NHLBI Workshop

Data Needs for Cardiovascular Events, Management, and Outcomes

Centers for Medicare & Medicaid Services (CMS) - Dr. Marcel Salive

Outline of Talk

  • Medicare Claims
    • The bills submitted by hospitals (Part A) and physicians (Part B) to CMS for reimbursement
  • Medicare Coverage Process
    • Ability to link to data collection
  • Quality Improvement
    • Expansion of public reporting
    • Implementation of pay-for-performance

Medicare is the national health insurance program for:

  • People age 65 or older
  • Some people under age 65 with disabilities
  • People with End-Stage Renal Disease
  • 34.9 million covered lives in 2004

Three General Categories of Available Data

Category 1: Medicare Eligibility & Enrollment Data
Category 2: Medicare Claims Data
Category 3: Medicaid Overview Data

The first category includes Medicare beneficiary eligibility and enrollment data. The second category of data includes the Medicare data associated with fee-for-service claims. The third category of data includes Medicaid eligibility, utilization, and demographics data.

(Category 1) The Medicare Eligibility and Enrollment category contains person-level entitlement information for Medicare beneficiaries. Each time a claim for services rendered is received for adjudication and payment, beneficiary entitlement status is verified using this information. The Enrollment Database (EDB) is the designated CMS repository of enrollment and entitlement data for persons who are or have ever been enrolled in Medicare.

(Category 2) Medicare claims data constitutes the second category. Processing claims for Medicare health insurance benefits is fundamental to the operation of the Medicare program. CMS ensures that payments are made for services that are medically appropriate, covered, and rendered to eligible beneficiaries by qualified providers. The detailed claims records provide a unique source of information on health care utilization and costs. From these records, analytic files are created to support program and policy development and evaluation, as well as health care analyses and research. The National Claims History (NCH) is the CMS designated repository for all claims and utilization data.

Data from both categories contributes to the beneficiary demographics information that is maintained in the Enrollment Database (EDB), including: name; temporary residence, & mailing addresses; FIPS state & county codes; SSA state & county codes; date of birth; sex; representative payee; and program service center.

(Category 3) The third category contains data that originates in the State Medicaid Claims Processing systems. The types of data found in this category include eligibility data, claims data, other encounter and utilization information, and provider data.

Advantages of Using Medicare Claims for Surveillance

  • Routinely collected
  • Virtually the entire population of patients and providers
  • Large numbers
  • Tied to reimbursement, so is complete
  • Fraud if not accurate
  • Unique identifiers allow episodes of care to be linked for complete follow-up

Part A (Hospital) Claims

  • Unique patient and hospital ID
  • Dates of admission and discharge
  • Admitting diagnosis and acuity
  • Procedures performed (ICD-9-CM)
  • Medical diagnoses (ICD-9-CM)
  • Discharge status
  • Discharge destination

Part B (Physician) Claims

  • Unique patient and physician ID
  • Surgical and diagnostic procedures (CPT)
  • Date of service
  • Diagnosis for which service performed

Medicare Coverage

  • Section 1862(a)(1)(A) of the Social Security Act
  • Coverage and payment limited to items and services
    • Found "reasonable and necessary"
    • For treatment of illness or injury...

Steps to Medicare Coverage Determination and Payment

  • Outside of CMS:
    • Congress determines benefit categories
    • FDA approves drugs/devices for market
  • Within CMS:
    • Coverage
    • Coding
    • Payment

What standards are used in an NCD

  • Evidence of improved health outcomes
  • Appropriate for Medicare population
  • Could be replicated in provider community

Medicare ICD expanded coverage

  • Effective 1/05
  • Based on the results of SCD-HeFT trial
  • Linked to submission of data to national ICD database
  • Can answer residual questions regarding safety & effectiveness in certain groups of patients & providers
  • Initial hypotheses included in Decision Memo

ICD Implant Data Form

  • One page printed form
  • Data elements include:
    • Demographics
    • Patient history & clinical characteristics
    • Medications
    • Provider information
    • Clinical indications
    • Complications

CMS Vision of Quality: "The right care for every person every time," where the "right care" corresponds to the 6 Institute of Medicine aims

  • Safety
  • Effectiveness
  • Efficiency
  • Patient-centeredness
  • Timeliness
  • Equity

CMS Strategies For Promoting The Quality Council Vision

  • Standards setting, regulation, enforcement
  • Public reporting
  • Payment policy including pay-for-performance
  • Technical assistance

Upcoming work to Support This Vision?

  • Launch the third phase of the QIO Program - promote the Quality Council vision through infrastructure for public reporting and p4p and the provision of assistance
  • Lay foundation for evidence-based improvement in drug safety/quality
  • Lay foundation for evidence-based improvement in efficiency of resource use

Quality Measures

  • Measures do not stay constant due to changes in care
  • Measure refinement and development of new measures is needed
  • Abstraction, electronic, and survey tools must be modified to support data collection
  • CMS will develop and implement data validation processes
  • Maintenance and upgrades to IT infrastructure necessary to support data collection, validation, and reporting.

Assistance Tools and Methodology

  • Tools/methodologies needed to help providers seeking improvement on new performance measures
  • CMS needs to refine the tools and methods we are implementing
    • New tools to support work on new measures
    • Refined tools for work on measures that have not shown substantial improvement

Single Measure Set, Multiple Uses...

  • CMS using all or subset of measures for:
    • Doctors office
    • Medicare care management project
    • Voluntary coordinated care improvement pilot
    • Physician Group Practice demonstration
  • External interest in common measure set...
    • Physician specialty boards
    • Health plans
    • Purchasers
    • Consumers

CMS Measurement Framework

  • Build from previous work
  • Use existing, accepted concepts or measures
    • HEDIS measures
  • Align with other measures
    • JCAHO hospital measures
  • Public comments/forums

CMS Basic Requirements

  • Must be scientific and clinical sound
    • Evidenced based in guidelines
  • Must be transparent (reproducible)
    • Technical specifications and other technical documents available to public (www.cms.hhs.gov/quality) at no charge
  • Should not add burden to provider
  • Should use existing data source when possible

Types of Measures

  • Process: Blood pressure checked? Flu shot given?
  • Outcomes: Mortality, morbidity, HgbA1c control
  • Structure: Staffing levels, IT infiltration

Measure Alignment*

  • AMA, NCQA and CMS worked to create a single set of measures for coronary artery disease (CAD), congestive heart failure (CHF), hypertension (HTN), diabetes mellitus (DM), and prevention
  • JCAHO & CMS have aligned the reported hospital measures
    * Alignment at the micro-specification level

Measure Development

  • Identify existing relevant measures
  • If none found or revisions required....
    • Extensive input from expert clinicians
    • Draft technical specifications & training manuals
    • Build data collection tool if none exists
    • Support warehouse construction, record layouts
    • Multiple rounds of testing and refinement
    • Validity testing (do I get the information I expected?)
    • Reliability testing (does someone else get the same answer?)
    • Send for endorsement

Risk Adjustment: Address risk adjustment and other data adjustment needs

  • Inclusions and exclusions into the numerators and denominators
  • Variety of other techniques available to adjust the data
  • Result: measures calculated so that one has an 'apples to apples' comparison
  • May not always need to be risk adjusted

Endorsement of Measures: National Quality Forum

  • Private non-profit entity to create standards for health care quality
  • Working in partnership with many, including CMS, to develop consensus around what measures are ready to be called standards
  • Final step in a long development process

Contact Information

Websites: www.cms.hhs.gov/coverage
1-800-MEDICARE
www.medicare.gov
Marcel Salive, MD, MPH
410/786-0297
Marcel.Salive@cms.hhs.gov

Back to Workshop Agenda

Skip footer links and go to content