Dec 31 2009

Setting Priorities

To resolve improper payments, try the approach HHS takes to nab low-hanging fruit first.

Photo: Randall Scott
Ultimately, HHS expects to combine data analysis­ and data mining so it can even identify problems at a specific doctor’s office, the department’s Charles Johnson says.

Last March, during a congressional hearing on improper government payments — overpayments to individuals, organizations, businesses, states and local governments — Sen. Tom Coburn (R-Okla.) sounded an unusually panicky note: “I think if the American people heard some of these numbers, they would vote us all out of office.”

He pointed out that the estimated $83 billion of thrown-away money over the previous two years could have helped pay for the war in Iraq or funded Katrina relief four times over. “This subcommittee is going to keep having hearings on the subject,” he vowed.

Sweet and Sour

One of the people on the subcommittee’s hot seat that day was McCoy Williams, director of the financial management and assurance team at the Government Accountability Office. He recalls delivering a somewhat positive, but qualified message. “I was able to report to the committee that in fiscal 2005, we had a governmentwide reduction of $7 billion. That was good, but unfortunately it came from one agency: the Health and Human Services Department.”

Although each agency will have its own set of challenges when attempting to reduce improper payments, HHS’ overall strategy is widely applicable.

“We decided that first we were going to find the low-hanging fruit,” says Charles Johnson, assistant secretary of HHS for budget, technology and finance. Many agencies also may find relatively easy fixes that provide a lot of bang for the buck. That’s a great place to start, Johnson and others believe.

48.5% reduction: HHS’ paid claims error rate dropped from 10.1 percent to 5.2 percent in a year once data analysis found many lacked supporting documentation.

To ride on HHS’ coattails, agencies need to set up a program to help identify and then grab hold of that low-hanging fruit. Phase 1 requires identifying the challenges of finding the data. During Phase 2, the agency must resolve any systemic problems and gather the data. Phase 3 involves sifting through the information to spot payment trends and possible causes for improper payments. Phase 4 kicks off the efforts to right the wrongs leading to errant ­payments.

An Aha! Moment

Each year, contractors for HHS process more than 1.2 billion Medicare claims. They are responsible for checking to ensure that each claim covers medically necessary services for eligible individuals — the very categories in which improper payments could be hiding. But, ultimately, meeting the mandate in the President’s Management Agenda to eliminate such improper payments falls on HHS.

The first step to any solution is to find the specific challenges. And with HHS facing more than a billion opportunities for errors, it needed a scientific way to locate and prioritize the major issues. Accordingly, it developed the Comprehensive Error Rate Testing (CERT) program, which analyzed 160,000 claims randomly selected by the department. Why 160,000? Because HHS adviser PricewaterhouseCoopers deemed that number statistically sound for evaluating improper payment problems over the entire claims process.

Most of the analysis must be done manually. HHS will accumulate the results in its new Healthcare Integrated General Ledger Accounting System, built on Oracle Federal Financials (HIGLAS). Besides providing a national paid claims error rate, the program categorizes the claims rate problems related to contractors, to providers and to regions. So far, it doesn’t rate individual contractors or providers.

As HHS began the analytical process, it quickly found one glaring issue. There were a large number of claims filed without documentation. “It’s a very easy thing to see: Either there’s documentation or there isn’t,” Johnson says. Accordingly, HHS put the contractors’ feet to the fire. It held a series of meetings and used Web and in-person training to help contractors reduce the incidence of no-documentation claims. Contractors then relayed the pressure to health-care providers, who also were given a carrot: The time needed to submit documentation was extended from 55 to 90 days. Williams says just this one effort about halved the paid claims error rate from 10.1 percent in fiscal 2004 to 5.2 percent last year.

The Big Payoff

Carla Lewis, assistant director of the financial management and assurance team at GAO, believes the primary lesson learned from HHS is not so much that documentation may be a relatively easy problem to fix — although it may also be a problem at other agencies — but how agencies focus on resolving their improper payments. “Agencies should consider prioritizing their problems by scope and ease of fixing the problem,” she says. That’s the real payoff HHS got from its documentation effort, Lewis adds.

Eventually, Johnson hopes to track individual providers, contractors and even Medicare recipients. For example, he expects that through data mining, he will be able to find patients who “doctor shop” — get treated for the same problem by multiple doctors who in turn kick back part of the fee to the patient. But, he points out, getting to such a granular level of analysis may require a law mandating that providers use standardized electronic medical records. Health-care policy-makers and legislators are moving in that direction, but he doesn’t expect any real traction and data access for his purposes for a number of years.