Data overload: Answers are here — now find them
Data overload: Answers are here — now find them
With Medicaid managed care programs expanding significantly over the last several years, most states receive huge amounts of data they have requested from the managed care organizations. Sometimes, it’s just too much.
State officials often don’t know how to use the data to better manage their programs.
It need not be that way, according to Princeton, NJ-based Mathematica Policy Research’s James Verdier, who conducted a survey of state agencies using data from the Center for Health Care Strategies.
As of June 2000, Mr. Verdier says, 42 state Medicaid agencies had implemented capitation contracting with at least one managed care organization (MCO) and were collecting large quantities of data in their Medicaid managed care programs.
Encounter data used included:
- MCOs;
- HEDIS (Health Plan Employer Data and Information Set) and other utilization and quality measures;
- CAHPS (Consumer Assessment of Health Plans) and other enrollee studies;
- External Quality Review Organization reports and focused clinical studies;
- reports on managed care organization enrollment and disenrollment;
- complaints and grievance records;
- reports on managed care organization financial performance;
- state budget tracking reports.
"Data are costly for MCOs to collect and report . . . and most states have limited resources to analyze the data and put them to effective use," Mr. Verdier says. "This puts a premium on using data strategically."
Mr. Verdier tells State Health Watch that his data survey came as a result of comments at a Center for Health Care Strategies purchasing institute that revealed state agency problems and frustrations with data collection and use. The agencies, he says, complained they were getting "tons of data from various places but were not able to use them to understand what was happening and to provide feedback effectively." And the managed care organizations were complaining that they have to submit a lot of data but nothing is done with them.
Work with what’s available
"Our suggestion is that states do the things that they can do most easily with the data that are readily available," Verdier says.
They should use "things such as enrollment and disenrollment trends; complaints and grievances; beneficiary surveys and financial reports, although they can be expensive; and key utilization measures such as hospital days, hospital admissions, emergency room visits, and drug data."
There are two approaches states are taking to be able to make more effective use of data: building up internal staff capabilities or relying on actuaries and other consultants.
States need to develop ways to use data to support their managed care strategies, which generally focus on some combination of four goals — improving MCO performance, demonstrating value to program funders, building provider support, and building consumer support and understanding.
Encounter data have potential
Although encounter data reported by MCOs represent potentially the richest source of data for Medicaid managed care, only a small number of states have been able to tap its potential, Mr. Verdier says.
Encounter data can be used to set MCO capitated rates and to monitor the volume and cost of Medicaid services down to the level of individual providers and beneficiaries. The data can be aggregated to show trends over time, comparisons among MCOs and providers, and patterns of care and beneficiary access, he says.
"To fulfill this promise, however, the data must be reasonably complete and accurate, states must have sufficient analytic resources, and there must be clearly defined uses and audiences for the data. If any of these ingredients is missing, encounter data tend to accumulate largely unused, MCOs increasingly question the utility of collecting and reporting them, and their promise remains unfulfilled," Mr. Verdier adds.
HEDIS and CAHPS are the most fully developed tools available to states to measure MCO performance, Mr. Verdier says. They permit standardized comparisons among MCOs and can show trends over time. Even though research and state experience suggest that consumers generally make only limited use of these measures in choosing among MCOs, they have become a valuable purchasing tool for Medicaid agencies and a significant stimulus for MCOs to improve their performance.
Because HEDIS and CAHPS are especially useful for public comparisons, Mr. Verdier says, states can use the measures to influence perceptions of key stakeholders including legislators, health care providers, advocacy groups, and the news media.
If states conclude that HEDIS and CAHPS are most useful as tools to monitor MCOs and for public accountability, rather than as direct aids to consumer choice, they can focus their publication and distribution efforts accordingly. Summary comparisons among MCOs can highlight dimensions likely to be of most interest to key stakeholders. Reports on plan performance could be made available at enrollment offices, on the Internet, and on request, rather than being mailed to every enrollee.
Looking for patterns and trends
Both states and MCOs track data on enrollee complaints and grievances for internal management purposes, looking for patterns and trends, and ensuring appropriate follow-up. Difficulties with standardization and interpretation limit usefulness of these reports for external reporting and comparisons among MCOs.
Enrollment and disenrollment trends are timely and easy to collect, but can be difficult to interpret unless the specific reasons for disenrollment are accurately obtained and recorded. Since disenrollment rates for reasons other than loss of Medicaid eligibility usually are very low, only major differences among MCOs or major changes in trends are likely to be significant. Verdier says it is important to be able to identify such differences and changes when they occur and to follow-up quickly to learn the underlying causes.
External Quality Review Organi-zations (EQRO) can be helpful because states receive enhanced federal matching payments for work the EQROs do and because they can provide a wide variety of managed care quality monitoring and reporting functions for states. Some of the functions include in-depth clinical studies, medical record reviews, encounter data validation and analysis, MCO readiness reviews, and beneficiary surveys.
Many states, according to Mr. Verdier’s survey, find MCO financial reports to be their most valuable monitoring tool, while other states make little use of them. The reports provide monthly or quarterly data on service utilization, revenues, and costs for each MCO, usually in standard forms prescribed by state insurance regulators or the Medicaid agency.
"The data are most useful when the Medicaid line of business is broken out separately," Verdier says. "States use financial reports primarily for internal monitoring of MCOs rather than for external reporting, although states in which MCOs have had solvency problems may find significant external interest in these reports. MCO financial reports can also be used to help set MCO capitated rates, especially when claims or encounter data are not sufficient for rate setting."
Maryland contracts with school
In Maryland, the state agency has had a long-standing contractual relationship with the Center for Health Program Management and Development at the University of Maryland-Baltimore County, according to Alice Burton, the state Medicaid agency’s director of planning and development administration. She tells State Health Watch that the university center has been the agency’s data warehouse as well as doing data analysis.
Ms. Burton says the agency has done a lot with encounter data in the past five years, using it in the rate-setting process for risk-adjusted patients. "We’ve developed a pretty complete set of encounter data," she says.
Rich source of information
"We’re using them to answer questions such as how access has changed. We see [the information] as a very rich data set that we can cut in different ways so it gives us a lot of flexibility," Ms. Burton explains.
She says establishing the encounter data set has been very difficult, and there only have been enough data to use in the last couple of years. "We now have physician data about 90% complete, and that’s enough to make good use of in our analyses. Actually using the data made it easier to step up collection of the data. When you use the data in rate-setting, there is a significant financial advantage to plans to have complete data." She says they have done better collecting outpatient data than inpatient data, and are still limited in the analyses they want to do.
While a report on the first five-years of Medicaid managed care in Maryland is being completed with help of the data, Burton says there are early indications that the state has made progress in expanding access to ambulatory care and well-child services. "We see that more people are generally getting into services, and we’ve made significant progress in rural areas."
Minnesota uses encounter data
In Minnesota, Medicaid director Mary Kennedy says her agency has collected six years worth of encounter data, with the last three years the most complete. "Each year they get better."
Minnesota, like Maryland, is using data for risk-adjustment, which accounts for 50% of the variation in rates paid to MCOs.
"Once we started using data in rate-setting, everyone was able to produce better data," Kennedy says. "We’ve tried to feed back data to the plans and to show contrasts with HEDIS and other public data."
Setting standards for collection
States wanting to make better use of data should set criteria for data collection that clearly specifies what elements are to be collected and reported, Ms. Kennedy says. They also should have an adequate and timely processing system, insist on data coming in as clean as possible, use data to provide feedback to plans, and use data in rate-setting.
Ms. Burton agrees, and adds that states should set priorities and concentrate on the high-priority areas initially. Thus, Maryland focused on data from the Centers for Medicare and Medicaid form 1500 first, and made a lot of progress in that particular area.
In summarizing recommendations to states, Verdier says:
- Medicaid agencies should approach their data collection and analysis efforts in a deliberate, strategic manner, paying careful attention to the audiences for their data.
- Agencies should involve MCOs in data planning and implementation efforts.
- They should perform a "data inventory" of the encounter data they require MCOs to submit.
- They should recognize that sound, timely information from MCOs on the financial performance of their Medicaid lines of business is critical to state program monitoring efforts and can compensate for gaps in encounter data.
[Contact Mr. Verdier at (202) 484-4520, Ms. Burton at (410) 767-5806, and Ms. Kennedy at (651) 282-9921.]
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.