Trends

Machine Learning in Claims Processing

By October 30, 2019 No Comments
Machine learning in claims processing

Challenge

Healthcare costs across the world have soared over the past decades at a rate at which, where United States spends more on healthcare than the national budget of half of the countries in the world. A big chunk of those expenses relate to pharmaceutical products.

AI and IoT in Healthcare: Need of Future

Opportunity

Here is the opportunity for Machine Learning Solutions – in US, by and large, most pharmaceutical transactions are captured electronically as claims. These claims, and the way they are processed, cover a myriad scope of information including patient demographics, disease states, drug utilization review, formularies, coverage and utilization review, contra-indications, etc. This information can be used for a number of reasons – better pricingof drugs based on their utilization and volume, better prediction of diseases and therapeutic journeys where we can guess over time which drugs a patient will require, and a more interactive engagement of the patient using virtual “friends” to guide them through their therapy and ensure compliance, adherence and better outcomes.

IoT can truly Transform Rural Healthcare in India

Over the past several years, solutions have been operationalized working with payors, PBMs on cost and efficacy predictions for new therapies for HIV (PrEP treatment) and Hemophilia (Gene Therapy). Another application is successfully helping Pharmacy Benefit Managers predict their most efficient drug pricing for patients who are not covered with insurance, cost efficiencies for seasonal and style drugs to name a few. GxWave™ leverages LSTM algorithm (Long Short-Term Memory) in predicting price efficiencies, claim volumes, call center call volumes, and average margins.

Approach

Solutions such as GalaxE’s GxWave™ platform with solutions such as Claims Neurology described below, utilizes their proprietary technology to extract business rules from adjudication systems and then use prior claims data to predict various “edits” or applicable rules such as prior authorization (where the use of a specific drug requires express approval from the physician) or adjustments and accumulators so that across their therapy, the price they pay is properly adjusted for the full range of medicines and medical services the patient consumes. These pathways for the adjudication of a claim are then trained on neural nets that learn the time based, formulary based, disease and therapy-based trends.

AI and Cybersecurity in Digital Healthcare

The trained nets are then used to predict and project the therapeutic journey of a patient, or the volume and timeframe for the consumption of specific therapy in a given market.

With all the consolidations of drug & device industries in play, their efficiency has to be at the highest point in order to drive patient costs lower. GxWave™ is helping these entities with improving efficiency. With a combined data set comprising elements of documentation across the development lifecycle of an application with governing procedures and its intended use, natural language processing techniques can derive information from previously unexplored data sets that can be analyzed to ensure compliance to regulations, adherence to organization policies and procedures and alignment with the documented intended use of the system. Healthcare IT consists of diverse applications with multiple critical integrations at various levels where regulatory impact can be ambiguous and could be left unassessed. A change in the landscape has to be simultaneously assessed for regulatory and risk impacts (includes business, security and privacy risks) without delays ensuring all impacts are being planned for before the change is implemented.

Healthcare IT market builds the foundation of Artificial Intelligence

Advantages

A number of advantages have surfaced with these solutions:

1. Immediate re-categorization or inspection of current non-regulated applications that could potentially be regulated due to change in feature/functionality.

2. Expose deficiencies within the system that could lead to a potential replacement for not being able to satisfy customer and regulatory requirements.

3. Allow for continual monitoring instead of the traditional approach of conducting periodic reviews.

4. Improve software documentation quality across the application lifecycle.

This is just the beginning. We expect that solutions like GxWave™ will utilize data from genomics, patient profiles, therapy histories and help generate the most medically and economically rational and effective therapies for patients in the very near future!

About the authors

Sandipan Gangopadhyay is the President and COO of GalaxE and plays a key role in GalaxE’s continued worldwide expansion and operational success Prior to this, Mr. Gangopadhyay spent over a decade in high profile roles in both Pharmaceutical and Information Technology companies around the globe and instrumental in setting up one of India’s first private Software Technology Parks. He has a Bachelor’s degree in Computer Engineering from Bombay University, is a member of the Indian Institute of Chemical Engineers, and is certified in the Governance of Enterprise IT.

Dheeraj Misra is the Chief Technical Officer and Senior Executive Vice President of GalaxE and has over 15 years of experience in the design, development, testing, porting and maintenance of application and system software for the healthcare industry. Prior to this, he spent a number of years in high profile roles at HCL Technologies, Context Integration, Eforce Global, and as a Research Specialist in Parallel Processing. He has a B.E. in Computer Engineering from REC, Allahabad.

Sandipan Gangopadhyay

About Sandipan Gangopadhyay

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: