Aetna Prescription 1

Integrate (Among Other Things, to Make Sure Duplicate Payments Cease)

Aetna’s customer and profit woes largely revolve around its ability to process claims fairly, efficiently and cheaply. “Reducing the labor associated with paying claims is a critical lever with regard to being profitable,” says Brad Holmes, an analyst with Forrester Research. But Aetna’s progress is hindered by multiple claims systems.

Integrating these systems was supposed to follow Aetna’s acquisition of U.S. Healthcare, known for systems that can handle large numbers of new users. It hasn’t happened.

Former CEO Huber boasted of moving the business onto one common platform. Where Aetna truly succeeded was with its data warehouse, fed by claims data from all products. However, the integration of operational systems didn’t extend to Aetna’s non-HMO products—including less restrictive forms of managed care that customers increasingly demanded. Also, some strengths of the U.S. Healthcare systems, such as autoadjudication, relied on the relative simplicity of the HMO business model. Those systems didn’t have the flexibility to support Aetna’s more tailored benefits plans. Older systems had to be retained.

“Huber was focused on buying market share for economies of scale, but that’s theory. You have to integrate claims processing to achieve any of it, and they didn’t do it,” says JP Morgan’s Price.

Aetna also failed to make sure its prices kept pace with increasing medical costs. “They essentially mispriced for several years in their at-risk HMO business,” says Greg Crawford, head of health care research at Fox-Pitt, Kelton Inc., an investment bank in New York. Aetna should have been able to see the cost trends by analyzing claims data, and its failure to do so reflects the fragmentation of claims systems, he says.

Stuck With Aging Systems

Systems issues also have caused Aetna to overpay some accounts. “We are overpaying substantially,” Rowe said in a speech last year at a Sanford C. Bernstein & Co. investors’ conference. “And my guess is there’s more to be discovered.” Aetna was losing millions because providers who resubmitted claims for lack of payment sometimes wound up being paid two or three times. Some smaller accounts were paid even though their policies had expired. These problems reflected the inability of its databases to share information or identify duplicate records. Aetna refuses to discuss how this has been addressed or how much waste it has eliminated.

In a November 2000 court filing, Aetna had to respond to allegations that it actually manipulated its systems to slow payments. “Aetna does not use computer software to administer claims processing in even a remotely uniform fashion,” the company claimed in its own defense. In the filing, Aetna said it maintained four claims systems. While HMO claims were processed through systems that try to automate decisions based on business rules, claims for PPO (preferred provider organization) plans were run through an older system that did little more than display claims on a screen so a clerk could make the judgment of whether to pay.

All four systems are still in use. Integration is highest in its HMO business, where all claims are handled by the system inherited from U.S. Healthcare. Elsewhere, however, Aetna continues to use two systems incapable of autoadjudication, called Aecclaims and the Managed Choice System. However, Aetna says it is moving to have all non-HMO claims managed by its Automated Claims Adjudication System (ACAS) by November. That would leave Aetna with two healthcare claims systems, instead of four.

Performance also has improved. Under the First Claim Resolution program, Aetna cut its HMO claims processing time to 4 days by April, down from 5.7 days a year earlier. Today, 62% of HMO claims are autoadjudicated, nearly double the rate of two years ago.

Holding onto existing claims systems is typical, says Ingram at Meta Group. “Companies don’t replace them very often because it’s a very disruptive, very high-risk process,” she says. Instead, they add on accessories such as fraud detection, she says. “It ends up being very piecemeal.”

Forrester’s Holmes believes Aetna will have to take the risks, however large, if it truly wants to lower administrative costs while eliminating hassles for consumers and providers. He suggests the need for a next-generation operating environment.

There are risks there, too. Although Aetna is a big user of IBM’s DB2, its claims systems continue to make significant use of Computer Associates’ IDMS, a mainframe-only data storage technology. IDMS has a reputation for being high-performance and reliable, and CA does sell an add-on product that allows it to mimic a relational database. Still, finding experts in the base technology is becoming harder as time goes on (see story at left, “Gotcha! Downsides of Using Mainframe Databases”).

“Aetna has been working on a major conversion to DB2 and some of that has happened,” says Del Barlett, a former U.S. Healthcare IDMS administrator who left about a year after the Aetna merger. “When you have the whole system revolving around IDMS, it is a monumental task.”

Because the principles for optimizing a traditional mainframe database are entirely different for a relational database, applications have to be rethought rather than merely migrated, he says. Barlett says he remains “an IDMS bigot” because he likes its performance and reliability—even though he recognizes that DB2 has advantages in terms of flexibility for ad hoc queries and developing new applications. The trade-off is that DB2 requires more computing power and administrative manpower, he says.