Contributed by Stephanie Kennedy
In attempting to better understand patterns and outcomes within the health care system, one of the most powerful tools researchers and policy-makers have is the ability to link data. Take, for example, the case of Vital Statistics and Medicaid data. By itself, Medicaid data cannot deliver information on key indicators of child and maternal health. Similarly, Vital Statistics (birth certificate) data does not always reflect the most accurate information on payer status. Examining the relationship between the two sets can be instrumental in identifying interventions to improve the overall health outcomes of mothers and their infants and putting them into practice. Specifically, data linkage is important because:
- Population health needs can be better understood using linked data. Medicaid and the Children’s Health Insurance Program (CHIP) –programs that provide health care for poor and low-income children and adults – cover almost fifty percent of all births in the United States. Linking data can help target specific sub-populations and identify certain health indicators otherwise unavailable from claims data or birth certificates alone.
- Linking data sets can improve the quality of birth certificate reporting. Examining the relationship between the two data sets positions states to better evaluate the accuracy of information about payment source included on birth certificates. Such findings can inform efforts to improve the quality of birth certificate reporting.
- Linked data will allow for measurement of various maternal and child health indicators. This includes birth outcomes, interconception care, pregnancy spacing, smoking cessation, and healthcare utilization. Linked data will also allow these outcomes to be tracked over time in order to monitor trends and even identify potential gaps in health program services.
In September 2015, AcademyHealth concluded Phase II of the Vital Statistics and Medicaid Data Linkage Training Project. This year-long interactive training project conducted with our partners from the Centers for Medicare and Medicaid Services (CMS) and the Division of Reproductive Health (DRH) within the Centers for Disease Control and Prevention (CDC) assisted states in linking and using Medicaid and Vital Statistics (birth certificate) data.
In particular, the project worked to improve participating states’ capacity to assess and report two measures in the CMS Core Set of Children’s Health Care Quality Measures for Medicaid and CHIP: (1) a measure of the rate of low birth weight, and (2) a measure of the rate of C-section delivery. To accomplish this, AcademyHealth developed a series of web trainings, facilitated expert technical assistance and hosted an in-depth and interactive training for participants. Dr. Craig Mason of the University of Maine and Dr. Russell Kirby of the University of South Florida along with state advisers and technical assistance from Mathematica Policy Research led states in addressing some of the most pressing challenges within data linkage today.
Prior to this training, the majority of Phase II states had limited knowledge and experience working with Medicaid claims and/or eligibility data or had never worked closely with linkage software SAS and FRIL. In addition to key technical skills, the training also afforded them the much-needed and rare opportunity to set aside designated time with their team to specifically devote to linking. One participant noted, “The best part [of the training] was sitting and talking with my collaborator…about the project. We are much farther ahead than we would have been if I had been doing it on my own.”
The process of linking data can sometimes take years to complete – accessing and sorting the data, “cleaning” the data sets, keeping detailed records and managing team turnover, not to mention numerous bureaucratic and policy hurdles all stand in the way. However, with time, patience, and persistence it can be achieved, resulting in incredibly valuable insights into larger public health questions and proving that, in the case of Vital Statistics and Medicaid data, two sets are better than one.
More information on the Vital Statistics and Medicaid Data Linkages Training Project can be found here.
Stephanie Kennedy is a Research Assistant with the State Health Policy and Technical Assistance team. In this role, Stephanie supports various state health projects, primarily focusing on AcademyHealth’s State-University Partnership Learning Network (SUPLN), and the Vital Statistics and Medicaid Data Linkage Web Trainings with the Centers for Disease Control and Prevention and the Centers for Medicare & Medicaid Services.