By all accounts, those problems continue to gum up the system. On Wednesday, the CMS, American's Health Insurance Plans and the Blue Cross and Blue Shield Association took the unusual step of releasing a joint statement addressing the issue.
“We are working together closely to resolve back-end issues between health plans and HealthCare.gov,” it said. “This is a very focused effort that is being driven by a team of experts from CMS, key outside contractors working closely with health plan representatives and overseen by CMS's general contractor, Optum/QSSI.”
But administration officials have repeatedly refused to provide any kind of metric for how pervasive the remaining problems are or how many erroneous forms have already been sent out to insurers. The continuing problems raise the specter that individuals who believe they have signed up for coverage could show up at the doctor's office after Jan. 1 and find out otherwise.
“There is a risk of chaos once people start to use services,” said Paul Ginsburg, president of the Center for Studying Health System Change. “What nobody knows is how quickly it can be fixed.”
With large numbers of people finally able to enroll, concerns will also turn to who is actually signing up for coverage and whether the risk pool is sufficiently balanced between sick and healthy individuals to provide a viable market for insurers.
The federal government hasn't released any demographic information for individuals that have signed up for coverage through HealthCare.gov. But some states have released initial data on who's actually purchasing coverage—and it shows that exchange enrollees are significantly older than the general populace.
In California, roughly a third of enrollees in October were 55 or older, even though that age group represents just 11% of the population. Likewise, in Minnesota, half of enrollees in October and November were over the age of 50. Just 23% were between ages 21 and 40.
The risk profile of the enrollees, Ginsburg said, could be a much a bigger problem for insurers than the data glitches. “Will the risk profile be somewhat in the range of what the insurers were projecting and what they set their rates on? Or will it be very different?”