Technology: Rx for Health Care Fraud Ills
By Diane Davis
Health care is a $1 trillion industry in the United States. Unfortunately, health care fraud is a nearly $100 billion industry.
Health care fraud affects everyone: Taxpayers must pay more to support Medicare and Medicaid; consumers must pay higher premiums or deductibles for health insurance; companies must shell out more to provide employees coverage.
If that money weren't being siphoned off, think of how much health care it might buy for those who desperately need it, or how it might be used to ensure the solvency of Medicare, or how it could be redirected to areas such as education.
For years, good old-fashioned investigative work has been used to ferret out fraud. But this approach is useful only after a crime has been committed. As is the case with health care itself, prevention is a much-preferred approach. Fortunately, technology has advanced so that there are new solutions that can ? and should ? be used to fight fraud.
To understand how this technology can help, it is important to understand the beast that is health care fraud. It comes in many forms.
Claims can be submitted for services that were never rendered or for more serious procedures than were actually performed. Or bills are submitted separately for a group of procedures that are typically reimbursed as one.
Kickback schemes involve hidden financial arrangements between health care providers and patients. In one documented case, underprivileged children were coerced into visiting a local clinic, where they were checked by every doctor for every conceivable malady. After their checkups, the kids were dropped off in their neighborhoods and given $20 apiece for their time. The doctors received hundreds of thousands of dollars in reimbursements.
Technology can help mainly through something called decision-support solutions. These are a group of technologies that use complex databases, mathematical algorithms and often artificial intelligence to transform raw data into more usable information.
Here's how it works: Tremendous amounts of data are generated in health care. Unfortunately, most of it is fragmented among numerous systems. By integrating these various pieces, a comprehensive view of a provider's billing practices can be generated. And by using advanced decision-support tools, patterns of care can be analyzed and payments that shouldn't be made can be identified.
For example, analyzing combined data will expose quickly providers who have billed for more hours than is possible to work in a day. It also can identify providers whose billings are always just under the allowable threshold. All kinds of irregularities can be unearthed.
This technology already is having a dramatic impact in Texas. The state recently implemented a Medicaid fraud and abuse detection system that incorporates advanced decision-support capabilities, including neural network technology or "learning" technologies. So far, the system's "hit rate" is nearly 100 percent, which means virtually every suspect the system identified was engaging in fraudulent activities.
By 2000, the new system is expected to uncover at least 2,000 additional fraud suspects and recover an additional $14 million annually.
The Medicare program has begun to take a similar approach. The Health Care Financing Administration, which oversees both Medicare and Medicaid, is implementing an information system that will integrate data from the Medicare Part A and Part B programs (hospital and physician billing, respectively) as well as from its managed-care programs. Once all this information is integrated, decision-support tools can be used to identify aberrant patterns.
The ultimate goal is to get a consolidated view of a provider's activity. This is complicated when the provider offers services to numerous state and federal programs plus employer-funded insurance programs ? each with its own designation for that provider.
With the help of the National Provider Identifier, which was mandated under the Health Insurance Portability and Accountability Act of 1996, this consolidated view is closer to reality. When fully implemented, this unique identifier will be used to create an integrated database with records of all providers and suppliers who are certified to bill Medicare or Medicaid.
With this tool, providers that have been excluded from one program or another can be flagged in the database. Then before a claim is paid, the database can be accessed for notification of previous fraudulent practices. This will provide a Better Business Bureau of sorts for the health care industry ? something that is badly needed.
When the potential savings are so large ? especially when the money could be used to provide legitimate medical benefits ? moving ahead with such innovations is as logical as it is prudent. For until health care follows the lead of other industries and establishes such technical monitors, it will continue to attract people looking to make an easy buck through unsavory practices.
Diane Davis is the fraud and abuse systems and products manager with Electronic Data Systems' State Health Care unit.