This is the accessible text file for GAO report number GAO-06-457R entitled 'Children's Health Insurance: Recent HHS-OIG Reviews Inform the Congress on Improper Enrollment and Reductions in Low-Income, Uninsured Children' which was released on March 9, 2006. This text file was formatted by the U.S. Government Accountability Office (GAO) to be accessible to users with visual impairments, as part of a longer term project to improve GAO products' accessibility. Every attempt has been made to maintain the structural and data integrity of the original printed product. Accessibility features, such as text descriptions of tables, consecutively numbered footnotes placed at the end of the file, and the text of agency comment letters, are provided but may not exactly duplicate the presentation or format of the printed version. The portable document format (PDF) file is an exact electronic replica of the printed version. We welcome your feedback. Please E-mail your comments regarding the contents or accessibility features of this document to Webmaster@gao.gov. This is a work of the U.S. government and is not subject to copyright protection in the United States. It may be reproduced and distributed in its entirety without further permission from GAO. Because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately. March 9, 2006: Congressional Committees: Subject: Children's Health Insurance: Recent HHS-OIG Reviews Inform the Congress on Improper Enrollment and Reductions in Low-Income, Uninsured Children: The Congress passed legislation creating the State Children's Health Insurance Program (SCHIP) in 1997 to reduce the number of uninsured children in families with incomes that are too high to qualify for Medicaid.[Footnote 1] For SCHIP, the Congress appropriated $40 billion over 10 years, with funds allotted annually to the 50 states, the District of Columbia,[Footnote 2] and the U.S. commonwealths and territories. States' participation in SCHIP is voluntary. States that do participate have three options in designing their SCHIP programs: expand the Medicaid program to include SCHIP-eligible children, develop a separate child health insurance program, or maintain a program that combines both of these options. Financed jointly by the states and the federal government, SCHIP offers a strong incentive for states to participate by offering a higher federal matching rate--that is, the federal government pays a larger proportion of program expenditures-- than the Medicaid program.[Footnote 3] While this incentive encourages efforts to reduce the number of uninsured children through state participation in SCHIP, there have been concerns that states might inappropriately enroll Medicaid-eligible children in SCHIP and thus obtain higher federal matching funds than allowed under Medicaid. In addition, there has been interest in assessing the progress states made to reduce the number of uninsured children, including the extent to which states met the objectives and goals established in their SCHIP programs.[Footnote 4] In particular, states must report their progress in reducing the number of low-income, uninsured children and may rely on certain national data sets, such as the Current Population Survey (CPS), or conduct their own surveys, to do so. In the Medicare, Medicaid, and SCHIP Balanced Budget Refinement Act of 1999 (BBRA), the Congress directed the Department of Health and Human Services (HHS) Office of Inspector General (OIG) to conduct a series of studies on two issues--determining the number of children who were enrolled in separate SCHIP programs but were eligible for Medicaid and assessing states' progress in reducing the number of low-income, uninsured children--every 3 years, beginning in fiscal year 2000.[Footnote 5] This provision required the OIG to only include in its studies states with separate SCHIP programs. BBRA directed that we review and report on the OIG's work. The OIG issued its initial reports in February 2001, and our assessment of the OIG's work was published in March 2002.[Footnote 6] The OIG's most recent set of reports on these issues was published in 2004 and 2005.[Footnote 7] This report reflects our evaluation of the OIG's recent reports. Specifically, we assessed the OIG's efforts to inform the Congress on (1) the number of Medicaid-eligible children enrolled in separate SCHIP programs and (2) states' progress in reducing the number of uninsured children, including the progress they have made in meeting the objectives and goals initially established in their SCHIP programs. To assess the OIG's work, we reviewed the OIG's methodologies and findings. We also interviewed the OIG officials who conducted these studies to clarify questions and to discuss their response to our prior recommendations.[Footnote 8] Finally, we examined the OIG's recommendations to the Centers for Medicare & Medicaid Services (CMS), which administers SCHIP. Our work was conducted from December 2005 through March 2006 in accordance with generally accepted government auditing standards. Results in Brief: The OIG's most recent set of reports on improper SCHIP enrollment and states' progress in reducing the number of low-income, uninsured children informed the Congress about these issues and included improvements from its initial studies. For example, in evaluating the number of children who were enrolled in separate SCHIP programs but were eligible for Medicaid, the OIG broadened the scope of its initial study to include a random sample of children's case files from the 34 separate SCHIP programs with available data. In its initial study responding to the BBRA mandate, the OIG only examined case files from 5 separate SCHIP programs. In its most recent report, the OIG estimated that only 1 percent of children were improperly enrolled in separate SCHIP programs. The confidence interval the OIG calculated for its enrollment error rate, which provides an estimated range of values that is likely to include the true error rate, was 0.3 to 2.6 percent. We believe that this confidence interval is relatively wide for such an analysis and is likely the result of the small sample of case files reviewed by the OIG. However, we recognize that even at its upper bound, the enrollment error rate for the population would be 2.6 percent. In addition, 7 percent of separate SCHIP case files did not include enough information to support enrollment decisions, but the OIG did not find any evidence in these case files to indicate that the enrollment decisions were inappropriate. In part to respond to our earlier recommendation that the OIG expand its scope beyond the separate SCHIP programs, the OIG further informed the Congress about improper SCHIP enrollment by conducting an additional study in 29 states with Medicaid expansion programs. This study, which also evaluated the enrollment decisions in a random sample of case files, identified 7 percent of sampled children as not meeting the state eligibility criteria for Medicaid expansion and 10 percent of case files as having missing documentation. Similarly, for its most recent review of states' progress in reducing the number of low-income, uninsured children, the OIG expanded its scope to include the 46 states that submitted SCHIP annual reports for fiscal year 2002. In its initial study responding to the BBRA mandate, the OIG only examined the annual reports of 5 states with separate SCHIP programs. The OIG also supplemented its most recent review by examining several national data sources on the uninsured. The OIG noted that states continue to face challenges in their efforts to measure the change in the number of low-income, uninsured children, and only 22 of the 46 states that submitted reports directly measured their progress in this area. One of the biggest challenges in measuring progress is the limitation in data sources--including the often-used CPS, which for various reasons, such as small sample sizes, has not produced reliable state-level estimates in the past. In light of these obstacles, the OIG recommended that CMS continue to work with states to address concerns about data sources used to measure such progress. We concur with this recommendation. In addition, absent state submission of data directly measuring changes in low-income, uninsurance rates through their SCHIP annual reports, the OIG suggested, and we concur, that CMS could itself measure such reductions by completing its own analysis of available CPS data, which now include the results of broader state samples. In commenting on a draft of this report, the OIG did not comment on our findings. The OIG provided technical comments, which we incorporated as appropriate. Background: Medicaid and SCHIP, joint federal-state programs to finance health care coverage for certain categories of low-income individuals, represent the primary source of health insurance coverage for low-income, uninsured children. Although Medicaid has provided coverage to children since 1965, SCHIP is a relatively new program, established in 1997. As of January 2006, 11 states had expanded their Medicaid programs to include children eligible for SCHIP, 19 states had separate SCHIP programs, and 20 states had combination programs.[Footnote 9],[Footnote 10] (See fig. 1.) Figure 1: States' Design Choices under SCHIP, January 2006: [See PDF for image] [End of figure] Medicaid program expenditures are shared between states and the federal government, and the share is determined using a formula that is based on a state's per capita income in relation to the national average. Federal matching rates for SCHIP are "enhanced"--they are established under a formula that takes 70 percent of a state's Medicaid matching rate and adds 30 percentage points, with an overall federal share that may not exceed 85 percent.[Footnote 11] In fiscal year 2006, the enhanced federal match rates for SCHIP ranged from 65 to about 83 percent while the federal match rates for Medicaid programs ranged from 50 to about 76 percent. Under SCHIP, each state is required to submit a SCHIP plan and an annual report, which must include a description of the state's progress in reducing the number of low-income, uninsured children. States may rely on the CPS, which is a monthly survey of a sample of American households conducted by the Census Bureau and collects information on characteristics of the labor force, to report progress in reducing the number of low-income, uninsured children.[Footnote 12] In particular months, the Census Bureau supplements its survey by incorporating additional questions. For example, the March Supplement historically asks respondents about their health insurance status and provides the only nationwide source of information on uninsured children by state. The OIG's Assessment of Improper Enrollment Identified Few Errors: By broadening the scope of its initial study, the OIG's most recent set of reports on the number of children who were enrolled in separate SCHIP programs but were eligible for Medicaid more fully informed the Congress on this issue. In contrast to its initial study in which the OIG reviewed case files for five separate SCHIP programs, the OIG's most recent work included a review of a sample of case files from all separate SCHIP programs for which data were available. Similar to its initial study, the OIG identified only 1 percent of children as being improperly enrolled in separate SCHIP programs. The confidence interval the OIG calculated for its enrollment error rate is relatively wide; however, even at its upper bound, the error rate would be 2.6 percent for the population. The OIG augmented this work by also evaluating enrollment decisions in a random sample of case files from 29 Medicaid expansion states. The OIG's Assessment of Improper Enrollment Broadened to Include Nearly All SCHIP Programs: The OIG broadened the scope of its initial study from a review of 5 separate SCHIP programs to a review of the 34 separate SCHIP programs for which data were available.[Footnote 13] From these programs, the OIG selected a random sample of 400 case files to assess enrollment decisions. After eliminating cases that did not fit study criteria, the OIG ultimately reviewed 386 case files. The OIG reviewed documentation within the case files, including the SCHIP application or the most recent eligibility redetermination; supporting income documentation; and calculation sheets states used to determine family income. The OIG did not verify the accuracy and completeness of the state case files; rather, it focused on whether the information in each file supported the eligibility determination reached by the state. If case files were missing documentation, the OIG determined if the files included any information that indicated enrollment decisions were inappropriate. Using the same methodology, the OIG also reviewed a random sample of case files in 29 Medicaid expansion programs with available data.[Footnote 14] Of the 400 case files randomly selected for this study, 357 met study criteria and were reviewed. This additional work was undertaken in part to respond to our earlier recommendation that the OIG expand its review to include Medicaid expansion programs. The OIG's review of enrollment decisions in both separate SCHIP and Medicaid expansion programs went beyond the BBRA mandate and more fully informed the Congress on this issue. The OIG's sample of case files was drawn from over 80 percent of all separate SCHIP and Medicaid expansion programs with available data. Nevertheless, our assessment is that this sample was small compared to the total SCHIP population, as it represented 0.01 percent of total separate SCHIP enrollees and 0.04 percent of Medicaid expansion enrollees. The small sample size resulted in a less precise estimate of the number of cases of inappropriate SCHIP enrollment. In discussing our assessment of the sample size, the OIG emphasized the increased work associated with broadening its scope from 5 to 34 separate SCHIP programs and reviewing the additional 29 Medicaid expansion programs. The OIG also explained that the size of its sample was influenced, in part, by available resources and competing priorities. The OIG's Reviews Identified Few Examples of Inappropriate Enrollment: The OIG's findings regarding the number of children improperly enrolled in separate SCHIP programs paralleled its earlier study on this topic, with only 1 percent of children (4 of 386 cases) identified as being inappropriately enrolled. In each of these 4 cases, the children were eligible for the respective state's Medicaid program. In its Medicaid expansion study, the OIG identified 7 percent of sampled children (24 of 357 cases) as not meeting the state eligibility criteria for Medicaid expansion. Of these cases, 21 had family incomes that were too low to qualify for Medicaid expansion, and the remaining 3 had family incomes that were too high to qualify. For both separate and expansion programs, enrollment errors were due to a variety of reasons, including caseworkers misinterpreting income information, multiplying daily wages by the wrong number of days, or basing a family's income on weekly as opposed to biweekly pay. For the separate SCHIP study, the OIG projected its error rate estimate to the population, and a 95 percent confidence interval was estimated as 0.3 to 2.6 percent.[Footnote 15] The confidence interval, which we consider to be relatively wide in light of the enrollment error rate of 1 percent, is likely a result of the small sample size.[Footnote 16] However, even at its upper bound, the error rate would be 2.6 percent for the population. The OIG did not project to the population for Medicaid expansion programs because of problems identified with population data provided by certain states, such as data that mistakenly included children who were enrolled in states' traditional Medicaid programs. In addition to the definitive cases of inappropriate enrollment identified above, the OIG noted that some case files--approximately 7 percent of the 386 separate SCHIP and 10 percent of the 357 Medicaid expansion case files--did not include complete documentation to support enrollment determinations. However, the OIG reviewed the documentation included in these case files and did not identify any information that indicated enrollment decisions were inappropriate. Further, the OIG explained that for the case files with missing documentation, income levels were toward the middle of the SCHIP eligibility range, as opposed to near the lower bound of the range closer to Medicaid eligibility levels. Therefore, errors in documentation or calculations of resources would have needed to be extensive for the children to be eligible for traditional Medicaid as opposed to SCHIP. We concurred with the OIG's reasoning. The OIG Identified Challenges States Face in Determining the Number of Uninsured and Opportunities for CMS Assistance: To assess states' progress in reducing the number of low-income, uninsured children, the OIG also broadened the scope of its mandated review to include all states that submitted SCHIP annual reports for fiscal year 2002 by June 1, 2003. The OIG's review of these annual reports indicated that states continue to experience challenges when determining their progress in reducing the number of low-income, uninsured children, primarily with data sources. We agree with the OIG's recommendation that CMS continue to work with states to address concerns about data sources used to measure their progress in reducing the number of low-income, uninsured children. The OIG's Review Indicated States' Efforts to Report Progress in Reducing Uninsured Children Are Hindered by Data Limitations: Similar to its most recent work on inappropriate enrollment in SCHIP programs, the OIG expanded its review of state efforts to measure changes in the number of low-income, uninsured children. The OIG reviewed the fiscal year 2002 SCHIP annual reports of the 46 states that submitted them by June 1, 2003. In its initial report, the OIG reviewed reports from 5 states' separate SCHIP programs.[Footnote 17] The OIG reviewed the annual reports to determine states' progress in meeting the strategic objective of reducing the number of uninsured children. While 22 states used CPS or state survey data to demonstrate changes in the uninsured population of children, the remaining 24 states did not respond directly to the objective.[Footnote 18] Instead, 19 of these 24 states used SCHIP enrollment data as a proxy for demonstrating their progress in reducing the number of uninsured children. Of the remaining states, 3 provided responses that did not measure insurance coverage or enrollment, and 2 did not respond. Further, the OIG augmented its assessment of state efforts by also reviewing national data on the uninsured--including data from the CPS, the National Health Interview Survey, and the Urban Institute. These sources were consistent with the majority of states' annual reports that indicated a reduction in the number of uninsured children. By expanding its scope, the OIG went beyond BBRA's requirements to inform the Congress on states' progress in reducing the population of uninsured children. The OIG emphasized, and we acknowledge, that efforts to measure progress in reducing the number of low-income, uninsured children in states continue to be hindered by multiple factors, such as limitations in data sources and the often prohibitive cost of conducting state surveys. For example, CPS data used by many states have well- established shortcomings--particularly with regard to state-level estimates--which can be unreliable and exhibit volatility from year to year because of small sample sizes. This is particularly true in states with smaller populations. Also, children who are enrolled in Medicaid are often undercounted in CPS data and may be mistakenly counted as uninsured. Finally, as noted in the OIG's recent report, the manner in which the Census Bureau asks respondents about their health insurance coverage during the past year may lead to respondents incorrectly answering the question. As a result, CPS data may overestimate the number of uninsured children.[Footnote 19] In addition to data source problems, the OIG noted that some states use changes in SCHIP enrollment to demonstrate progress in meeting this objective. However, we agree with the OIG that increases in SCHIP enrollment are not a valid measure of reductions in the number of low- income, uninsured children. For example, an increase in SCHIP enrollment can be the result of children moving from private health insurance coverage to public insurance under SCHIP. In addition, declines in the economy and increased unemployment can lead to some children losing their private health insurance coverage and enrolling in SCHIP, and others becoming uninsured because they are ineligible for SCHIP. The OIG Suggested CMS Assist States in Future Efforts to Estimate Uninsured Children: In its most recent report, the OIG recommended, and we agree, that CMS should continue to work with states to determine whether ongoing CPS sample size improvements have alleviated concerns about limitations in the CPS data. In 1999, the Congress appropriated $10 million annually for the Census Bureau for fiscal year 2000 and subsequent fiscal years to improve the reliability of CPS data for estimating the uninsured population of low-income children. Specifically, in response to concerns about the reliability of state-level estimates, the Census Bureau increased the survey sample size for each state, which may improve the accuracy of CPS estimates of low-income, uninsured children.[Footnote 20] Although these improved data were available in March 2002, not all states used these data in their fiscal year 2002 SCHIP annual reports. Of the 12 states that used CPS data to determine their progress in reducing the number of low-income, uninsured children, only 4 used the March 2002 data in their reports. The remaining 8 states relied on data from prior years. The OIG did not explore the reasons why these 8 states did not incorporate the March 2002 data in their reports. Further, CPS data are easily accessible and are available at no cost. Therefore, absent state submission of data measuring changes in low- income, uninsurance rates through their SCHIP annual reports, the OIG officials suggested, and we concur, that CMS could itself measure such reductions by completing its own analysis of CPS data. Agency Comments: We received comments on a draft of this report from the HHS-OIG (see the enclosure). In commenting on a draft of this report, the OIG did not comment on our findings, but the OIG did provide technical comments, which we incorporated as appropriate. We are sending a copy of this report to the Inspector General of HHS and other interested parties. In addition, the report is also available at no charge on GAO's Web site at http://www.gao.gov. If you or your staffs have questions about this report, please contact me at (312) 220-7600 or firstname.lastname@example.org. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. Susan Anthony, Assistant Director; Kevin Milne; Dae Park; and Sari B. Shuman made key contributions to this report. Leslie G. Aronovitz: Director, Health Care: Enclosure: List of Committees: The Honorable Arlen Specter: Chairman: The Honorable Tom Harkin: Ranking Minority Member: Subcommittee on Labor, Health and Human Services, Education, and Related Agencies: Committee on Appropriations: United States Senate: The Honorable Charles E. Grassley: Chairman: The Honorable Max Baucus: Ranking Minority Member: Committee on Finance: United States Senate: The Honorable Ralph Regula: Chairman: The Honorable David R. Obey: Ranking Minority Member: Subcommittee on Labor, Health and Human Services, Education, and Related Agencies: Committee on Appropriations: House of Representatives: The Honorable Joe Barton: Chairman: The Honorable John D. Dingell: Ranking Minority Member: Committee on Energy and Commerce: House of Representatives: Comments from the Department of Health and Human Services: DEPARTMENT OF HEALTH & HUMAN SERVICES: Office of Inspector General: Washington, D.C. 20201: MAR 3 2006: Ms. Leslie G. Aronovitz: Director, Health Care: U.S. Government Accountability Office: Washington, DC 20548: Dear Ms. Aronovitz: Enclosed are the Department's comments on the U.S. Government Accountability Office's (GAO) draft report entitled, "CHILDREN'S HEALTH INSURANCE: Recent HHS-OIG Reviews Inform the Congress on Inappropriate Enrollment and Reductions in Low-Income, Uninsured Children" (GAO-06- 457R). These comments represent the tentative position of the Department and are subject to reevaluation when the final version of this report is received. The Department provided several technical comments directly to your staff. The Department appreciates the opportunity to comment on this draft report before its publication. Sincerely, Signed by: Daniel R. Levinson: Inspector General: Enclosure: The Office of Inspector General (OIG) is transmitting the Department's response to this draft report in our capacity as the Department's designated focal point and coordinator for U.S. Government Accountability Office reports. OIG has not conducted an independent assessment of these comments and therefore expresses no opinion on them. COMMENTS OF THE DEPARTMENT OF HEALTH AND HUMAN SERVICES ON THE U.S. GOVERNMENT ACCOUNTABILITY OFFICE'S DRAFT REPORT ENTITLED, "CHILDREN'S HEALTH INSURANCE: RECENT HHS-OIG REVIEWS INFORM THE CONGRESS ON INAPPROPRIATE ENROLLMENT AND REDUCTIONS IN LOW-INCOME, UNINSURED CHILDREN" (GAO-06-457RI: The Department of Health and Human Services (HHS) appreciates the opportunity to comment on the draft report. The Medicare, Medicaid, and SCHIP Balanced Budget Refinement Act of 1999 (BBRA) mandates that every 3 years the Office of Inspector General (OIG) determine the number of separate SCHIP enrollees, if any, who are eligible for Medicaid and the reduction in the number of low-income, uninsured children. Your report carries out the additional BBRA mandate that GAO review these two OIG studies. Our comments follow. In the first paragraph on page 2, GAO indicates that Congress asked OIG to study two issues-inappropriate enrollment and the reduction in the number of low-income, uninsured children. In regard to the first issue, BBRA asks OIG to determine the number of children, if any, who are enrolled in separate SCHIP but are eligible for Medicaid (rather than inappropriate enrollment.) Congress's concern, as expressed through the mandate, is that children eligible for Medicaid would instead be enrolled in SCHIP, thus allowing a State to claim a higher match rate. This clarification should be noted throughout when it describes the purpose of this study. In the second paragraph on page 8, when discussing Medicaid expansion, GAO's presentation implies that OIG projected its findings to the universe of children enrolled in Medicaid expansion programs. OIG did not. We suggest deletion of the sentence that discusses OIG not calculating a confidence interval because a discussion of confidence intervals is not warranted when no projection has occurred. If GAO chooses not to delete this sentence, we suggest that GAO replace the current language, "The OIG did not calculate a confidence interval for the error rate estimate for the Medicaid expansion programs.." with the following language, "The OIG did not project to the universe for Medicaid expansion programs.." [End of section] (290512): FOOTNOTES  Medicaid is a federal-state program that provides health care coverage to certain categories of low-income adults and children. SCHIP was established as title XXI of the Social Security Act by the Balanced Budget Act of 1997, Pub. L. No. 105-33, § 4901, 111 Stat. 251, 552, and is codified at 42 U.S.C. § 1397aa, et seq.  The District of Columbia is included among our discussion of states for purposes of this report.  Federal funds are allotted to states for SCHIP programs up to a specified amount each year. See 42 U.S.C. § 1397dd.  The SCHIP statute includes a provision requiring states, in establishing their programs, to specify strategic objectives and performance goals for providing child health assistance. See 42 U.S.C. § 1397gg.  Pub. L. No. 106-113, App. F., § 703, 113 Stat. 1501A-321, 1501A- 401-402.  GAO, Children's Health Insurance: Inspector General Reviews Should Be Expanded to Further Inform the Congress, GAO-02-512 (Washington, D.C.: Mar. 20, 2002).  Department of Health and Human Services Office of Inspector General, SCHIP: States' Progress in Reducing the Number of Uninsured Children (Washington, D.C.: August 2004); Determining if Children Enrolled in Separate SCHIPs Were Eligible for Medicaid (Washington, D.C.: June 2005); and Determining if Children Classified as SCHIP Medicaid Expansion Meet Eligibility Criteria (Washington, D.C.: October 2005).  See GAO-02-512. In that report, we noted that the OIG's findings could not be generalized to all SCHIP programs because sample cases were limited to five states' separate SCHIP programs. Therefore, we recommended that the OIG (1) expand its scope when conducting subsequent mandated studies and (2) review enrollment practices in states with Medicaid expansion programs, in addition to separate SCHIP programs, to further inform the Congress about the appropriateness of enrollment. Because state evaluations of reductions in the uninsured had limitations, we also suggested that the OIG review other available literature on changes in the uninsured population for its next study.  Prior to September 30, 2002, Tennessee had a Medicaid expansion program under SCHIP, which covered children born before October 1, 1983, and who were under age 19 with family incomes up to 100 percent of the federal poverty level. After September 2002, Tennessee discontinued its SCHIP program because all enrolled children had aged out of the program.  At the time of the OIG's review, New York had a combination SCHIP program. New York's program changed to a separate SCHIP program as of April 1, 2005, when all of the children enrolled in the state's Medicaid expansion aged out of the program.  For example, a state with a 50 percent Medicaid match receives a 65 percent match under SCHIP.  The CPS is the primary source of information on the labor force characteristics of the U.S. population, and estimates obtained from the CPS include employment, unemployment, earnings, and hours of work.  Two separate SCHIP programs--Michigan and Rhode Island--were unable to provide necessary data to the OIG.  Michigan's Medicaid expansion program was unable to provide necessary data to the OIG.  The OIG did not report confidence intervals for its initial review of inappropriate SCHIP enrollment.  A confidence interval provides an estimated range of values, within which the true error rate for the population will likely fall. For this study, the OIG calculated that the true enrollment error rate could be from 0.3 to 2.6 percent, which is 70 percent below and 160 percent above the estimated error rate.  Four of the five states excluded from the most recent OIG review- -Connecticut, Hawaii, Minnesota, and Nevada--were excluded because they did not submit their SCHIP annual reports by June 1, 2003. The remaining state, Tennessee, was not required to submit an annual report because there was no one enrolled in its SCHIP program.  Of the 22 states that directly demonstrated changes in the uninsured population of children, 12 states used CPS data and 10 states used state survey data.  Although the CPS asks respondents if they had health insurance coverage within the past year, the question is asked at a specific point in time and may result in respondents answering incorrectly. For example, those who had health insurance at some time during the year, but who are uninsured at the time of the survey, may mistakenly answer the question with their current uninsured status, which can lead to an overestimate of the uninsurance rate.  When reporting uninsurance rates, the Census Bureau reports 3-year averages. Therefore, at least 4 years of data will need to be collected to measure the full impact of the expanded sample. The 4 years of data will allow for a comparison of the change in the rate of low-income, uninsured children from two consecutive 3-year averages.