AMCHP 2005 Conference Linking Birth and Medicaid Data for MCH Policy and Program Development Presentation to Association of Maternal and Child Health Programs Mary Alice Lee, Ph.D. Amanda Learned, B.A.
AMCHP 2005 Conference Overview Describe methods used to link birth certificate data with Medicaid data –Development and evaluation of data matching algorithm –Interagency collaboration and data sharing Illustrate ways in which linked data can be used to inform MCH policy and program development
AMCHP 2005 Conference Connecticut’s HUSKY Program HUSKY A=Medicaid managed care program –Children <19 in families with income <185% FPL –Parents/caretaker relatives in families with income <100% FPL –Pregnant women <185% FPL Covers 215,000 children and 91,000 adults (01/2005)
AMCHP 2005 Conference Independent Performance Monitoring in HUSKY A Supported by Connecticut General Assembly State-funded, with additional funds from private foundations for special studies Value-added: –Accountability for federal and state dollars –Accountability for program change –Information and data sharing for policy development, surveillance and health planning –Data quality assurance
AMCHP 2005 Conference Limitations of Existing Data on Births to Mothers in Medicaid CT birth certificate does not have field for payer HUSKY A health plan reports on prenatal care and birth outcomes are not always complete HUSKY A health plans report on adequacy of prenatal care once the mother is enrolled Encounter records for mother and baby cannot be readily linked
AMCHP 2005 Conference Interagency Collaboration: The Bridge Worked with staff from state Medicaid and public health agencies to develop an algorithm for matching CT birth data and HUSKY A enrollment data Designed study and obtained approval from CT Department of Public Health Human Investigations Committee for release of birth data Worked with agency staff on design of interagency data-sharing agreement
AMCHP 2005 Conference Data Elements for Linkage Mother’s Social Security number (SSN) Mother’s exact first + last names (NAME) Mother’s exact date of birth (DOB) Baby’s date of birth (DOB) (to verify that mother was enrolled when the baby was born)
AMCHP 2005 Conference STEP 1 Match mother’s SSN and Verify mother’s enrollment on baby’s DOB 9,173 records matched BIRTH DATA 43,075 births HUSKY A ENROLLMENT DATA 293,548 ever enrolled STEP 2 Match mother’s NAME and DOB and Verify mother’s enrollment on baby’s DOB 457 records matched 9,630 RECORDS MATCHED
AMCHP 2005 Conference Evaluation of Match (1) For records with match on SSN: –Did NAME match? –Did DOB match? For records with match on NAME and DOB: –Was SSN missing or different?
AMCHP 2005 Conference STEP 1 Match mother’s SSN and Verify mother’s enrollment on baby’s DOB 9,173 records matched BIRTH DATA 43,075 births HUSKY A ENROLLMENT DATA 293,548 ever enrolled 3% did not match on mother’s exact DOB 5% did not match on exact mother’s first NAME 10% did not match on exact mother’s last NAME <1% did not match on DOB, first or last NAME STEP 2 Match mother’s Exact First Name and Exact Last Name and Verify mother’s enrollment on baby’s DOB 457 records matched - 14% missing SSN on birth certificate - 86% with SSN that did not match 9,630 RECORDS MATCHED
AMCHP 2005 Conference Evaluation of Match (2) Two HUSKY A health plans provided mother’s SSN: –97% of records found in linked dataset Two HUSKY A health plans provided mother’s name and maternal date of birth: –85% of records found in linked dataset
AMCHP 2005 Conference Results Births to mothers in HUSKY A Births to other mothers US births 2000 Number (%)9,630 (22.4%)33,445 (77.6%)4,058,814 Births to teens23.2%3.3%11.8% LBWT9.6%6.8%7.6% Preterm births13.2%11.1%11.6% First trimester prenatal care 76.0%88.2%83.2%
AMCHP 2005 Conference Extent of HUSKY A Coverage By maternal residence: –60% of births to Hartford mothers –52% of births to New Haven mothers –44% of births to Bridgeport mothers –Nearly half births in five other towns By race/ethnicity: –48% of births to African American mothers –49% of births to Hispanic mothers
AMCHP 2005 Conference How Data Have Been Used to Inform Policy and Program Development To show how publicly funded health program contributes to maternal and infant health in CT To identify health risks among low income women that can be addressed through Medicaid and public health programs To identify opportunities for increasing access to and coordination between programs
AMCHP 2005 Conference Medicaid Coverage for Smoking Cessation Bill raised in state legislature that would have made treatment for tobacco dependence a covered benefit
AMCHP 2005 Conference Subspecialty Care for VLBWT Babies Question raised about access to appropriate level of care for high risk mothers and newborns in Medicaid managed care
AMCHP 2005 Conference Impact of Premiums on Pregnant Women in Medicaid In 2003, CT proposed charging premiums for pregnant women in Medicaid in families with income >50% federal poverty level Estimated 2,000 low-income pregnant women would fail to enroll or lose coverage Repealed in 2004
AMCHP 2005 Conference Effect of Timely Enrollment on Prenatal Care and Birth Outcomes Question raised by HUSKY A health plans about when pregnant women enroll
AMCHP 2005 Conference Future Opportunities for Study Linked birth-Medicaid data and HUSKY A encounter data: –Infant health and health care –Dental care in pregnancy Linked birth-Medicaid data and enrollment data from other publicly funded programs –WIC participation –Early intervention programs
AMCHP 2005 Conference Challenges Continuing state funding for independent performance monitoring Enhancing resources for data management and analysis of large datasets Complying with HIPAA restrictions on disclosure of protected health information Evaluating effects of program change Improving data quality
AMCHP 2005 Conference For more information Mary Alice Lee, Ph.D. Senior Policy Fellow Connecticut Voices for Children 60 Gillett Street Suite 204 Hartford, Connecticut / fax Amanda Learned, B.A. Systems Analysis Manager MAXIMUS, Inc Sunset Hills Road Reston, Virginia com