TennCare officials find collecting data on quality, cost and accessibility to be much harder than they thought
TennCare Data
Nearly three years after its controversial launch, TennCare has learned the hard way that systematic collection of good, reliable data in a managed care environment is no easy feat.
That’s a lesson other states invariably will be pressed to learn, as Medicaid managed care programs continue to proliferate and expand. The proof of all those programs lies in the data — the raw material for evaluating the impact of managed care on the quality, cost and accessibility of services for Medicaid recipients.
"Anywhere you go in Medicaid, those are going to be the three themes of [data] measurement," notes Jo Anne Lutz, director of marketing and communications for The Codman Research Group, Inc., whose managed care information system TennCare has contracted to use. "The overarching issue," she continues, "is: How do we guarantee that the change in the reimbursement environment is not adversely impacting access to care and quality of care?"
Theresa Clarke, assistant commissioner for Tennessee’s Department of Finance and Administration, which oversees TennCare, admits that she thought data collection would be a relatively simple matter. After all, she says, the state had years of experience collecting fee-for-service claims data as a Medicaid agency. Under TennCare, it is collecting essentially the same data.
The trick, however, is that these data are now being culled from an "encounter-based" environment, rather than a traditional fee-for-service environment. In a fee-for-service system, providers have a very strong incentive for recording and submitting proper claims information: Without it, they don’t get paid.
But that incentive disappears under managed care. It then becomes the plan’s responsibility to ensure that each of its providers is recording and reporting all of the required data for all their encounters in the proper fashion, using agreed-upon definitions that are designed to apply across all participating health plans. "Even if all those things are being done, you then have to figure out: Do I have all the data?’" explains an official from the Health Care Financing Administration. "It’s hard to tell whether something is missing."
TennCare had other problems as well. Many of the 12 managed care organizations under contract to provide services to the state’s 1.2 million Medicaid recipients are new plans, formed for the purpose of contracting with TennCare. "They’re just trying to get the basics," says Bart Perkey, director of assessment and evaluation at the Metro Health Department for Nashville and Davidson County.
Plans find reporting difficult
And even the more established plans had never been subjected to the kinds of reporting requirements imposed by TennCare, Ms. Clarke notes. "The MCOs have found it incredibly difficult," she says. "We really got off to a slow start with the collection of the data."
Plans under contract with TennCare are required to report their encounter data in an electronic format, using nationally accepted claims standards, such as the HCFA 15 form for professional services and the UB-92 form for hospital services. "We collect the same information we required providers to submit to us when we [operated] as a Medicaid agency," Ms. Clarke explains.
Plans have six months in which to file an encounter claim. After a claim has been adjudicated, the plan has until the 15th day of the following month to submit the relevant encounter information, Ms. Clarke says.
Those data were not forthcoming initially, she notes. But the state had carved itself a powerful — and effective — club. Every month, the state withholds 10% of each plan’s capitated payment. That withhold is paid the following month, provided the plan meets certain obligations, including proper data filing. At first, "some plans accumulated large amounts of withholds," Ms. Clarke recalls. Now, she says, most data are submitted on a timely basis.
State officials also had to make sure the plans understood TennCare’s data definitions, Ms. Clarke says. During the past six months, the plans have been asked to submit a raft of self-reported data for a number of utilization indicators. "We then calculated the same information from the encounter data, then laid them side by side to try to understand the variances," she explains. Those cross-checks have helped clarify data definitions and highlighted data processing glitches. Now the plans are resubmitting many of those data, Ms. Clarke says.
Ms. Clarke believes the pay-off is in sight. TennCare, she says, has been rebuilding its entire database, which, by the end of this month, will have accurate information on all encounters through March 1996. Data from the program’s first 18 months alone amount to 55 million encounter records, Ms. Clarke says. "It really is overwhelming when you think about analyzing it."
HCFA says it’s awaiting updated 1994 data from TennCare, although a preliminary look at early data indicated they were of "pretty good" quality. HCFA’s evaluation contractor for the TennCare program, Mathematica, will run the data through a validity check to ascertain their accuracy and comprehensiveness.
The information system designed by Codman will permit TennCare to create risk-adjusted utilization profiles across all types of services and settings, conduct small-area analysis by county, examine population-based health status and compare managed care providers, Ms. Lutz says. The database contains two years of fee-for-service data from prior to TennCare’s implementation and will soon contain two years of post-implementation data, she adds, so that TennCare officials will be able to assess the impact of managed care on services.
But plenty of other observers are interested in what those data have to say — and some are worried.
Kathy Wood-Dobbins, executive director of the Tennessee Primary Care Association, which represents the state’s community health centers, notes that, after nearly three years of operation, TennCare has yet to undergo any meaningful scrutiny by either HCFA or the state legislature.
"We don’t understand what’s happening with clinical care" under TennCare, Ms. Wood-Dobbins says. "We’ve received assurances, but no evidence."
She notes that since TennCare has been moved from the state Health Department to the Department of Finance and Administration, it’s unclear how the state will integrate Health Department data with TennCare data. That, she says, is of particular concern to public health advocates.
But Ms. Clarke says those data will be integrated and analyzed. For example, she says, TennCare officials have been working with the Health Department’s Center for Health Statistics to match TennCare data with other state data to identify pregnancy outcomes. "The encounter record doesn’t contain a lot of outcome information, but it does provide information that we can link with other data sources," she explains.
Ultimately, she says, TennCare officials want to be able to compare contracting plans on performance indicators such as rates of low birthweight, prenatal care visits and mammograms. "We absolutely want to do that," Ms. Clarke says.
It will, however, take more time. "We are in the position where we have a lot of data that we believe are pretty accurate, and we hope that we’ve corrected most of the problems," Ms. Clarke says. But before undertaking any kind of "report card" project, program officials need to establish a feedback and review process with the plans, she says. "We’ve got one more level of testing before we could publish report card data," says Ms. Clarke, adding that she hopes to produce such a report next year.
In the end, she believes the data will show "that our enrollees are receiving quality services, and that there’s every opportunity for our enrollees to receive better care than they were receiving under fee-for-service."
This article was written for SHW by Mary Darby, a health care writer in Washington, DC.
TennCare officials find collecting data on quality, cost and accessibility to be much harder than they thought
Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles
both enjoyable and insightful. For information on new subscriptions, product
trials, alternative billing arrangements or group and site discounts please call
800-688-2421. We look forward to having you as a long-term member of the Relias
Media community.