In 2018, Venmo suffered from a high volume of fraudulent transactions. Instead of taking measures to tightly control its data, the company responded by simply eliminating certain features. Only 2 percent of customers conducted transactions via the Venmo website, but the website accounted for 15 percent of net losses. Moving forward, the peer-to-peer payments company will only allow transfers via its mobile app.
This is the type of fraudulent activity we often hear about in the news — large tech companies mishandling their customers’ data. But it’s nowhere near the most damaging or the most prevalent form of data fraud.
Consider healthcare. In the past two years alone, 89 percent of healthcare organizations experienced some loss of control over their data, creating enormous costs. Across all industries, the average cost per compromised record is $148, but healthcare must contend with a staggering $408 per loss. That’s nearly three times higher than the average, and it’s the highest cost of any industry for the eighth consecutive year.
The nature of the healthcare industry is what makes it particularly susceptible to fraud, waste, and abuse. Three key pillars within it (patients, hospital staff, and data management personnel) are often siloed. Even though they manage and interact with the same highly confidential data — sometimes via personal devices — they do not share a common communication platform, which increases the likelihood of a data security breach.
When poor data management processes go unchecked, fraud has an opportunity to slip right through the cracks.
Figuring Out Fraud
Traditionally, when we think about healthcare fraud, a practice known as “phantom billing” comes to mind.
For example, an anesthesiologist could claim to administer more units of anesthesia than he or she actually did, then pass fraudulent billing onto the insurance companies. Similarly, a dentist could write multiple insurance claims for X-rays that patients never received. It’s unrealistic to expect insurance companies to reach out to patients and confirm that they got X-rays. Instead, they foot the bill, fund the fraud, and ultimately pass the costs on to patients.
Another more insidious form of healthcare fraud occurs largely because of sloppy data management, miscommunication, and inefficient business practices. To illustrate, let’s say a patient stays in the hospital for 20 days and incurs a total bill of $100,000. If the hospital makes a mistake and submits the bill without key information, it can cause delays that add significant costs into the system.
In this case, the insurance company will hold the bill for around 15 days while it tries to unsuccessfully process it, at which point it goes back to the hospital with a request for more information. When the back-and-forth process drags out for 60-plus days, thousands of dollars are wasted on administrative fees. Furthermore, the $100,000 bill becomes vulnerable to fraudulent practices such as balance billing, which occurs when hospitals illegally bill patients for outstanding balances after insurance companies submit their portion of the bill.
This fraud may not have been carried out by individuals with bad intentions, but that distinction matters very little to those who suffer the consequences.
The Future of Fraud Detection
The pace at which fraud techniques advance and evolve means that we need to stay two steps ahead at all times. At its core, limiting fraud is about controlling information, whether it’s customer or corporate data. More specifically, prevention and detection require a mix of stricter internal data management policies and better data management systems.
Efficiency is an enemy of fraud. With a seamless and comprehensive data management system that integrates data from multiple sources (e.g., patient to doctor, doctor to data management system, data management system to insurance company), time-intensive mistakes can be kept to a minimum and fraud can become easier to prevent.
Organizations are also starting to tap machine learning solutions. Algorithms can comb both structured and unstructured data to look for and flag anomalies. Healthcare organizations can then rely on the experience of just a few staff to verify fraudulent transactions that might otherwise have remained buried beneath the surface. The power of artificial intelligence enables fraud to be detected and remedied much earlier in the process, saving the healthcare industry and its customers a lot of money.
While the incorporation of machine learning and artificial intelligence into the fraud-detection process is exciting, the reality is that many healthcare organizations are too reliant on legacy platforms to immediately take advantage of these technologies. To step into the future of fraud detection, these organizations need to migrate to modern platforms that support intelligent data management.
For organizations that do, fighting fraud will be easier than ever, and more than that, they will gain a competitive advantage through cost savings and improved patient experience. The business case is clear; it’s time to bring healthcare fraud detection into the 21st century.