Innovation Transforms Utilization Management and Healthcare Future
The healthcare industry is facing a number of challenges, from rising costs and insurance premiums to workforce shortages, cybersecurity threats and complex payment processing and invoicing systems.
With administrative costs accounting for one-quarter to one-third of the $4.3 trillion spent on healthcare in a year, utilization management programs are critical to reducing these costs relative to overall program management spending. Utilization management is widely recognized as a valuable process for evaluating the medical necessity, efficiency and appropriateness of healthcare services, procedures and facilities.
From a care quality perspective, utilization management helps Medicaid programs and health plans achieve the “4 Rs:” delivering care at the right time, in the right place, to the right patient, at the right cost. Here’s how innovation is transforming the process.
Advanced Tools of the Trade
Artificial intelligence and machine learning (AI/ML) tools use advanced data analysis to continuously learn over time — with little human intervention required outside of providing the data. It’s exciting to consider AI/ML’s ever-accelerating value in healthcare, particularly when it comes to addressing administrative burden in today’s overstretched workforce.
AI/ML is reducing administrative costs by enabling greater clinical efficiency, significantly reduced manual labor, increased diagnosis speed and accuracy, and better use of data and resources. When it comes to utilization management, these tools can greatly improve and streamline utilization management processes, such as case identification and decision making. By connecting disparate data, AI/ML creates a more complete patient profile, finding things we may have missed otherwise and better informing the decision-making process.
Streamlining the Healthcare Administration Continuum
When applied to utilization management, AI/ML can help states and healthcare organizations achieve numerous benefits, including:
- Improved patient outcomes due to more accurate information and decision-making support, leading to better patient care
- Reduced medical costs by ensuring patients are getting the right care from the right physician at the right place at the right time, eliminating overspending
- Improved access to data analytics, which can be used along with machine learning to create models used within artificial intelligence
- Reduced administrative burden and costs through more efficient processes and more readily accessible and comprehensive conclusions
Identifying Patterns, Raising Flags
AI/ML can see outside typical patterns. Plus, it has the capability to continually learn from corrections to case decisions, expand on any rules put in place and pull from medical necessity guidelines to reach increasingly smart, complex conclusions. The results will help improve patient outcomes.
For example, a cardiologist is recommending knee surgery for a patient rather than a heart-related procedure. This might raise questions since it’s not that physician’s specialty. AI/ML can analyze historical requests and patterns to support that recommendation.
Another example might be a service recommended for a patient. Based on the metadata and a review of medical necessity criteria, the service is not recommended for a patient under 18 years of age. This patient is child. AI/ML was able to identify important decision-making criteria in this case.
Going Beyond Rules-based Engine Abilities
Equally as important to understanding AI and ML’s abilities is also knowing what they cannot do. They cannot provide systemic decisioning based on rules engines.
A rules engine may be the foundational knowledge of the “brain,” but it’s not AI. It will do exactly what you’ve told it to do and nothing more, nothing less. It’s important to be informed and understand that capabilities marketed as AI/ML may not, in fact, be AI/ML.
AI/ML leaves it to the Medicaid enterprise system or other or other healthcare information systems to think about parameters based on the metadata such as patient or physician information and diagnosis and service information. This includes historical decisions, corrections to past systemic decisions, trend analysis, the latest medical necessity criteria and other metadata. The system is not merely running data through a rules engine and by doing it automatically, it is increasing efficiency and reducing administrative costs by reducing manual work.
Used in software for medical devices, AI/ML is able to learn from real-world use and improve its performance, assisting healthcare providers and improving patient care. Virtual nursing assistants use AI/ML to reduce hospital and clinic visits by performing timely checks via voice — managing progress and lightening the burden on medical professionals.
Healthcare AI/ML Future
Healthcare AI/ML will continue to gain traction as this technology incorporates into nonclinical administrative processes, medical devices, patient engagement and medical training. It presents limitless potential to maximize knowledge, resources and efficiency within the healthcare industry.