Robotic Process Automation – Futureproofing The Health Insurance Industry
May 21, 2019
AiRo Digital Labs wins Bot-a-Thon in Automation Anywhere Partner Bootcamp, New Delhi, 10th June 2019
June 11, 2019

Predictive analytics to reduce Payers losses in FWA

Very recently, losses exceeding USD 1 billion were reported in one of the biggest frauds ever in the history of American health insurance. Though not a directly payer fraud, it was engineered by an international telemarketing network involving call centers in the Philippines and Latin America.

It involved millions of patients who were already registered under the Medicare public insurance system and offered a scheme providing wrist, knee and other braces to elderly and/or disabled patients.

The trick was to approach doctors, pay them for prescribing braces for patients that did not need them, and then presenting bills to Medicare, which finally amounted to more than USD 1.7 billion. The US Department of Justice finally caught up with them and charges were announced against 24 people for fraudulent activities.


It was a multi-billion, multi-state operation and the accused ranged from senior executives with telemedicine companies, the owners of dozens of medical equipment firms to- of course, doctor. The losses were totted up to more than USD 1.2 billion and administrative penalties were enforced against 130 orthopedic equipment suppliers, according to DoJ.

While this is one of the biggest scams that can be called FWA, errors and other illegal activities that require massive payouts are not uncommon in the healthcare payers’ space. Case in point was an organized medical fraud in July 2017 that took down USD 1.3 billion, involved over 400 doctors, nurses, and pharmacists. The accused billed the Medicare, Medicaid, and other health insurance programs for drugs ordered and not purchased or given to the patients.


Totting up the bill, the losses can be staggering – the National Healthcare Anti-Fraud Association in the US estimates about USD 80 billion annually. These are official estimates, but the actual figure- with unreported and undetected losses, could be as high as over USD 200 billion. Estimates say the losses can amount to between 3 to 10 percent of the total annual spending on healthcare in the US.

But there’s more. Only about 5 percent of these losses are ever recovered, making them a challenge not only for the healthcare payers sector but also a burden to carry for taxpayers.

While there is no panacea for this particular evil, healthcare providers use a number of tools to fight these frauds and oversights. No payment or policy is foolproof, so the best payer companies can do is to be alert to the loopholes and weaknesses in the system. In practical terms, the goal has to be to retain the losses due to frauds to the limits of the industry norm- which is 3-5% of the annual payments.


If identifying loopholes and risks be the objective, predictive analytics can help in highlighting the risks at all levels. In fact, over the 4 quarters, the use of this technology has grown by almost 13% in the community, says a survey report. A recent new survey of 200 provider and payer executives from the Society of Actuaries shows an increasing usage of predictive analysis, growing 13% from 2018. In the survey, almost 60% of respondents saying they are using the technology in their organization. They plan to devote almost 15% of their IT spending to this technology in 2019 as well since they feel the investment will pay almost 15% savings over the next five years.

Of the rest, almost 30% say they plan to implement it over the next 12 months to 5 years.

Data visualization and machine learning, two very critical components of predictive analytics, will definitely drive strategies for fraud and irregularities’ detection in payouts. Costs will be reduced not only by increased efficiency but also by the acquired ability to notice anomalies, unnatural data states, and unexpected incidents.

Using predictive analytics also reduces the human error factor- data collation or filing by human resources that could slip up and create massive logjams, by way of unintentional fraud. The new parameters allow an enormous amount of insights into strengths, weaknesses, integrity, and loopholes in the payment system.


Systemic weakness in the adjudication and payment system that can be exploited by providers, is the greatest risk where most financial compromise can happen. It has a high chance of being identified by strong predictive analytics algorithms. The way to do this would be to score claims, providers, and processes for their level of risk propensity using predictive analytics and machine learning parameters and then identify the areas where the probability of losing money is the greatest.

While there are a number of predictive analytics tools that could help with this endeavor, an ideal one would be that would encompass all the various organization teams that work together for any payout. Including data and ensuring stakeholder buy-in from teams like Legal, Claims Operations and Provider Relations would help create a strategy that will leverage technology as well as strategy to minimize losses due to FWA.

Comments are closed.