Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 65
Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary 8 Monitoring Children’s Health Insurance Coverage Under CHIPRA Using Federal Surveys Genevieve Kenney and Victoria Lynch The Urban Institute The Children’s Health Insurance Program (CHIP) was reauthorized for an additional 4.5 years in February 2009 through the Children’s Health Insurance Program Reauthorization Act (CHIPRA, P.L. 111-3). CHIPRA contained a number of provisions designed to expand eligibility for public coverage among children and to increase take-up of coverage among uninsured children who were already eligible for Medicaid and CHIP. This paper assesses possible data sources for monitoring the impacts of CHIPRA on children’s health insurance coverage. The following section provides background on CHIP, CHIPRA, and other recent federal policy changes that also have important implications for Medicaid and CHIP coverage for children and their parents. Subsequent sections discuss key research questions and data needs and describe the strengths and limitations of different data sources that are available at the federal level. The closing section suggests ways to improve existing federal surveys so that they provide more useful information for evaluating CHIPRA and related policy changes. THE CHILDREN’S HEALTH INSURANCE PROGRAM CHIP was created in 1997 in an effort to close coverage gaps facing low-income families who did not have access to affordable private coverage for their children but had incomes that were too high to qualify for Medicaid. CHIP was funded as a block grant to states but with higher federal matching rates than states typically received under Medicaid. States
OCR for page 66
Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary had flexibility over theIR design of CHIP, including eligibility thresholds, outreach, retention, enrollment policies, and, within parameters set down under the statute, benefits and cost sharing. All states chose to expand coverage for children through CHIP and implemented policies aimed at simplifying enrollment processes, many of which were also adopted under Medicaid (Kaye et al., 2006). Since the late 1990s, when these policy changes were adopted, uninsured rates have fallen among children, both those made newly eligible for public coverage under CHIP and those who were already eligible for Medicaid (Dubay and Kenney, 2009; Hudson and Selden, 2007). Importantly, gains in health insurance coverage appear to have translated into improvements in access to care and increased preventive care receipt among children (Davidoff et al., 2005; Kenney and Change, 2004; Kenney and Yee, 2007). Despite this progress, at the time when CHIPRA was passed, research indicated that millions of children were uninsured despite being eligible for Medicaid or CHIP and that many children enrolled in public coverage were not receiving recommended levels of care (Dubay et al., 2007). Moreover, uninsured rates among low-income children continued to vary widely across states (DeNavas et al., 2008). In an effort to address these gaps, CHIPRA provided states with new tools to address shortfalls in enrollment as well as access and quality. CHIPRA included new outreach and enrollment grants and bonus payments to states that adopted five of eight enrollment/retention strategies and that experienced Medicaid enrollment that exceeded targeted growth rates.1 States were also given new options to use Express Lane Eligibility options to facilitate eligibility determination and enrollment and for meeting documentation requirements (U.S. Department of Health and Human Services, 2010). CHIPRA allowed states to use federal dollars to cover legal immigrant children who had been in the United States less than 5 years (previously coverage for such children had to be funded exclusively with state funds). It also provided states with additional federal allotments for CHIP to cover the costs of enrolling more eligible children and of expanding eligibility (e.g., to higher income groups). In 1 These include (1) Adopting 12-month continuous eligibility for all children; (2) eliminating the asset test for children; (3) eliminating in-person interview requirements at application and renewal; (4) using joint applications and supplemental forms and the same application and renewal verification process for the two programs; (5) allowing for administrative or paperless verification at renewal through the use of pre-populated forms or ex parte determinations; (6) exercising the option to use presumptive eligibility when evaluating children’s eligibility for coverage; (7) exercising the new option in the law to use Express Lane; and (8) exercising the new options in the law in regard to premium assistance (Georgetown Center for Children and Families, 2009).
OCR for page 67
Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary addition, CHIPRA included a number of provisions designed to improve access to care and the quality of care for the children served by Medicaid and CHIP. In addition, both the American Recovery and Reinvestment Act (ARRA) of 2009 and the Patient Protection and Affordability Act (PPACA) of 2010 included provisions that could affect children’s coverage.2 ARRA included enhanced matching rates to states that maintain their Medicaid eligibility thresholds for children and adults, in an effort to induce them to continue supporting Medicaid coverage during the current recession. The enhanced matching rates, which are at least 6.2 percentage points higher than regular matching rates, were implemented on October 1, 2008, extended in August 2010, and are slated to continue with a phase-down period through June 2011. PPACA contained a number of important policy changes that could affect both Medicaid and CHIP coverage for children. It legislated comprehensive health reform, including an expansion of Medicaid to adults and children up to 133 percent of the federal poverty level (FPL), a maintenance of effort requirement through 2019 on state Medicaid and CHIP coverage for children, the provision of new subsidies for the coverage of families with incomes up to 400 percent of the FPL, the creation of health insurance exchanges, and coverage mandates for both individuals and employers. PPACA also provided 2 additional years of federal funding for CHIP, beyond what was in CHIPRA, through 2015. It is not clear how long states will be able to continue CHIP, given that new federal funds for CHIP are allocated only through 2015 and that the federal matching rates under CHIP rise by as much as 23 percentage points at that time, which means that a given allocation will be spent more quickly. This paper focuses on the monitoring of children’s coverage under CHIPRA. However, as we approach 2014, the policy questions will begin to focus on how the provisions of PPACA affect children’s health insurance coverage. Although the questions of interest will expand to reflect the broader, comprehensive reforms adopted under PPACA relative to CHIPRA, the data issues discussed here will still be relevant for assessing PPACA. MONITORING CHILDREN’S COVERAGE UNDER CHIPRA As indicated above, CHIPRA gave states new tools, incentives, and resources to expand health insurance coverage to low-income children. Just 1 year after CHIPRA was passed, 15 states had expanded their eligibility thresholds under Medicaid/CHIP, 19 states had removed the ban on 2 Patient Protection and Affordable Care Act of 2010 (P.L. 111-148).
OCR for page 68
Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary covering legal immigrant children who had been in the country less than 5 years, and 18 states had proposed improvements to their enrollment and retention processes or received approval to use the new Express Lane Eligibility Option. In 2009, 9 states qualified for bonus payments, and $40 million in outreach grants was allocated to 42 states (another $10 million in tribal outreach grants was funded in February 2010). These policy changes are expected to produce changes in the coverage distribution of children made newly eligible for coverage (i.e., immigrant children in states lifting the 5-year ban and children in income groups made newly eligible for coverage in states that expanded eligibility) and children who were already eligible for coverage under Medicaid and CHIP (in states that streamlined eligibility and retention processes, etc.). In order to monitor children’s coverage under CHIPRA, the following types of questions regarding children’s health insurance coverage and participation in Medicaid and CHIP need to be addressed: Did uninsured rates fall among children following enactment of CHIPRA? If so, by how much? Did uninsured rates fall more among some groups of children (defined by race/ethnicity, income, age, health status, etc.) than among others? If so, by how much? Did uninsured rates fall among Medicaid-eligible children? CHIP-eligible children? Among groups of children made newly eligible for coverage (e.g., immigrant children in the country for less than 5 years, children in income groups targeted by eligibility expansions in particular states, etc.)? Did uninsured rates fall more in some states than in others? Did Medicaid/CHIP participation rates increase? Were increases greater for some groups of children? How much do Medicaid/CHIP participation rates vary across states? Do differences in participation rates across states narrow over time? Did Medicaid/CHIP participation rates change over time? Did they increase more in some states than in others? How much do Medicaid participation rates differ from CHIP rates? Have those differences narrowed over time? Did rates of public coverage increase over time? What was happening to rates of private coverage over the same time period? How did rates of public and private coverage change for different groups of children (defined by race/ethnicity, income, age, health status, etc.)? How did rates of public and private coverage change across states?
OCR for page 69
Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary To what extent are the observed changes in uninsurance, public coverage, and private coverage among children attributable to CHIPRA? To what extent can the observed coverage changes be attributed to individual provisions in CHIPRA? To what extent can the observed coverage and participation changes be attributed to gains in particular states? To what extent can the observed coverage and participation changes be attributable to specific policy changes related to CHIPRA that were adopted? To address these questions, valid state and national survey estimates of insurance coverage are needed for the period prior to CHIPRA (to establish a baseline) and for several years following the implementation of CHIPRA-related policies but before the major provisions of PPACA are enacted. These estimates can be derived only from survey data because other sources, such as administrative records for Medicaid and CHIP, do not include information about children who are not enrolled. Establishing a pre-CHIPRA baseline is somewhat difficult because, although CHIPRA was enacted in February 2009, a similar reauthorization bill had been voted on several times in 2007; even though it did not ultimately become law until over a year later, states may have begun making eligibility and enrollment changes to their CHIP and Medicaid programs for children in 2007 and 2008 in anticipation that the bill would ultimately pass. Valid estimates are needed on how uninsured rates and rates of different types of coverage (e.g., employer-sponsored insurance [ESI], private nongroup coverage, Medicaid/CHIP coverage, other coverage) are changing for children age 18 and under. Because of the critical role that states play in designing and implementing their Medicaid and CHIP, it is essential to have precise annual estimates of the distribution of children’s coverage in each state. It is also critical that the survey data that are used to derive valid coverage estimates permit the identification of children who are eligible for Medicaid and CHIP to assess how coverage and participation rates are changing among children who are targeted by those two programs. Without such information, it is not possible to track how well Medicaid and CHIP are doing at reaching eligible children or to assess whether enrollment, retention, and related outreach efforts are increasing participation. Simulating eligibility for Medicaid/CHIP requires information on the child’s health insurance unit (such as family size, income, immigration status, etc.) that is used to determine eligibility for Medicaid and CHIP in each state (Dubay and Cook 2009; Kenney et al., 2010b). Finally, assessing the impacts of CHIPRA and related policy changes requires establishing a
OCR for page 70
Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary counterfactual for what would have happened in the absence of a particular policy change or set of policy changes. FEDERAL SURVEYS FOR MONITORING CHILDREN’S COVERAGE Four federal household surveys are available to monitor health insurance coverage on an annual basis: the Current Population Survey (CPS), the National Health Interview Survey (NHIS), the American Community Survey (ACS), and the Medical Expenditure Panel Survey (MEPS). In this paper we include some information on the MEPS but focus more on the CPS, the NHIS, and the ACS because they have larger sample sizes that make them better suited for monitoring coverage over time at the state and national levels and for identifying how changes may relate to changes in public policies. The key features of each of these surveys differ with respect to tracking coverage over time nationally, by state, and for types of populations that are of special concern to policy makers. Three federal agencies are responsible for producing estimates from the CPS, the ACS, the NHIS, and the MEPS. The Census Bureau produces the CPS and the ACS. The National Center for Health Statistics (NCHS), which is part of the Centers for Disease Control and Prevention (CDC), produces the NHIS. The Agency for Healthcare Research and Quality (AHRQ) produces the MEPS. Unfortunately, no federal agency is responsible for publishing tables that include estimates from all four sources or for providing guidance on what to make of the different estimates and which survey to use for different tracking needs. Another problem with the presentation of published estimates is that there is often a lack of clarity about the insurance concept and reference period reflected in the estimate. For example, the official publication of the CPS 2008 coverage estimates includes the calendar year in its table titles but does not note that the uninsured estimate is designed to represent uninsurance throughout the calendar year, or that the Census Bureau advises that the uninsured estimate can be interpreted as representing a point in time—whether that means the interview date (around March of the following year) or some point during the year the survey asks about is not made evident (DeNavas et al., 2009). The official publication of ACS 2008 coverage estimates includes the calendar year in its table titles but one must know that the ACS is a rolling survey to realize that the estimate is for an average day during the calendar year.3 The official publication of NHIS coverage 3 U.S. Census Bureau American FactFinder Table B27001. Health Insurance Coverage Status by Age for the Civilian Noninstitutionalized Population. Available: http://factfinder.census.gov/servlet/DatasetTableListServlet?_ds_name=ACS_2008_1YR_G00_&_type=table&_program=ACS&_lang=en&_ts=294308138878 [October 2010].
OCR for page 71
Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary estimates indicate which uninsurance estimates cover which time periods (“time of the interview,” “at least part of the past year,” and “more than a year”), but many people monitoring coverage may not know that these represent estimates of an average day during the calendar year (Cohen and Martinez, 2009). The table titles for the MEPS coverage estimates available on the AHRQ website indicate the time period (“first half of” calendar year or simply the calendar year) but do not indicate that the estimates represent uninsurance throughout those periods. This is especially problematic for anyone just looking at one of the several calendar years for which only half-year estimates are available, because it is less likely that the data user will figure out (i.e., by comparing calendar year estimates) that the half-year estimates are not point-in-time and thus not comparable to estimates from most other surveys.4 There is no consensus on how many children are uninsured at a point in time or throughout the year (Office of the Assistant Secretary of Planning and Evaluation, 2003). For example, the most recent year for which full-year estimates of uninsurance are available for more than one survey (assuming that the CPS estimate is not a valid measure of full-year uninsurance) is 2007, and those show a range from 3.7 million in the NHIS to 7.9 million in the MEPS (Cohen et al., 2007).5 Not only is there disagreement about how many children lack health insurance coverage at a particular point in time nationally, but state-level estimates vary across surveys as well (Blewett and Davern, 2006; Call et al., 2007). In terms of coverage estimates, the main methodological differences between the CPS, the NHIS, and the ACS relate to sample size, level of detail used in questions collecting information about coverage, other subjects asked about, characteristics of the interview, and postcollection processing (Davern et al., 2009). Other features of sampling and the particular population controls used in weighting may also explain differences in the survey estimates and their suitability for monitoring coverage. We discuss the validity of the NHIS, the CPS, and the ACS in terms of likely misclassification of coverage type, particularly Medicaid/CHIP, given current research findings and ongoing questions about the validity of coverage estimates. We also discuss sampling design, the questions included, and the validity of variables used to study key population subgroups of interest, such as children who are eligible for Medicaid/CHIP. 4 Online MEPS-HC tables available: http://www.meps.ahrq.gov/mepsweb/data_stats/quick_tables_results.jsp?component=1&subcomponent=0&year=2007&tableSeries=4&searchText=&searchMethod=1&Action=Search [June 2010]. 5 Online MEPS-HC tables available: http://www.meps.ahrq.gov/mepsweb/data_stats/summ_tables/hc/hlth_insr/2007/alltablesfy.pdf [June 2010].
OCR for page 72
Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary National Health Interview Survey There is a general consensus that the NHIS produces the most valid coverage estimates (Kenney, Holahan, and Nichols, 2006).6 The NHIS is a health-focused survey that includes many features to aid respondents in understanding the coverage question and recalling details required to correctly answer it. The NHIS features that may strengthen validity include area sample frame; well-trained interview staff that work exclusively on this survey; fairly high response rate; usually an in-person interview; a knowledgeable respondent (interviewers encourage older children to report about themselves but indicate that they want to speak about coverage with individuals who are knowledgeable about the coverage status of household members); a questionnaire that defines concepts and probes respondent memory as it collects information; breadth of content on other health-related data, which potentially helps respondents understand distinctions between coverage types and accurately classify the coverage status of the individuals whom they report about; asking about coverage source at the time of the survey, which is associated with lower measurement error; asking about Medicaid and CHIP using state-specific names; a low level of item nonresponse on insurance sequence; asking for many details about coverage (e.g., type of managed care, copayments, deductibles, need for referrals), which may help define relevant concepts and help respondent recall coverage details; asking about periods without coverage and when the child last had coverage (for use in estimating full-year uninsurance) and why it stopped (potentially helping the respondent to recall more details required to determine the child’s true coverage status); verifying no Medicaid for children with no reported coverage; asking about citizenship, place of birth, and family relationship, which are some of the important variables needed to simulate eligibility in Medicaid and CHIP; asking about medical visits and other uses of coverage or evidence of acting uninsured; 6 Results from the Medicaid Undercount project suggest that underreporting of Medicaid/CHIP is lower in NHIS than CPS; available: http://www.census.gov/did/www/snacc/ [October 2010].
OCR for page 73
Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary asking for the name of the insurance plan so the name can be matched to a list of insurance plans by state in a postcollection data processing phase and potentially used to recode misreported coverage type; and has been in production for many years and with attention to maintaining a credible time series. There are also important limitations for using the NHIS to monitor coverage. The most problematic of these is the sampling design, which limits the geographic and other subpopulation estimates that are possible as well as raising validity questions. First, the sample size is too small to produce precise annual state (and substate) estimates for most states. Second, most states have only a very small number of primary sampling units, a fact that raises concerns about the representativeness of the state-level estimates produced by the survey.7 Third, because of data confidentiality concerns, access to state identifiers is available only through data centers. The ability to use the NHIS to simulate Medicaid/CHIP eligibility is also limited by the quality of the income data, as well as the possible underreporting of the Medicaid/CHIP information coverage, despite all the efforts to measure coverage accurately (U.S. Census Bureau, 2009). The NHIS is also limiting because of the timing of the data release and what is excluded from the published estimates. There is an early release that enables some important coverage evaluations before the survey is fully prepared; however, it is still about 9 months after the interviews are completed, it does not include published estimates for children aged 0-18 separately, and it does not provide valid estimates by income. Current Population Survey The CPS has historically played an important role in monitoring coverage. Besides being a relatively large survey using high-quality data collection methods, it is an income- and employment-focused survey and is considered to have valid data on those domains, which are integral for eligibility simulations and other coverage-related analyses. Features that may strengthen validity include area frame; well-trained interview staff working exclusively on this survey; telephone and in-person interviews; high response rates; 7 NHIS PSUs cover only about 25 percent of U.S. counties.
OCR for page 74
Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary sample size is large enough for precise state estimates for large states annually; family-level questioning about coverage source, which helps get more coverage reported in large households; state-specific names and separate questions about Medicaid and CHIP; other probes and definitions (the Civilian Health and Medical Program of the Department of Veterans Affairs [CHAMPVA], direct purchase); a question on directly purchased coverage that emphasizes that it is not related to a current or former employer; asking for detailed information about coverage, including who is the policy holder, who is covered by the same policy, who is covered by someone outside the household, and employer contributions; asking several times about any other type of coverage not yet talked about; verifying the absence of insurance coverage; logical coverage edits performed by the Census Bureau to correct some likely reporting errors; asking about citizenship, place of birth, family relationship, supports from people outside the household, firm size, as well as income and employment-related factors, which are some of the important variables needed to simulate eligibility in Medicaid and CHIP; asking about health status; has been in production for many years and with attention to maintaining a credible time series; and the release of estimates and public-use files with state identifiers 5 to 6 months after the data are collected. The most critical limitation of using the CPS to monitor coverage is the known measurement error with the coverage questions because of confusion, recall bias, and other issues with the retrospective reference period (Pascale, Roemer, and Resnick, 2009). As a result of these, there is more apparent underreporting of Medicaid/CHIP coverage and considerable uncertainty about what the estimates mean, especially compared with other surveys (Davern et al., 2009a; Kincheloe, Brown, and Frates, 2006). In addition, the sample size and the number of primary sampling units is small in many states, which raises concerns about the representativeness and precision of the state estimates. Historically there has also been concern about bias in the imputation process for coverage variables; however, new imputations are being implemented at the Census Bureau to address this problem (Davern et al., 2007). The CPS is also missing
OCR for page 75
Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary information about access to, the need for, and use of health services and spending. The 9-month lag between the end of the calendar year reference period and the release of the published estimates and edited data limits the ability to track coverage in real time; however, the estimates and data are more timely if they are interpreted as representing some time closer to the interview data in March (just 6 months before the release). The published estimates are also limiting because they do not include children aged 0-18. American Community Survey The ACS, an annual survey designed to provide intercensal estimates of the information contained on the decennial census long form, added information on health insurance coverage in 2008. Although the ACS is still too new of a resource for studying trends in children’s health insurance coverage, it has a number of important strengths relative to the other surveys: The most important strengths of the ACS are its very large sample and its sample frame (which samples every county and census tract in the country), allowing for: 1-year coverage estimates for areas with a population of 65,000 or more; starting in 2011, 3-year coverage estimates for areas with populations of 20,000 or more; and starting in 2013, 5-year coverage estimates for all statistical, legal, and administrative entities. The coverage information refers to the time of the survey. It is possible to put together a variety of substate estimates, including public-use microdata areas, large counties, large metropolitan areas, etc. Comparisons with the employment-focused CPS suggest that the ACS also has fairly robust income- and employment-related data, for use in eligibility simulations and other studies. Although most data are collected by mail, the Census Bureau computes a 98 percent response rate, which is very high (Griffin and Hughes, 2010). The release of estimates and public-use files with state identifiers (8 to 9 months after the end of the survey period, implying an average lag between data collection and data release of 14 to 15 months). Since the ACS coverage data are new as of 2008, the survey cannot provide an extended pre-period for studying trends in children’s cover-
OCR for page 76
Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary age prior to the adoption of CHIPRA-related policy changes. Moreover, research is just now being conducted on the validity of the ACS estimates. Although overall the unadjusted ACS estimates of uninsured children were close to the CPS estimates, they were somewhat higher than the NHIS estimates for the same period, and reports of direct purchase of insurance on the ACS are very high (Turner, Boudreaux, and Lynch, 2009). A major limitation of the ACS for monitoring coverage is that much of the data is collected by mail (56 percent of responses in 2008), which means that most respondents complete the survey without the aid of an interviewer. Another major concern is that the coverage question includes no distinction among Medicaid, CHIP, and other sources of government insurance, and no state-specific names for Medicaid/CHIP were provided in 2008 (they were available to interviewers in computer-assisted modes starting in 2009). In addition, there is only one itemized list of coverage types (rather than a detailed series of patterned questioning, defining, and probing, as in the NHIS and the CPS), which could also introduce more measurement error in the reporting of coverage type. Also of concern is the absence of a statement that insurance purchased directly should not have anything to do with a current or former employer as well as the absence of questions about coverage details (managed care, premiums, employer contributions, and other questions that probe memory, define concepts, and can be used to recode misreports). In addition, the ACS does not include a verification of uninsurance or questions about duration of uninsurance. Another concern with the 2008 ACS estimates is that there was relatively little postcollection processing on the ACS to remedy possible reporting error. By contrast, the NHIS, for example, gives field representatives the opportunity to indicate concern about the validity of coverage reports and also collects the name of the person’s plan and uses it to reclassify coverage type. The CPS, for another contrasting example, uses other coverage-related information collected about the person or family to reclassify coverage on a logical basis. The ACS has a number of other content limitations. In particular, family relationship information is not directly available for analysis, making it much more difficult to identify health insurance units for eligibility simulations; there is no information on the child’s general health status or the parents’ firm size. And while the ACS sample is very large and its published estimates cover a variety of important geographic areas, the sample released for public-use data is smaller and excludes many geographic identifiers (e.g., congressional districts), which makes it more difficult to track meaningful coverage changes for smaller states and smaller subgroups, short of gaining access to a Census Bureau data center (which requires a comprehensive application that takes several months and must meet stringent requirements).
OCR for page 77
Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary IMPROVING FEDERAL SURVEYS Published estimates and public-use files could be made more useful for monitoring children’s coverage in a number of ways: Modify questionnaires to address known problems with coverage validity and collect other information needed to monitor coverage, especially as coverage options change with reforms: CPS: Add a question about current coverage (being tested now); NHIS: Expand income series (some changes are currently under way); NHIS and MEPS: Include a variable for the coverage status shown on the insurance card that respondents are asked to collect at the beginning of the interview; and ACS: Add a clarification to the itemized question about direct purchase (to emphasize that it is coverage that is unrelated to a current or former employer); add state-specific Medicaid/CHIP names (if adding the actual name is not feasible, add “CHIP or the children’s health insurance program in your state” or add an insert that includes a list of the Medicaid/CHIP names in different states); include more definitions of coverage types in the booklet of directions for mail respondents/interviewers and refer to their availability in the introduction to the health insurance question; add questions on firm size and general health status, verification of uninsurance, any government assistance paying for health insurance premiums (primarily for use in recoding coverage), and health insurance plan name (also primarily for use in recoding coverage). Perform call-backs of selected cases in the ACS and the CPS reported to have direct purchase, starting with those identified as low-income or logically covered by Medicaid, military, or Employee State Insurance (ESI). Use results to edit erroneous reports and to refine rules for logical coverage editing. Use all the explicit and implied information about coverage that is collected about each child and his or her family to create an edited set of coverage variables if other, more reliable reported information implies the original coverage variable is incorrect (Lynch, Boudreaux, and Davern, 2010). ACS and CPS: Create an edited version of the variable for directly purchased coverage that excludes sample people who appear to have coverage from Medicaid/CHIP or the military or other employer. Extend current logical editing rules for Medicaid/CHIP. Our research indicates that such rules for the 2008 ACS
OCR for page 78
Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary reduces the uninsured estimate for children aged 0-18 from 8.2 to 7.3 million (which is very close to the NHIS uninsured estimate) and increases the estimated number of children with Medicaid/CHIP as their primary coverage on the ACS from 19.8 to 24.4 million, which is 6.0 percent lower than the comparable administrative count for June 2008 (Lynch, 2010). Conduct more research on administrative records to identify reasons for errors in reporting children’s coverage and provide data users with methods to adjust for them: ACS, CPS, NHIS, and MEPS: Adopt methods from the Medicaid Undercount project’s research on reporting errors about enrollees of all ages (in the CPS, the NHIS, and MEPS) to research on just children (U.S. Census Bureau, 2008); and ACS, CPS, NHIS, and MEPS: Develop models to adjust children’s coverage estimates and make them available to data users, as has been done for all ages in the CPS and the NHIS (Davern et al., 2009b). Test validity of logical coverage edits against administrative data. Conduct targeted methodological research to identify survey features that can be modified to reduce reporting errors about coverage: ACS: Reassess reports of direct purchase and the method of assigning responses that are written in or reported as other coverage. For example, do not code write-in cases with reported/logical ESI, military, or Medicaid/CHIP as also having direct purchase. Assess how Massachusetts sample children with subsidized and unsubsidized coverage through the Health Connector are being reported and use findings to refine data collection strategies aimed at identifying children who end up obtaining coverage through health insurance exchanges under reform. Reexamine terms used to describe coverage types in mail mode and assess respondent ability to correctly classify coverage; and CPS, NHIS, and MEPS: Examine causal mechanisms for factors identified as predictors of Medicaid misreport. Conduct more interagency research on differences in coverage and other variables needed to study coverage: Explain differences in estimates of coverage (especially nongroup coverage and Medicaid/CHIP); and Explain differences in estimates of family income from the ACS, the CPS, and the NHIS. Present published estimates that cross-reference the other federal estimates, explain possible reasons for discrepancies across the
OCR for page 79
Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary surveys, cover more policy-relevant constructs, and include more explicit and accurate labels: ACS: Re-release 2008 estimates with the logical coverage edits adopted for 2009 and beyond. Describe published estimates as point in time or an average day in the calendar year; MEPS: Improve the policy relevance of published estimates: add point-in-time, part-year, and full-year estimates for those recent years that do not publish those estimates; link to the NHIS (the MEPS sampling frame) to provide more information on changes in coverage; CPS: Explicitly describe published estimates as approximately point in time or an average across the prior year; and ACS, CPS, NHIS, and MEPS: Include estimates for the different policy-relevant definitions of “children,” meaning both for children aged 0-18 and 0-17. Release published estimates with an introduction that informs policy makers and other users about complexities, including the likely possibility that measuring coverage is becoming more complex as the types and numbers of plans increase; the fact that the time frame for a person’s coverage status is important because a person’s health insurance status can change over time; the fact that how coverage type is defined is important because individuals may categorize their status differently from technical definitions; and the fact that these complexities are part of the reason estimates differ by survey. Include more documentation about measurement problems and provide more information in published materials to help readers correctly interpret and understand estimates. Give data users more information so they can estimate more concepts and have more flexibility dealing with limitations and comparing across surveys: ACS: provide the month of interview and person-level rural/urban or metropolitan variables (using as up-to-date information on metropolitan statistical area boundaries or local area population density as possible); CPS and ACS: Provide flags for logical editing that include an indicator of the reason for an edit; and NHIS: Provide flags for reason for coverage recode. It is important to recognize that although some of these changes are geared toward improving the validity of the coverage estimates for children, they could also introduce breaks in the time series for a particular survey. Thus, in order to assess changes in coverage over time, it
OCR for page 80
Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary would be important to make needed adjustments to the estimates so that they are comparable before and after the changes are made. In addition, we emphasize recommendations that are feasible to implement without requiring a large increase in funding. However, if we broadened the scope to include important policy questions related to access to care and service use among children, we would strongly recommend increasing the capacity of the federal surveys (such as the NHIS) to produce valid state estimates on these questions. This would require both changing the survey’s sample frame or including more primary sampling units and increasing the number of children included in the sample each year. Such an expansion would provide important information about how well individual states are doing at achieving the ultimate objective of CHIPRA and of health reform more broadly, which is to improve the health and functioning of children and adults. However, it would be critical for the state identifiers to be released so that states’ progress could be studied with public-use files. We have focused on measures designed to improve the information available on insurance coverage for children in the current coverage environment. However, it will be important for federal surveys to anticipate the new coverage options that will be available under health reform and that they adjust survey questions and content accordingly to allow the tracking of coverage at the national and state levels and for key population subgroups. Given that the major pieces of health reform are not slated to be implemented until 2014, there is time to test out new questions and to coordinate questionnaire changes across surveys so that they are in place in 2012-2013. ACKNOWLEDGMENT We appreciate the helpful feedback from the other participants at the Workshop on Evaluating Databases for Use in Uninsured Estimates for Children. This paper reflects the views of the authors and does not necessarily represent the views of the Urban Institute, its sponsors, or its trustees. REFERENCES Blewett, L.A., and Davern, M. (2006). Meeting the need for state-level health insurance estimates: Use of state and federal survey data. Health Services Research, 41(3 Pt 1), 946-975. Call, K.T., Davern, M., and Blewett, L.A. (2007). Estimates of health insurance coverage: Comparing state surveys with the Current Population Survey. Health Affairs, 26(1), 269-278.
OCR for page 81
Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary Cohen, R.A., and Martinez, M.E. (2009). Health Insurance Coverage: Early Release of Estimates from the National Health Interview Survey, 2008. Report from the National Center for Health Statistics. Available: http://www.cdc.gov/nchs/data/nhis/earlyrelease/insur200906.pdf [October 2010]. Cohen, R.A., Martinez, M.E., and Free, H. (2007). Health Insurance Coverage: Early Release of Estimates from the National Health Interview Survey, 2008. Report from the National Center for Health Statistics Available: http://www.cdc.gov/nchs/data/nhis/earlyrelease/insur200806.pdf [October 2010]. Davern, M., Rodin, H., Call, K.T., and Blewett, L.A. (2007). Are the CPS uninsurance estimates too high? An examination of imputation. Health Services Research, 42(5), 2038-2055. Davern, M., Quinn, B., Kenney, G., and Blewett, L.A. (2009a). The American Community Survey and health insurance coverage estimates: Possibilities and challenges for health policy researchers. Health Services Research, 44(2), 593-605. Davern, M., Klerman, J.A., Baugh, D., Call, K.T., and Greenberg, G. (2009b). An examination of the Medicaid undercount in the Current Population Survey (CPS): Preliminary results from record linking. Health Services Research, 44(3), 965-987. Davern, M., Klerman, J.A., Ziegenfuss, J., Lynch, V., Baugh, D., and Greenberg, G. (2009c). A partially corrected estimate of Medicaid enrollment and uninsurance: Results from an imputational model developed off linked survey and administrative data. Journal of Social and Economic Measurement, 34(4), 219-240. Davidoff, A., Kenney, G., and Dubay, L. (2005). Effects of the State Children’s Health Insurance Program expansions on children with chronic health conditions. Pediatrics, 116(1), e34-e42. DeNavas, C., Proctor, B.D., and Smith, J. (2009). Income, Poverty, and Health Insurance Coverage in the United States: 2008. U.S. Census Bureau. Current Population Reports, series P60-229. Washington, DC: U.S. Government Printing Office. Dubay, L., and Cook, A. (2009). How Will the Uninsured Be Affected by Health Reform? Washington, DC: Kaiser Commission on Medicaid and the Uninsured. Dubay, L., and Kenney, G. (2009). The impact of CHIP on children’s health insurance coverage: An analysis using the National Survey of America’s Families. Health Services Research, 44(6), 2040-2059. Dubay, L., Guyer, J., Mann, C., and Odeh, M. (2007). Medicaid at the ten-year anniversary of SCHIP: Looking back and moving forward. Health Affairs, 26(2), 370-381. Georgetown Center for Children and Families. (2009). The Children’s Health Insurance Program Reauthorization Act of 2009. Washington, DC: Georgetown Health Policy Institute. Griffin, D., and Hughes, T. (2010, May). Mixed Mode Data Collection in the American Community Survey. In the American Association of Public Opinion Research Conference Proceedings. Chicago, IL. Hudson, J., and Selden, T. (2007). Children’s eligibility and coverage: Recent trends and a look ahead. Health Affairs, 26(5), w618-w629. Kaye, N., Pernice, C., and Cullen, A. (2006). Charting SCHIP III: An Analysis of the Third Comprehensive Survey of State Children’s Health Insurance Programs. Portland, ME: National Academy for State Health Policy. Kenney, G., and Change, D. (2004). The State Children’s Health Insurance Program: Successes, shortcomings, and challenges. Health Affairs, 23(5), 51-62. Kenney, G., and Yee, J. (2007). SCHIP at a crossroads: Experiences to date and challenges ahead. Health Affairs, 26(2), 356-369. Kenney, G., Holahan, J., and Nichols, L. (2006). Toward a more reliable federal survey for tracking health insurance coverage and access. Health Services Research, 41(3), 918-945. Kenney, G., Lynch, V., and Cook, A. (2010, June). Medicaid and CHIP Eligibility and Participation rates. Urban Institute, ACS working paper.
OCR for page 82
Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary Kenney, G., Lynch, V., Cook, A., and Phong, S. (2010). Who and where are the children yet to enroll in Medicaid and the Children’s Health Insurance Program? Health Affairs, 29(10), 1920. Available: http://content.healthaffairs.org/cgi/reprint/29/10/1920 [October 2010]. Kincheloe, J., Brown, E.R., and Frates, J. (2006). Can we trust population surveys to count Medicaid enrollees and the uninsured? Health Affairs, 25(4), 1163-1167. Lynch, V. (2010). Applying Logical Coverage Edits for Analyzing Medicaid/CHIP Participation and Coverage in the 2008 ACS, Revised Edition. Washington, DC: Health Policy Center. Lynch, V., Boudreaux, M., and Davern, M. (2010, July). Applying and Evaluating Logical Coverage Edits to Health Insurance Coverage in the American Community Survey. Census Bureau, Housing and Household Economic Statistics Division working paper. Office of the Assistant Secretary for Planning and Evaluation (ASPE). (2003). Understanding Different Estimates of Uninsured Children: Putting the Differences in Context. U.S. Department of Health and Human Services working paper, May 2003. Pascale, J., Roemer, M., and Resnick, D. (2009). Medicaid underreporting in the CPS: Results from a record check study. Public Opinion Quarterly, 73, 497-520. Turner, J., Boudreaux, M., and Lynch, V. (2009). A Preliminary Evaluation of Health Insurance Coverage in the 2008 American Community Survey. U.S. Census Bureau, Housing and Household Economic Statistics Division working paper, October 2009. U.S. Census Bureau. (2008). SNACC Research Project to Understand the Medicaid Undercount. The University of Minnesota’s State Health Access Center, the Centers for Medicare & Medicaid Services, the Department of Health and Human Services Assistant Secretary for Planning and Evaluation, the National Center for Health Statistics, and the U.S Census Bureau. Phase II Research Results: Examining Discrepancies Between the MSIS and the CPS ASEC. Available: www.census.gov/did/www/snacc/SNACC_Phase_2_report.pdf [October 2010]. U.S. Census Bureau. (2009). SNACC Research Project to Understand the Medicaid Undercount. The University of Minnesota’s State Health Access Center, the Centers for Medicare & Medicaid Services, the Department of Health and Human Services Assistant Secretary for Planning and Evaluation, the National Center for Health Statistics, and the U.S Census Bureau. Phase IV Research Results: Estimating the Medicaid Undercount in the National Health Interview Survey (NHIS) and Comparing False-Negative Medicaid Reporting in NHIS to the Current Population Survey (CPS). Available: http://www.census.gov/did/www/snacc/docs/SNACC_Phase_IV_Full_Report.pdf [October 2010]. U.S. Department of Health and Human Services. (2010). Children’s Health Insurance Program Reauthorization Act: One Year Later, Connecting Kids to Coverage. Available: http://www.insurekidsnow.gov/chip/chipra_anniversary_report.pdf [October 2010].