U.S. flag

An official website of the United States government

DQ Atlas: Designing the Future of Medicaid & CHIP Data Quality

Although Medicaid and CHIP cover more than 75 million Americans, collecting high-quality and timely data for each state’s Medicaid and CHIP program is incredibly difficult. Not only do states have flexibility in the design and implementation of these programs, but collecting data and submitting it to the Centers for Medicare and Medicaid Services (CMS) in a standardized format is a demanding administrative and information technology task for both states and the federal government. Taken together, these conditions create an environment in which Medicaid and CHIP data users often struggle to interpret their results, as it can be complicated to determine if anomalous research findings are the result of state variation in program design, the presence of one or more data quality problems relating to a state’s data submission to CMS, or both.

The Data and Systems Group within the Center for Medicaid and CHIP Services at CMS, with the support of their vendors, has worked to solve this problem by designing and releasing the DQ Atlas. DQ Atlas is a free, interactive, web-based tool that shows data quality assessments for states across critical research areas with the goal of informing whether CMS’s Medicaid and CHIP data is usable, reliable, and accurate for analyzing a particular topic. DQ Atlas has the potential to revolutionize the ability of stakeholders to examine the usability of Medicaid and CHIP data and foster a broader community of researchers interested in understanding one of the nation’s largest and most critical insurance programs. 

DQ Atlas, T-MSIS, and the TAF

DQ Atlas builds upon CMS’s foundational Medicaid and CHIP dataset, the Transformed Medicaid Statistical Information System (T-MSIS), through which states submit data to CMS on a monthly basis. CMS and our state partners have been collaborating for several years to create and improve T-MSIS, which is now the most comprehensive national dataset on Medicaid and CHIP beneficiaries, providers, service utilization, managed care, and expenditures. However, given the size and complexity of T-MSIS, CMS created a research-optimized format called the T-MSIS Analytic Files (TAF) to facilitate analysis, research, and data-driven decision-making on key dimensions of the Medicaid and CHIP programs. TAF consist of beneficiary-level data and contains information on enrollment, demographics, service utilization, and payments in Medicaid and CHIP. TAF expand on their precursor, MAX, by containing hundreds of additional data elements and collecting information on a monthly, rather than quarterly, basis. TAF data is currently available for calendar years 2014-2018, as well as preliminary files for 2019.

TAF support internal and external research across a number of key areas, including cost and use, program evaluation, policy assessments, health care quality, and provider participation, both within and across states. Most recently, TAF have also become a critical data source for CMS’s Medicaid and CHIP Scorecard, which was designed to increase public transparency about the programs’ administration and outcomes. The Scorecard uses a multitude of data sources, including quality measures and more traditional administrative data. The use of these types of established measures serve as a foundation for the Scorecard, but can also be limiting in terms of the topics covered. Great potential exists for TAF to further support the Scorecard in highlighting which areas of the Medicaid and CHIP program need additional focus. Further, continuing to leverage TAF for the Scorecard will also lead to the use of more timely data that will ultimately drive improvements in areas such as state and federal alignment, beneficiary health outcomes, and program administration.

Although TAF represent the most robust Medicaid and CHIP data CMS has collected, there is considerable variation in data quality across states, time periods, file types, and topic areas. As previously mentioned, this is a function both of the complexity of program design and variation across states regarding their capacity to collect information and submit it to CMS in a standardized format. Because of the difficulty associated with collecting and reporting T-MSIS data, CMS and its state partners routinely collaborate to improve the quality of state T-MSIS submissions. CMS is working one-on-one with states to help them address data quality issues across 32 top priority items, and while states have made significant progress in addressing these items, data quality varies by state and by topic.

The DQ Atlas Advantage

Understanding the variation in data quality across states is essential in fostering data-driven Medicaid and CHIP programs. DQ Atlas is the first step in filling this gap by providing both detailed and summarized data quality information for all states across a wide range of topics and years. Further, it is the first time that CMS has released an interactive, web-based tool for assessing data quality. Unlike prior Medicaid and CHIP data quality products, which often included hundreds of pages of table output for each year and file type, DQ Atlas allows users to quickly search through over 80 analytically important topics grouped into nine thematic topic areas:

  • Enrollment Benchmarking
  • Enrollment Patterns Over Time
  • Beneficiary Information
  • Claim Files Completeness
  • Service Use Information
  • Provider Information
  • Non-Claim Records
  • Payments
  • Expenditure Benchmarking

In addition to searching across a single topic, DQ Atlas also allows users to compare across topics or examine multiple topics within a specific state.

Within DQ Atlas, information is grouped into DQ Assessments, Background and Methods documents, and DQ Snapshots. DQ Assessments are the charts, maps, and tables in Atlas that show state-level results for a particular data quality evaluation. For each state, the DQ Assessment assigns states a value ranging from unusable to low concern that indicates whether state’s TAF data is usable, reliable, and accurate for addressing a particular research question. Background and Methods documents explain more about the steps CMS has taken to determine state DQ Assessments for each topic. This includes information about the topic’s importance, the DQ Assessment criteria, the relevant TAF data elements, and related topics. Finally, users can also download DQ Snapshots, which summarize all of the DQ Assessment information for all states and topics. DQ Topic Snapshots summarize the information by topic area, while DQ State Snapshots summarize topic areas within a particular state and year.

The Future of DQ Atlas

Although DQ Atlas is an incredibly robust tool, CMS expects the tool to grow by adding additional topic areas and years of data in the near future. As Medicaid and CHIP agencies continue to make significant improvements in the quality of their T-MSIS submissions, DQ Atlas can continue to expand its focus on more specific and detailed research topics. Additionally, as TAFs for new content areas are released, such as the Provider File and the Managed Care Plan File, DQ Atlas can expand to cover these additional research areas.

Data transparency is a critical driver in achieving the triple aim of better care, better health, and smarter spending. DQ Atlas represents the most comprehensive and user-friendly data quality tool CMS has released for Medicaid and CHIP data to-date. By leveraging DQ Atlas to contextualize the data quality of Medicaid and CHIP data analyses, CMS continues to work towards transforming health care for all Medicaid and CHIP beneficiaries.

Messages from CMCS
Jessie Parker, Kim Proctor, Julie Boughn, Douglas Olson, Karen LLanos


Collections: Messages from CMCS