Why the Health Care Exchange Has Been Such a Mess

When federal and state governments rolled out health insurance exchanges on Oct. 1, many voices, including this blog, complained about the technical glitches that cropped up. I wondered, too, why this whole thing was such a mess and whether better testing could have prevented some of the problems. 

But after chatting with Leighton Ku, a Professor at the George Washington University in Washington, D.C., I’ve changed my tune somewhat. I still think that good automation and simulation technology could have helped the exchanges with more robust front-end testing. 

At the same time, though, I now realize it’s a miracle that these initiatives got off the ground in the first place. 

For one thing, this was one of the largest programming exercises in history by federal and state governments. Ku, a Professor in the Department of Health Policy at GWU, said he doesn’t know how much money was spent on the exchanges.

“But if you told me it was in the hundreds of millions of dollars, I wouldn’t be surprised,” he said. “There were thousands of programmers, systems engineers (and others related personnel) across the many places that were doing it.” 

Sheer Complexity of a Healthcare Exchange 

The idea is for the exchanges to be the rough equivalent of an Orbitz or Travelocity for health insurance, according to Ku, whose titles including Director of the Center for Health Policy Research at GWU’s School of Public Health and Health Services. 

Consumers and small businesses, in theory, should be able to compare plans and select the ones with the right combination of prices and features for their particular circumstances.  People close to the poverty line or who are part of the “working poor” may also eligible for a federal tax credit that should make the purchase of health insurance more affordable for them. 

However, several problems have cropped up, Ku said. 

For one, the health reform law allows states to either set up these exchanges on their own, or have the federal government run them instead. Because of that, there are at least 18 different models for how the exchanges work. 

Beyond that is the sheer complexity of what the exchanges are trying to do, something far greater than what, say, an Orbitz or Travelocity tries to pull off in travel. In the District of Columbia, for instance, people can choose from among nearly 300 health plans, Ku said. 

Another tricky part is the necessary coordination between the exchanges on the one hand and a large number of federal and state databases on the other. Under the law, for example, the exchanges can’t serve undocumented immigrants. That means everyone who applies for insurance coverage must have verification that they are citizens or legal immigrants, which means checking with the Social Security Administration and the Homeland Security Department. 

Biggest Issue: Time 

Perhaps the biggest issue, though, is the compressed timeframe in which the exchanges were set up. In D.C., Ku said, the contract to set up the system was established in March, and the exchange was built between March and October. 

That’s extremely fast, Ku said, especially considering that during the testing of the District’s exchange system, it also had to adopt roughly 1,000 federal standards.

The District’s system was tested to ensure it was operating the way it was supposed to, Ku said. 

Alas, the speed with which exchanges were built, combined with changes that the government instituted during their construction, meant that many exchanges simply weren’t ready for prime time on Oct. 1. Exchanges faced high standards and thorough testing, which meant that if they passed, they were essentially good to go.

“On the other hand, a lot of places couldn’t get there,” Ku said. “There were delays in some places because during testing, there were problems that cropped up, and they had to debug the programs.” 

Not to Mention the Volume 

Last but not least was the sheer volume of web traffic that hit the exchanges when they opened. Right out of the gate, the federal government’s exchange site saw 8 million people check in to see what it had to offer. Expectations had been for 7 million people to sign up over the course of a number of months. 

“Heady demand caused capacity problems,” Ku said. 

The good news is that the exchanges are facing up to their technical problems. 

“It was a compressed time schedule. There are always glitches any time any new big thing begins,” Ku said. “Hopefully now that we’ve opened up most of these web sites, there can be an opportunity to build capacity, debug programs and get the programs read on time. The initial hope is those software problems can be resolved.” 

Let’s hope so. And let’s hope that strong development testing is a big part of the effort. Too much is riding on this to use anything less.