CALIFORNIA California Program on Access to Care

Transcription

1 CALIFORNIA PROGRAM ON ACCESS TO CARE California Policy CALIFORNIA Research POLICY Center RESEARCH CENTER California Program on Access to Care University of California, Office of the President 1:00 p.m. Welcome and Introductions Gilbert Ojeda, Director, CPAC, UC Office of the President Andres Jimenez, Director, CPRC, UC Office of the President University of California POLICY BRIEFING Access to Care for Vulnerable Populations Wednesday, August 23, :00 p.m. to 3:00 p.m. California State Capitol, Room 2040 A G E N D A 1:10 p.m. The Effect of Public Reporting and Pay-for-Performance on California s Safety Net Hospitals R. Adams Dudley, MD, MBA and Lauren Goldman, MD, UC San Francisco 1:35 p.m. Impact of Welfare Reform on Access to Medical Care, Mental Health Services, and Substance Abuse Treatment for CalWORKS Participants Deborah Podus, PhD, UC Los Angeles 2:00 p.m. Increasing Access to Health Care for the Poor: What Can We Learn from the SB 87 Experience? Jane Mauldon, PhD and Kamran Nayeri, PhD, UC Berkeley 2:25 p.m. Dental Check-up of the Healthy Families Program Umo Isong, PhD, UC San Francisco 2:50 p.m. Questions/Comments 3:00 p.m. Adjourn

2 Protecting Vulnerable Patients from Unintended Consequences of Public Reporting of Hospital Quality and Hospital Pay-For-Performance L.E. Goldman, MD, MCR, S. Henderson, PHD, and D. Dohan, PhD, R.A, Dudley, MD, MBA University of California, San Francisco DNN01K

3 Widespread Movement to Publicly Report Health Care Quality Multiple stakeholders Employer groups Health plans Consumer groups Governmental organizations

4 Public reporting of quality Comparative data among health care providers By state, county, city, and zip Hospitals and physicians Mandatory statewide hospital reports of risk-adjusted mortality for: Heart attacks, pneumonia, hip fracture, Coronary Artery Bypass Grafts Who is reporting data on the internet? Employers Center for Medicare and Medicaid Services Foundations

5 Pay-for-performance Tie reimbursement to scores on performance indicators Multiple domainsincluding Patient safety Quality Information technology Patient satisfaction Demonstration projects CMS- inpatient hospital pay-forperformance com/all/quality/hqi/

6 Public reporting of quality Proponents of public reporting argue: Disclosure of comparative data permits consumers to choose the highest quality provider Creates market pressure to improve quality Provides mechanisms for purchasers to create incentives for all providers to improve

7 Quality Improvement at Safety Net hospitals Safety net hospitals provide a substantial level of care to lowincome, uninsured, and vulnerable populations with limited or no alternative access to care Certain hospital characteristics used to define safety net hospitals: Hospitals serving a high percentage of Medicaid & uninsured patients Public ownership Rural location

8 Public reporting of quality Proponents of public reporting argue: Disclosure of comparative data permits consumers to choose the highest quality provider Vulnerable patients may have few choices among providers Geographic barriers, limited or no insurance, clinical conditions

9 Public reporting of quality Proponents of public reporting argue: Disclosure of comparative data permits consumers to choose the highest quality provider Vulnerable patients may have few choices among providers geographic barriers, insurance limitations, clinical conditions Creates market pressure to improve quality Different market incentives Rural location may have fewer competitors Underinsured and uninsured patients may have few alternatives

10 Public reporting of quality Proponents of public reporting argue: Disclosure of comparative data permits consumers to choose the highest quality provider Vulnerable patients may have few choices among providers geographic barriers, insurance limitations, clinical conditions Creates market pressure to improve quality Different market incentives Rural location may have fewer competitors Underinsured and uninsured patients may have few alternatives Provides mechanisms for purchasers to create incentives for all providers to improve Main purchasers driving pay-for-performance may not be major purchasers in safety net hospital system

11 Public Reporting and pay-for-performance & Safety net hospitals Public expectation of public reporting and pay-for-performance may increase expenditures on performance measurement without compensatory increase in revenue. Expenditures without any compensatory increase in revenue could be detrimental to long term goals to improve quality at safety net facilities

12 How do these quality improvement initiatives affect hospitals that serve vulnerable populations? Aim #1 To compare quality of care between California safety net hospitals and other hospitals before widespread initiation of pay-for-performance. Aim #2 To describe perceptions of safety net hospital representatives about values of and problems with participation in public reporting and pay-forperformance

13 Methods- Aim #1 Cross-section analysis of safety net vs. non-safety net hospital quality Subjects All non-federal California acute care hospitals Databases Hospital Quality Alliance 1 st 2 quarters (1/04-6/04) 2004 OSHPD financial database

14 Methods- Aim #1 Hospital Quality Alliance Starting in 2003, CMS sponsored this national voluntary public reporting initiative 10 performance measures 5 acute myocardial infarction 2 congestive heart failure 3 community acquired pneumonia

15 Analysis- Aim #1 Outcome: 10 HQA performance indicators % of eligible patients who received appropriate care 2 safety net hospital definitions > 50% of the discharges reimbursed by Medicaid or uninsured Public hospital ownership

16 Public vs. Private hospitals, adjusted Measures N Public Hospitals Nonpublic hospitals 95% CI ASA on arrival for heart attack % 95% 0.46 (0.30 to 0.70) ASA on discharge for heart attack % 93% 0.32 (0.20 to 0.54) Beta-blocker on arrival for heart attack % 87% 0.46 (0.32 to 0.67) Beta-blocker on discharge for heart attack % 88% 0.37 (0.25 to 0.55) ACE inhibitor for heart attack % 75% 0.40 (0.24 to 0.67) Assessment of left ventricle function in heart failure % 72% 0.77 (0.51 to 1.18) ACE inhibitor for heart failure % 81% 0.56 (0.42 to 0.75) Oxygen assessment in pneumonia % 98% 0.51 (0.22 to 1.22) Pneumovax in pneumonia % 24% 0.55 (0.29 to 1.06) Appropriate antibiotic timing in pneumonia % 69% 0.97 (0.76 to 1.22 *adjusted for teaching status, number of staffed beds, number of emergency room visits

17 Public vs. Private hospitals, adjusted Measures N Public Hospitals Nonpublic hospitals 95% CI ASA on arrival for heart attack % 95% 0.46 (0.30 to 0.70) ASA on discharge for heart attack % 93% 0.32 (0.20 to 0.54) Beta-blocker on arrival for heart attack % 87% 0.46 (0.32 to 0.67) Beta-blocker on discharge for heart attack % 88% 0.37 (0.25 to 0.55) ACE inhibitor for heart attack % 75% 0.40 (0.24 to 0.67) Assessment of left ventricle function in heart failure % 72% 0.77 (0.51 to 1.18) ACE inhibitor for heart failure % 81% 0.56 (0.42 to 0.75) Oxygen assessment in pneumonia % 98% 0.51 (0.22 to 1.22) Pneumovax in pneumonia % 24% 0.55 (0.29 to 1.06) Appropriate antibiotic timing in pneumonia % 69% 0.97 (0.76 to 1.22 *adjusted for teaching status, number of staffed beds, number of emergency room visits

18 Public vs. Private hospitals, adjusted Measures N Public Hospitals Nonpublic hospitals 95% CI ASA on arrival for heart attack % 95% 0.46 (0.30 to 0.70) ASA on discharge for heart attack % 93% 0.32 (0.20 to 0.54) Beta-blocker on arrival for heart attack % 87% 0.46 (0.32 to 0.67) Beta-blocker on discharge for heart attack % 88% 0.37 (0.25 to 0.55) ACE inhibitor for heart attack for heart attack % 75% 0.40 (0.24 to 0.67) Assessment of left ventricle function in heart failure % 72% 0.77 (0.51 to 1.18) ACE inhibitor for heart failure % 81% 0.56 (0.42 to 0.75) Oxygen assessment in pneumonia % 98% 0.51 (0.22 to 1.22) Pneumovax in pneumonia % 24% 0.55 (0.29 to 1.06) Appropriate antibiotic timing in pneumonia % 69% 0.97 (0.76 to 1.22 *adjusted for teaching status, number of staffed beds, number of emergency room visits

19 High Medicaid and Uninsured vs. Other Hospitals, adjusted Measures N High Medicaid & Uninsured Other hospitals 95% CI ASA on arrival for heart attacks % 95% 0.64 (0.39 to 1.07) ASA on discharge for heart attacks % 92% 0.42 (0.19 to 0.89) Beta-blocker on arrival for heart attacks % 86% 0.60 (0.39 to 0.93) Beta-blocker on discharge for heart attacks % 87% 0.41 (0.23 to 0.73) ACE inhibitor for heart attack % 74% 0.79 (0.44 to 1.43) Assessment of left ventricle function in heart failure % 73% 1.06 (0.74 to 1.52) ACE inhibitor in heart failure % 80% 0.91 (0.67 to 1.25) Oxygen assessment in pneumonia % 99% 0.29 (0.133 to 0.63) Pneumovax in pneumonia % 24% 0.62 (0.33 to 1.16) Appropriate antibiotic timing in pneumonia % 69% 0.63 (0.53 to 0.77) *adjusted for teaching, number of staffed beds, number of emergency room visits

20 Conclusions High Medicaid/uninsured and public hospitals had lower measured quality at the onset of the HQA public reporting project in several quality indicators across 3 different clinical conditions Differences between safety net and other hospitals varied depending on the performance indicator

21 Possible explanations Safety net hospitals may have had minimal QI efforts directed to these measurements prior to this initiative True quality differences Measurement differences

22 Limitations Voluntary participation Cross-sectional analysis Evaluated processes of care not outcomes Unknown relationship between statistical and clinical differences

23 Policy Implications At baseline, safety net institutions had lower measured quality of care Important to assess quality improvement and disparities in care over time to evaluate if these quality differences lead to health care disparities Important to assess how quality improvement strategies affect quality of care at the safety net facilities

24 Methods- Aim #2 Qualitative study of California safety net hospital executives Opinions about public reporting and pay-for-performance Suggestions to improve public reporting and pay-forperformance

25 Subjects Random stratified sample of California safety net hospital executives High Medicaid and uninsured Public Rural

26 Semi-structured telephone interviews Conducted 6/05-9/05 Elicited Opinions about public reporting and pay-for-performance Suggestions to improve and minimize harm to safety net hospitals Performed thematic analysis N-Vivo qualitative software

27 Findings 37 hospitals participated 9 rural 10 public, high Medicaid/uninsured 9 public, low Medicaid/uninsured 9 private, high Medicaid/uninsured

28 Benefits of participation Used data for benchmarking and QI projects Develop feasible goals Used to show stakeholders hospital quality

29 Biases against safety net hospitals

30 Lack of resources influences measured quality If I lose a biller, I can t just put an ad in the paper and say I need a biller, and one comes and starts to work. There s not another biller in this whole community that knows how to bill medical billing.

31 Lack of resources for trainings Performance will depend on how well trained the extractor is; those who know which documents to take the information from significantly improve their hospital s results. Others can overlook things.

32 Less likely to be computerized We re not as computerized as the private sector. Hospitals with online charting of information have the ability track results and focus on improvement in a way we cannot match.

33 Difficulty gaining physician buy-in This is a medically underserved area and we need the existing physicians. There s no incentive for them to help us with public reporting or pay-for-performance.

34 Economic benefits weren t necessarily applicable We re the only show in town. There s no competition. Ours isn t a market driven commodity. People aren t buying it on the basis of price or reputation. We get whatever the market doesn t want. I m not sure that public reporting will drive paying patients to our institutions.

35 Small sample size It takes manpower. Whenever we didn't have enough patients to get counted, it was a waste of our time. I m concerned about the impact of our small numbers. I want to see some data before we start attaching potential negative financial consequences.

36 Concern that Case-Mix affects quality measurements We have many patients who wait until they re very sick to go to the hospital, then get discharged to the street and don t return for follow-up appointments. Our return rate was less than 1%, so we stopped surveying our patients. They don t have telephones or change telephone services monthly or change addresses monthly.

37 Suggestions Sponsors of public reporting and pay-for-performance should offer educational resources to safety net hospitals. Support for implementation of electronic medical record for hospitals with limited resources due to payer-mix. Comparisons hospitals with similar geographic constraints and case-mix. Consider how social factors influence quality measurement Consider measurements applicable to safety net hospitals such as availability of translators

38 Limitations Qualitative design prevents hypothesis testing In-depth interview limit sample size Perspective of safety net hospital executives

39 Policy Implications Public reporting and pay-for-performance are likely to force safety net hospitals to either commit additional resources to performance measurement or risk appearing unwilling to share performance data Policy makers should consider strategies to reduce the cost to these institutions, enhance the capacity for them to improve through trainings and resource support. Failure to do so may exacerbate pre-existing disparities in resources and quality of care for patients using safety net hospitals.

40 Impact of Welfare Reform on Access to Health, Mental Health, and Substance Abuse Services for CalWORKs Participants Deborah Podus, Ph.D., P I M. Douglas Anglin, Ph.D., Co-P I UCLA Integrated Substance Abuse Programs Semmel Institute for Neuroscience and Human Behavior Funded by the California Program on Access to Care-California Policy Research Center and the Robert Wood Johnson Foundation-Substance Abuse Policy Research Program CPAC-CPRC Policy Briefing, Sacramento, CA, August 23, 2006

41 Background Welfare reform and the emphasis on moving welfare recipients from welfare to work heightened concern about substance abuse as a barrier to employment nationally. Reasons for greater attention by policymakers: - Assumption that drug and alcohol abuse is prevalent in the welfare population - Belief that substance abuse fosters welfare dependence - Belief that substance abuse treatment would be an effective way to move substance-abusing participants from welfare to work Many welfare agencies across the country instituted substance abuse screening and treatment programs based on these assumptions.

42 Research Problem Many welfare departments found that fewer clients than expected disclosed substance use to agency staff. Some possible reasons for low rates: - Rates of substance use were lower than estimated - Problems with program design or implementation Our study examined several research questions with respect to this problem focusing on Los Angeles County: - How prevalent is substance use among CalWORKs clients? - How does substance use affect approval for CalWORKs? - What is the impact of substance use on on access to health, mental health, and substance abuse treatment services among those approved for CalWORKs?

43 Study Methods Study had two components: Baseline prevalence interview and a 9-month follow-up interview with a subset of subjects. Baseline interview: - Probably eligible applicants for CalWORKs - CalWORKs participants undergoing a routine annual review Follow-up interview: only respondents who were probably eligible applicants at baseline Data were collected between November 2000 and May 2002 Eligibility criteria: - At least 18 years of age and - Speak either English (61.9%) or Spanish (38.1%)

44 Baseline Prevalence Study Baseline recruitment and interviews were conducted at all 24 CalWORKs offices in Los Angeles County. Recruitment was based on a flyer system and depended on collaboration among DPSS administrators, CalWORKs district office administrators and staff, and the UCLA research team. Baseline interviews were conducted with 287 probably eligible applicants and 224 CalWORKs participants. In addition to interviews, 77.2% of respondents voluntarily provided a urine sample; 94.8% took a Breathalyzer test; 67% consented to the release of certain CalWORKs administrative data.

45 Follow-up Cohort Study Follow-up interviews were conducted with a 155 (of 254) probably eligible applicants who were interviewed at baseline. The follow-up period was about 9 months. Interviews were conducted in person (73.2%) and by telephone (26.8%). Respondents who participated in a faceto-face interview could also provide a voluntary urine sample. Response rate at follow-up: 61%.

46 Baseline Sample Characteristics (n=509) Female 95.9% Mean age 32.1 years Ethnicity Latino 61.2% White 9.2% African American 26.8% Other 2.6% Marital status Married 16.8% Widowed/separated/divorced 28.0% Never married 55.2% Born in the United States 51.7%

47 How prevalent is substance use?

48 Respondents Report Substance Use Is Prevalent in Their Communities (n=509) Top three responses to an open-ended item asking subjects to name the most important problem in their community: Gangs 20.7% Alcohol and drug use 20.3% Crime and violence 8.5% Of six possible community problems, substance abuse was the one rated extremely serious or very serious most often. Top five: Alcohol and drug use 55.7% Lack of programs for kids 46.6% Environmental pollution 40.0% Crime and violence 36.8% Inadequate health care 30.2% Subjects reporting immediate family and/or close friends who have had a serious alcohol or drug problem: 46.9%

49 Reluctance to Disclose Use Complicates Estimation of Drug Use Prevalence % Tested % Tested positive % Tested Positive positive and denied use and denied use (n=393) in last 3 days in last 30 days Marijuana 9.2% % Cocaine/ Amphetamines/ Opiates 6.6% 92.3% 84.6%

50 Prevalence of Problematic Drug Use Based on Multiple Measures Respondents were grouped by their risk of having an AOD problem. Those at higher risk should be assessed. Measure is conservative. High risk of problematic use, 10.0% (n=51): Test positive or report recent opiate/stimulant use; use marijuana 10+ days in last 30; intoxicated 7+days of last 30; had AOD problem in last 30 days; or AOD treatment in last 6 months Medium risk of problematic use, 14.7% (n=75): Not high risk and test positive for or report other AOD use; AOD treatment > 6 months ago; report DUI arrest; or record of any AOD-related arrest Low, 57.0%, or presumed low, 18.3%, risk of drug use (n=383): Not at high or medium risk; those at low risk tested negative for all drugs tested; those presumed low did not provide a urine test

51 What is the relationship among problematic drug use, CalWORKs participation, and access to care?

52 Problematic Substance Use and Welfare Participation CalWORKs applicants who were at risk for problematic substance use were not less likely than other applicants to be approved for CalWORKs benefits. 95.3% problematic users vs. 78.9% other applicants Among those approved for CalWORKs benefits, those at risk for problematic substance use were not less likely than others to feel they were treated fairly. 87.8% problematic users vs. 70.9% other applicants Participants at risk for problematic use were slightly more likely to report that their benefits had been reduced for failure to follow program rules, but the difference was not significant. 39% problematic users vs. 32.6% other applicants

53 Dissemination of Information about Substance Abuse and Mental Health Services Informed About AOD Services Informed About Mental Health Services Informed About Medicaid % Recipients who said they were informed about substance abuse services, mental health services, and Medi-Cal (n=127)

54 Substance Abuse and Mental Health Screening for Referral to Assessment Respondents were less likely to report being screened for substance abuse or mental health problems than to report being informed about substance abuse and mental health services: % reported being asked about their substance use % reported being asked about their mental health Respondents at risk for problematic use were no more likely to report being screened that other respondents % problematic users vs. 51.8% other applicants reported being asked about their substance use % problematic users vs. 33.7% other applicants reported being asked about about their mental health

55 Receipt of Health and Mental Health Services CalWORKs aid was an effective pathway to Medi-Cal health insurance for most respondents regardless of their level of drug use. It was especially effective in providing health coverage for respondents children: - Over 95% of respondents reported health coverage for their children The percentage of cohort study respondents who reported having received mental health services in the last 6 months increased between baseline and follow up, but service use was low given that the majority of respondents reported 1 or more emotional problems at baseline. - Percentage of problematic users who reported receiving mental health services went from 12.2% at baseline to 17.1% - Percentage of low risk respondents receiving mental health services went from 8.1% to 9.3%

56 Receipt of Substance Abuse Services The percentage of problematic users who reported receiving substance abuse treatment in the last six months was unchanged between baseline and follow up: - Percent who reported recent treatment at baseline: 7.3% - Percent who reported recent treatment at follow-up: 7.3% The percentage of respondents at low risk of problematic use who reported getting substance abuse treatment went from 0% at baseline to 2.3% at follow up. Among cohort study participants who authorized release of of their administrative data, none were recorded as being referred for clinical assessment or as having received substance abuse treatment through the CalWORKs program.

57 Conclusions I Findings from the baseline study: Welfare applicants and recipients are concerned about substance use: - Over 50% view substance use as a serious problem - Almost 50% had family/friends who have had a serious AOD problem Approximately 10% of respondents are at high risk of problematic use (i.e., use that impairs their functioning) and another 15% at moderate risk. Thus, about 25% are suitable candidates for referral for clinical assessment for substance abuse. This estimate, based on multiple measures including urine tests, is conservative. Reluctance of substance using respondents to divulge their substance use complicates the estimation of rates of use and the identification of those individuals who would benefit from clinical assessment.

58 Conclusions II Findings from the cohort follow-up study: Applicants at risk of problematic use are not less likely to be approved for CalWORKs benefits. This a positive, important study finding. Many recipients are informed about the availability of substance abuse and mental health services, but rates could be higher. Respondents are less likely to be screened about their use of drugs or emotion problems than they are to be informed about substance abuse and mental health services. CalWORKs is an effective link to Medi-Cal health insurance, but is less effective in linking respondents who need it to mental health and especially substance abuse treatment.

59 Conclusions III Though initiated as a program to promote employability, CalWORKs substance abuse and mental health services should also be considered an important component of health policy. Relatively low rates of problematic use in the CalWORKs population, uneven implementation of screening, and limited access to care for those who need it suggest a need to better target screening efforts. Study limitations: the sample was not randomly selected; follow-up response rate was 61%; drug use estimates are conservative. Study strengths: use of multiple data sources.

60 Acknowledgments We express our appreciation to the UCLA-ISAP CalWORKs study research team and to the many colleagues at ISAP that contributed their time and expertise to this research project. Thanks also to the many staff and administrators from the Los Angeles County Department of Public Social Services who facilitated our research efforts and compiled administrative data for research study purposes. Their research collaboration was essential to the conduct of the study. For more information contact: Deborah Podus, Ph.D. UCLA-ISAP 1640 So. Sepulveda Blvd, Ste. 200 Los Angeles, CA Ph: (310)

61 Increasing Access to Health Care for the Poor: What Can We Learn from the SB 87 Experience? Jane Gilbert Mauldon, Sung Man Cho, and Kamran Nayeri University of California, Berkeley Survey Research Center/UC DATA Funding provided by the California Program on Access to Care, California Policy Research Center

62 An impact evaluation of the Senate Bill (SB) 87 Enacted in Sep to ensure that Medi-Cal eligible individuals leaving cash aid (TANF, SSI) do not lose Medi-Cal coverage Previously, many Medi-Cal eligible leavers lost coverage (e.g. Mauldon, Nayeri and Dobkin 2002). Requires counties to continue leavers on Medi-Cal until an ex parte review evaluates their eligibility. If necessary, county staff contact clients by phone and letter for needed information Reduced leavers paperwork to stay on Medi-Cal : no Medi-Cal eligibility redetermination form to continue benefits and no quarterly reporting requirements.

63 Research Questions: 1. Has SB 87 increased the rate of Medi-Cal enrollment of TANF and SSI leavers? 2. Has SB 87 increased their enrollment in regular Medi-Cal programs? 3. How has Temporary Medi-Cal use changed for TANF leavers? How the new Redetermination and Appeal categories have been used for SSI leavers? 4. Has SB 87 reduced cross-county differences in the Medi-Cal enrollment of TANF and SSI leavers? 5. How have ethnic and language subgroups benefited from SB 87?

64 Data and Methods Use Medi-Cal Eligibility Data System (MEDS) 1/ /2004 to identify leavers and track Medi-Cal status. Leavers are those with at least one month of cash aid, followed by three consecutive months of no cash aid. Outcome Measure: Medi-Cal enrollment at the fourth month after leaving aid. Exclusion: Because of a Medi-Cal waiver in effect Los Angeles county is excluded from the evaluation study.

65 Study Samples For TANF leavers: The Pre-SB 87 sample had its last month on aid January May 2001 The Post-SB 87 sample had its last month on aid June June 2004 For SSI leavers: SSI impacts were complicated by the Craig v Bonta lawsuit The Pre-SB 87 sample had its last month on aid January 1999 April 2002 The Post-SB 87 sample had its last month on aid January 2004 August 2004

66 Background: SSI Appeals and Redetermination codes Medi-Cal Code 6N (In Appeal) was first used in June 2001, just before SB 87 implementation date of July 1, 2001 Litigation meant that pre-sb 87 rules applied until June 2002 This is outcome month Sept It and subsequent months are excluded, until May 2004 initiates post-sb 87 Medi-Cal Code 6E (pending SB 87 Redetermination) was first used in May 2003 Half of recent exits in that month put into 6E. Over the following six months, one-quarter were put into 6E. Use of 6E declines steadily until April 2004, as the backlog is processed; stabilizes at 12% 6E, 2% 6N.

67 Findings: TANF Medi-Cal coverage rose after SB 87 By 13 percentage points for TANF leavers By 15 percentage points for SSI leavers The increase was associated with SB 87, not other factors The Medi-Cal growth for TANF leavers was in regular, not Temporary, coverage: Suggests that prior to SB 87, many Medi-Cal eligible TANF leavers were dropped from Medi-Cal SB 87 helped address this serious problem

68 Medi-Cal coverage rose among TANF leavers, both overall and even more in regular Medi-Cal. 80% 70% 60% 50% 40% 30% 20% 10% 0% All Medi-Cal Regular Medi-Cal Before SB 87 After SB 87

69 Post-TANF coverage in the typical county increased. Counties at the bottom came up towards the middle Medi-Cal total enrollment rate Maximum Median Minimum Sample: The 35 largest counties (95% of population) PreSB87 Time_period withsb87

70 More TANF leavers were in regular Medi-Cal programs in 57 of 58 counties 1 ; the median rose 16 points Regular Medi-Cal enrollment rate Maximum Median Minimum Alameda PreSB87 Time_period withsb87 1. The sole exception is San Mateo, with 0.4% of the state s welfare population. Its total post-tanf Medi-Cal enrollment rose, however.

71 The seven largest counties vary widely in their effects. Enrollment in San Bernardino County rose most, followed by Orange and Santa Clara Counties. Alameda County s post-tanf Medi-Cal enrollment fell. 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Alameda Orange Riverside Sacramento San Bernardino San Diego Santa Clara All others (not LA) Before SB 87 After SB 87

72 An implication of this cross-county variation: The exiters who benefited the most lived in counties that previously had low post-tanf Medi-Cal rates One third (33%) of TANF exiters lived in ten counties that previously had low Medi-Cal enrollment: Kern, Kings, Monterey, Madera, San Bernardino, San Joaquin, Santa Clara, Solano, Stanislaus, and Ventura Counties In these counties: Pre-SB 87 % on Medi-Cal = 46% Post-SB 87 % on Medi-Cal = 70% Gain in coverage of 26 percentage points Compare to statewide effect, which was 61% to 74%.

73 Findings: SSI Counties enrolled more SSI leavers in Medi-Cal The effect of SB 87 was concentrated in younger clients (under age 65). Half of SSI leavers are working-age adults, and 10% are children The elderly leaving SSI enrolled in Medi-Cal at roughly the same rate (20%) pre- and post-sb 87 Enrollment in regular Medi-Cal changed little with SB 87. Instead, SB 87 put many younger SSI leavers in Redetermination status

74 The median county s post-ssi Medi-Cal rate rose. County rates had been similar, but now diverged Maximum Median total_rate Median Minimum PreSB87 Time_period withsb87

75 SSI leavers enrollment in regular Medi-Cal programs rose very little overall after SB Maximum regular_rate Median Minimum PreSB87 Time_period withsb87

76 The coverage gains of SB 87 are seen mainly for leavers under age 65. (The proportion exiting due to death is presumably high among older leavers) Pre-SB 87 Post-SB Under Age 18 Age Age Age Age 85- plus

77 All the coverage gains and coverage differences are in the Redetermination/Appeal status Pre-SB 87, % in Redetermination/ Appeal Post-SB 87, % in Redetermination/ Appeal 0 Under Age 18 Age Age Age Age 85- plus

78 Findings: Temporary Medi-Cal statuses for TANF and SSI leavers The temporary codes are place-holders for coverage during the ex parte evaluation Counties that had used Temporary Medi-Cal heavily for TANF leavers reduced their use All counties began to use the temporary Redetermination code for SSI leavers Counties varied in their use of this code

79 Use of Temporary Medi-Cal for TANF leavers fell among high-use counties; the median changed little Alameda Temporary Medi-Cal enrollment rate Maximum Maximum Median Minimum PreSB87 Time_period withsb87

80 SB 87 effectively created temporary Medi-Cal programs (Appeals and Redeterminations) for SSI leavers. The use of these statuses varies by county Maximum Median in_appeal_rate Minimum PreSB87 Time_period withsb87

81 Counties still differ more in their rates of temporary coverage for TANF leavers than for SSI leavers. Maximum Median Minimum

82 The seven largest counties illustrate that counties differed initially in their use of Temporary Medi-Cal for TANF leavers, and then became more similar. 80% 70% 60% 50% 40% 30% 20% 10% 0% Alameda Orange Riverside Sacramento San Diego San Bernardino Santa Clara All others (not LA) Before SB 87 After SB 87

83 The county data also illustrates that counties responded in different ways to the new post-ssi temporary categories (Redetermination and Appeal) 30% 25% 20% 15% 10% 5% 0% Alameda Orange Riverside Sacramento San Diego San Bernardino Santa Clara All others (not LA) Before SB 87 After SB 87

84 Findings: Race/ethnic differences Ethnic differences appear among TANF leavers Ethnic/language differences most pronounced in rates of Regular Medi-Cal enrollment Rates of Regular Medi-Cal enrollment may be influenced by geographic concentration and community networks and cohesion Asians, Pacific Islanders and Vietnamese do better: These groups had higher post-tanf Medi-Cal rates than other groups prior to SB 87, AND They showed larger SB 87 impacts Non-English speakers in these groups did especially well.

85 Enrollment in Regular Medi-Cal shows ethnic/language differences pre- and post-sb 87 80% 70% 60% 50% 40% 30% 20% 10% 0% Other SE Asian/ Not English Other SE Asian/ English Lang. Vietnamese/ Vietnamese Vietnamese/ English or Other Asian-PI/ Not English Asian-PI/ English Lang. Latino, Not English Latino, English Lang. Black White Pre-SB 87, on Regular Medi-Cal Post-SB 87, on Regular Medi-Cal

86 Conclusions and Implications SB 87 has substantially increased Medi-Cal enrollments among eligible TANF and SSI leavers SB 87 has reduced cross-county disparities in post-tanf Medi-Cal enrollment, by replacing variation in Temporary Medi-Cal use with a more uniform use of regular Medi-Cal programs Use of SB 87 redetermination category accounts for almost all the increase in post-ssi Medi-Cal. County staff interviews suggest that further simplification of Medi-Cal redetermination paperwork will enhance the impact of SB 87. In particular, county staff offered ideas on improving MC 355 form.

87 Future research: How should we interpret the fact that the gains in post-ssi coverage are because cases are now in Redetermination? Are most of these cases ultimately denied? What explains the larger SB 87 gains among Asian/Pacific Islander and SE Asian non-english speakers? (It is not due to their counties of residence.) Social cohesion and well-organized community groups? Why are post-tanf enrollments not higher still closer to 100%? Are some eligibles still not being enrolled? If so, Why?

88 Dental Check-up of the Healthy Families Program Umo Isong, DDS MPH PhD Jane Weintraub, DDS MPH Center to Address Disparities in Children s Oral Health University of California, San Francisco

89 Background Dental insurance coverage facilitates access to dental care Insured children are more than twice as likely to use dental services as children with no dental insurance

90 SCHIP State Children s Health Insurance Program (SCHIP) targets uninsured children with family incomes too high to qualify for Medicaid, up to 200% of the federal poverty level (FPL)

91 Design choices States are allowed to choose from one of three program designs: Medicaid expansion Stand-alone SCHIP Combination of both

92 California s program California s child health insurance program combines Medicaid expansion with a separate SCHIP component, the Healthy Families (HF) program

93 Design choices Design has programmatic and financial implications for access: Dental benefits Other design options include: Eligibility requirements Cost sharing provisions Payment arrangements

94 Access to dental care Other factors at the local and state level may also affect dental access: Availability and accessibility of providers Perceived need for care

95 Access to dental care SCHIP statute requires each state to monitor their progress toward achieving performance goals One recommended performance measure is access to dental care States use varying approaches and data to monitor access to care

96 Project Aims Determine the use of dental services during 2001 by children ages 2 to 11 years with Healthy Families dental coverage

97 Project Aims Compare the use of dental services by children with Healthy Families dental coverage (HF) to that of children with private dental coverage (PC), Medicaid dental coverage (DC), or no dental coverage in 2001

98 Project Aims Identify factors which influence access to dental care among insured and uninsured children in the state

99 Methods Data from the 2001 California Health Interview Survey were used Data analyzed on 10,454 children included: Child s most recent dental visit Child s type of health insurance Child s dental insurance status

100 Methods Additional multivariate analysis was performed on ~9200 children with: Continuous health insurance coverage, or No health insurance coverage during the year

101 Key findings Nearly one in four children (23.4%) had no dental coverage at the time of the survey More than half (57.1%) of the children reporting no dental coverage were eligible for either Healthy Families or Medi-CAL insurance

102 Type of child s dental coverage Coverage type Prevalence (%) None 23 Denti-CAL 19 Private 52 HF 5

103 Dental coverage Approximately 7.6% of children with Healthy Families insurance were not aware they had dental coverage Represents more than 21,000 children ages 2 to 11 years statewide

104 Demographic characteristics HF children were more likely to be from near poor, working households than DC children HF children were less likely to be from English-only speaking homes (26%) than either DC children (36%) or PC children (65%)

105 Dental service use by CA children (%) Coverage type ADV PDV None Denti-CAL Private HF *ADV=Annual Dental Visit; **PDV=Preventive Dental Visit

106 Continuous insurance Children with continuous insurance were more likely to use dental services than those with coverage for only part of the year (76% vs. 63%) HF children were less likely to have continuous insurance (87%) than DC children (94%) or PC children (97%).

107 Usual Source of Care (USC) Children with no USC were less likely to use dental services than children with a USC (49% vs. 75%) HF children were more likely to have no USC (2.5%) than DC (1.4%) or PC (1.0%) children.

108 Multivariate model (n=9,178) Continuously insured DC children were more likely to use dental services than their HF peers, despite adjusting for factors such as poverty Continuously insured PC children were more likely to use preventive dental services than their HF peers, despite adjusting for factors like USC

109 Limitations The project used data from 2001 Data from CHIS were based upon self-reported data Comparisons to the 2003 CHIS were planned but were not possible due to variations in key questions

110 Recommendations 1. Improve HF program participation among plans and providers 2. Increase the number of HF providers, particularly in rural communities 3. Improve enrollment and retention of children in the HF program

111 Recommendations 4. Increase HF program participation by HF-eligible children 5. Provide families language assistance with their applications 6. Assign each enrolling child to a provider, preferably one that speaks their language

112 Recommendations 7. Inform families of enrolling children of the dental benefits of the child 8. Eliminate or reduce financial barriers where feasible, e.g. the maximum monthly premium payable per family 9. Perform ongoing monitoring of access to care using independent sources of data.

113 Acknowledgements Supported by funds from the California Program on Access to Care (CPAC), California Policy Research Center, and the UCSF Center to Address Disparities in Children s Oral Health.