Since the survey of CHIP and Medicaid enrollees and disenrollees is a major focus in this final report, additional information about its design is provided here. The survey was fielded in 2012 and included 12,100 CHIP enrollees in 10 states and 3,400 Medicaid enrollees in 3 of these states.18 The CHIP and Medicaid samples were drawn from three distinct enrollment domains:
- Recent enrollees were children who had been enrolled in the given program for 3 consecutive months, preceded by a gap in coverage of at least 2 months, at the time of sampling.
- Established enrollees were children who had been enrolled in the program for at least 12 consecutive months at the time of sampling.
- Recent disenrollees were children who were disenrolled from the program for at least 2 months, and who were previously enrolled for at least 3 months prior to their month of disenrollment, at the time of sampling.
The types of questions asked and the reference period for the questions varied depending on the sample domain for the child. As summarized in Table I.3, recent enrollees were asked about experiences before enrolling in CHIP or Medicaid, established enrollees were asked about experiences while enrolled, and recent disenrollees were asked about experiences since leaving CHIP or Medicaid. Questions about enrolling in the program were asked only of recent enrollees and questions about the reasons for leaving CHIP or Medicaid were asked only of recent disenrollees. A core set of demographic and health status questions were asked of all sample members.
Table I.3. Survey Content by Sampled Group
|SurveyContent and Timeframe for Sample Group|
|Content Area||Recent Enrollees||Established Enrollees||Recent Disenrollees|
|Enrollment||Reason for enrolling in CHIP or Medicaid Application experiences||Renewal experiences||Renewal experiences|
|Access, Use, Family Wellbeing||Access, Use, Family Wellbeing 12 months before enrollment||Access, Use, Family Wellbeing past 12 months (while enrolled)||Access, Use, Family Wellbeing since disenrollment|
|Insurance Coverage||Child’s coverage during 12 months prior to enrollment||Parent’s coverage and access to coverage at time of interview||Child’s coverage after disenrollment|
|Disenrollment||Reason for leaving CHIP or Medicaid|
The survey was conducted by telephone with an in-person follow-up component for households that could not be located or contacted by phone. For all sample members, the interview was conducted with the person most knowledgeable about the health care needs and services received for the sampled child. Typically, that person was either a parent or a legal guardian of the child. For in-person interviewing, the field locator provided the individual with a cell phone for completing the interview with Mathematica’s survey operations center, thus ensuring a consistent mode of interview (phone) for all sample members. Each sample member was weighted to reflect the population represented and also to correct for nonresponse.
The analyses combined bivariate and regression methods using appropriate sample weights and took into account the demographic and socioeconomic characteristics of sampled children and their families. The analysis of impacts of CHIP and Medicaid on access, service use and family wellbeing employed a comparison group design. The survey methods report provides additional details on the design and content of the survey as well as on the methods used in the analysis of the survey data (Smith et al. 2014)
Findings were integrated across the various evaluation components to address a number of overlapping research questions, as summarized in Table I.4.19 While this final report summarizes the main findings, individual reports provide further details on the methods employed and detailed findings for each evaluation component (see Appendix Exhibit B.1 for a full listing of evaluation reports).
Table I.4. Summary of Key Research Questions and Data Sources
|Research Topics/Questions||Surveys of CHIP (10 states) and Medicaid (3 states) Enrollees and Disenrollees||Case Studies in 10 states||Survey of CHIP Program Administrators (47 states)||National Survey Data (NSCH, CPS, ACS)||State Enrollment Data (10 states); other State Program Data (all states)|
|Program Context and Design Features
• Key CHIP design features
• How and why these features changed over time
• How design features influence key program outcomes
|Outreach, Application, Enrollment and Renewal
• Effective and ineffective outreach strategies
• Awareness of Medicaid and CHIP among low-income families
• Enrollee experiences with application and renewal
• Trends in program enrollment and factors influencing trends
|Retention and Disenrollment
• Trends in duration/retention, churning, and transitions between Medicaid and CHIP
• Why children disenroll, and whether they obtain other coverage after leaving
|Access, Utilization, Content of Care, and Family Wellbeing
• Experiences of CHIP enrollees seeking and obtaining care.
• How this compares with experiences before enrollment
• CHIP’s impact on the type of care received, content of care, and family well-being
|Relationship Between CHIP and Other Coverage
• How CHIP alters or factors into the movement of low-income children between public coverage, private coverage, and uninsurance
|Impact on Uninsured Children
• Coverage trends among CHIP’s target population
• Participation in public coverage among eligible low-income children
|Implications for Health Reform
• Influence of reforms on CHIP programs, to date and expected in future
Notes: CHIP=Children’s Health Insurance Program; CPS=Current Population Survey; ACS=American Community Survey.
18 For the CHIP survey, the target population included current and former enrollees in separate CHIP and Medicaid expansion CHIP components. For the Medicaid survey, the target population included only enrollees and disenrollees in the “traditional’ (Title XIX-funded) Medicaid program and not those in the Medicaid expansion CHIP component. Further details on the methods (including variable construction and sample sizes for different analyses) are part of a separate report on the survey (Smith et al. 2014). That document also includes the survey instrument and a detailed description of the sampling and data collection approach to the survey of CHIP and Medicaid enrollees and disenrollees.
19 Further details on the design are contained in the design report for the evaluation, Harrington et al. (2011).