, reduce manual processing costs and risk, increase consistency of output, and develop continuous self-learning processes. These best-in-class platforms will allow providers to manage prior authorization, coding, and billing needs through a combination of AI and specialists.
Providers should avoid completely automated solutions for prior authorizations and instead look to a solution that does not remove the deep knowledge and expertise factor gained from specialists. Many general-purpose AI solutions are nothing more than spot analytics solutions branded as AI, which require significant investments and deliver uncertain results.
While AI may one day evolve to the point of completely automating the prior authorization process, providers today should remain committed to incorporating the human factor for better decision making and outcomes.