Established Child Health Care Quality Measures--HEDIS(R)
HEDIS®: Health Plan Employer Data and Information Set
|Select to download Microsoft® Word® File (95 KB).|
- Child Measures Included
- Comparisons and Trends
- Benchmarking and Databases
- Service Delivery and Units of Analysis
- Length-of-Enrollment Requirements
- Data Issues
- Sample Sizes
- Resource and Burden Issues
- Development Process
- Criteria Used
- More Information and User Support
The Health Plan Employer Data and Information Set (HEDIS®) is a widely used measurement set focused on clinical services and utilization, initially designed for use by purchasers and managed care organizations (MCOs). HEDIS® is a registered trademark of the National Committee for Quality Assurance (NCQA), which maintains and updates it annually under the direction of a broadly representative Committee on Performance Measurement.
Online Resource: For more information on NCQA, go to: http://web.ncqa.org
Child Measures Included
The HEDIS® measurement set contains measures of:
- Preventive and well care.
- Care for selected chronic conditions.
- Use of services.
- Perceptions of care (incorporating the managed care version of CAHPS®).
- Descriptive measures.
HEDIS® 2003 was designed to collect data for measurement year 2002 for reporting in 2003. It contains 52 measures, including 23 measures designed for or applicable to children. The measures are categorized by NCQA-designated measure type as follows:
Type: Effectiveness of Care. Measures for which clinical standards of care exist.
- Childhood immunization status.
- Adolescent immunization status.
- Chlamydia screening for women.
- Use of appropriate medications for people with asthma.
Type: Access/Availability of Care. Measures related to timely provision of needed care.
- Children's access to primary care practitioners.
- Annual dental visit.
Type: Satisfaction with the Experience of Care. Measures of perceptions of care.
- The CAHPS® 3.0H Child Survey. This revised Consumer Assessment of Health Plans (CAHPS®) child instrument includes a screener for children with chronic conditions as well as several composite measures of health services related to such conditions. It also includes several HEDIS®-specific items related to MCOs. NCQA protocols for administering CAHPS® 3.0H must be followed if the results are intended for use in NCQA accreditation.
- Online Resource: For more information on the CAHPS® 3.0H Child Survey, go to: http://www.ahrq.gov/chtoolbx/measure2.htm#cahpsexpandedsurvey
Type: Use of Services: Measures of utilization, generally without standards of appropriate utilization levels. These include several measures related to older adolescents.
- Well-child visits in the first 15 months of life.
- Well-child visits in the third, fourth, fifth, and sixth years of life.
- Adolescent well-care visits.
- Frequency of selected procedures: Myringotomy, tonsillectomy.
- Inpatient utilization: General hospital/acute care.
- Ambulatory care.
- Inpatient utilization: Nonacute care.
- Mental health utilization: Inpatient discharges and average length of stay.
- Mental health utilization: Percentage of members receiving services.
- Chemical dependency utilization: Inpatient discharges and average length of stay.
- Chemical dependency utilization: Percentage of members receiving services.
- Outpatient drug utilization.
- Births and average length of stay, newborns.
- Discharges and average length of stay, maternity care.
- Cesarean section rate.
- Rate of vaginal birth after cesarean section.
Use of HEDIS® measures is widespread among employer-based MCOs, and the Center for Medicare & Medicaid Services (CMS) requires that Medicare+Choice MCOs use HEDIS® measures.
HEDIS® is also widely used by State Medicaid agencies to assess their contracted MCOs and, to a lesser extent, their primary-care case management programs (PCCMs). In its survey of State Medicaid programs, the National Academy for State Health Policy found that 34 States collected HEDIS® measures from their risk-based managed care programs in 2000. That same survey found that 19 States collected HEDIS measures for stand-alone risk-based State Children's Health Insurance Programs (SCHIPs). States using HEDIS® measures include New York and Washington State.
Online Resource: For some State examples, go to: http://www.ahrq.gov/chtoolbx/measure8.htm
Comparisons and Trends
When HEDIS® specifications are followed, HEDIS® results (other than those designated as "descriptive") are reliable and valid. The results can be used for statistically significant comparisons with HEDIS® results produced for other plans or programs and at other times (trending), provided that comparable populations are used. HEDIS® results based on samples can be generalized to the relevant universe.
Auditing of results to ensure consistent application of specifications and protocols helps ensure comparability of results. NCQA requires auditing of HEDIS® results submitted for MCO accreditation purposes and for public reporting through NCQA's Quality Compass.
For more information on HEDIS, go to: http://web.ncqa.org/tabid/59/Default.aspx
For more information on Quality Compass, go to: http://web.ncqa.org/tabid/177/Default.aspx
Some States or other users customize HEDIS® measurement specifications. Such customization can limit the comparability of measurement results. However, results produced using customized specifications can be credibly compared if other users have adopted the same changes. If the same customized specifications are used year after year, results can be compared over time to derive trends.
Benchmarking and Databases
Since 1998, the American Public Human Services Association (APHSA) has worked with NCQA to improve the availability and robustness of Medicaid HEDIS® benchmarks. The project received initial funding from The Commonwealth Fund and is currently being supported by CMS. APHSA has produced several reports that present national Medicaid averages on key measures.
The fifth-year (2002) report on this project includes data from 176 MCOs in 33 States, the District of Columbia, and the Commonwealth of Puerto Rico. Thirteen HEDIS® measures are covered, five of which are child specific.
Online Resource: Download the report in PDF format. Go to: http://www.nasmd.org/resources/publications.asp
The fourth-year (2001) report on this project contained data on 13 HEDIS® measures from 167 MCOs in 28 States and the Commonwealth of Puerto Rico. Results for the third-year (2000) report cover 12 HEDIS® measures from 167 MCOs in 27 States plus the Commonwealth of Puerto Rico
Additionally, NCQA makes health plan performance benchmarks (commercial, Medicaid, and Medicare) available in its annual State of Health Care Quality Report. For the commercial market and Medicaid, NCQA releases detailed plan-specific performance information through its Quality Compass. NCQA also makes its HEDIS®/CAHPS® database available to researchers.
For the State of Health Care Quality Report, go to: http://web.ncqa.org/Default.aspx?tabid=447
For Quality Compass, go to: http://web.ncqa.org/tabid/177/Default.aspx
Service Delivery and Units of Analysis
HEDIS® was designed to assess the performance of MCO service delivery systems. However, some States, including Colorado and Massachusetts, are using HEDIS® to assess services delivered to Medicaid beneficiaries enrolled in their fee-for-service (FFS) and PCCM programs as well as their MCOs.
Some HEDIS® results are published on a statewide basis as well as for individual plans and programs. With the assistance of a workgroup of Medicaid experts, NCQA produced a report discussing the methodological issues involved in using HEDIS® in FFS and PCCM settings.
Many HEDIS® measures have length-of-enrollment (or "continuous enrollment") requirements that identify those individuals whose treatment information can be included in calculations of measures assessing the performance of MCOs. To be included in the calculation of rates for HEDIS® measures involving services or treatments delivered in set time frames (e.g., preventive services, screenings, well-care visits), managed care plan members must be enrolled for a minimum of 12 months, with no more than one break of 45 days. For other measures, the required period of continuous enrollment varies.
Because HEDIS® measures are based in part on the premise that MCOs are accountable for providing defined services to enrolled members, the minimum period of enrollment is designed to give MCOs a reasonable opportunity to fulfill that responsibility prior to measurement.
There are no continuous length-of-enrollment requirements for HEDIS® use-of-services measures other than well-child and adolescent well-care visits, as no set periodicity requirements exist. CAHPS® 3.0H enrollment requirements are the same as for CAHPS® 3.0.
Data sources for the calculation of HEDIS® measures are administrative data, medical records, or, for CAHPS® 3.0H, survey data. Encounter data or claims data are potential HEDIS® data sources when applied to FFS or PCCM systems. Data completeness and accuracy are critical issues. Issues involving the data sources for HEDIS® include the following:
- MCO administrative databases may not be current or complete, especially for services covered under a capitation arrangement.
- Accessing medical records is costly and time consuming.
- Medical records may be incomplete and hard to decipher.
- Encounter data are often incomplete or inaccurate.
- Claims databases were designed to track the flow of money, not care, and may lack important data elements.
Improvement is always possible, however. States and others with experience can share lessons learned and examples of progress.
For additional information on data sources, go to: http://www.ahrq.gov/chtoolbx/develop.htm#data
For additional information on encounter or claims data, go to: http://www.ahrq.gov/chtoolbx/develop.htm#encounter
HEDIS® requires samples or groups large enough to produce statistically reliable results. HEDIS® measurement technical specifications are very detailed.
Online resource: For specifications, go to: http://web.ncqa.org/tabid/59/Default.aspx
Resource and Burden Issues
The production of performance measures is a data-driven activity. MCOs with little experience in producing HEDIS® results will need to devote noticeable resources to this in the first years, with senior staff oversight.
Application to FFS and PCCM programs requires particular attention in early years. In all instances, senior-level agency responsibility and sufficient staff resources are needed to ensure useful results.
The HEDIS® measurement development process involves numerous components:
- Review of the existing research literature and additional in depth research.
- A committee with expertise on the particular issue.
- NCQA staff support.
- Consensus on the best measurement approach according to set criteria of:
- Relevance to purchasers and consumers.
- Scientific soundness.
- Development of precise and detailed measurement specifications, including instructions on sampling.
- Statistical testing.
- Field testing by MCOs.
- Final approval by the broad-based Committee on Performance Measurement.
Criteria used in developing HEDIS® measures include:
- Is the measure relevant for consumers, purchasers, health plans?
- Does it measure prevalent conditions? Serious conditions?
- Does it assess activities with high cost? Does it encourage the use of cost-effective, clinically effective options?
- Are there actions that health plans can take to improve their performance? Is there potential for improvement?
- Does clinical evidence document links among interventions, clinical processes, and outcomes?
- Is the measure reliable—that is, does it produce the same results when repeated with the same population in the same setting?
- Is the measure valid—that is, does it make sense logically and clinically (face validity)? Does it correlate well with other measures of the same aspects of care (construct validity)? Does it capture meaningful aspects of this care (content validity)?
- Is the measure an accurate gauge of what is actually happening?
- If the measure is appreciably affected by variables beyond the health plan's control, is an appropriate case-mix or risk-adjustment strategy in place?
- Are there safeguards to ensure reasonable comparability of data sources?
- Are there clear operational definitions, data specifications, and data collection and reporting specifications?
- Is the burden imposed on the health plan justifiable in terms of improved outcomes?
- Is the data collection method in keeping with accepted standards of member confidentiality?
- Are the data available to the health plan during the required period?
- Can the measure be audited to prevent manipulation?
More Information and User Support
NCQA sells HEDIS® documents describing specifications and protocols for the administration of HEDIS® (including CAHPS® 3.0H, which has some unique protocols) and provides several levels of user support for a variety of charges. Go to: http://web.ncqa.org/tabid/59/Default.aspx