Welcome and Introductions 2:00 P.M. Gilbert Ojeda, Director, CPAC, UC Office of the President California Program on Access to Care University of California Challenges in Medi-Cal THURSDAY, MARCH 27, 2008 2:00 P.M. TO 4:00 P.M. CAPITOL BUILDING ROOM 3191 SACRAMENTO, CALIFORNIA AGENDA Presentations 2:05 P.M. Impact of Medicare Part D on Access to Antipsychotic Drugs and Hospital Costs among Medi-Cal Dual Eligibles in California Martha Shumway, PhD, Professor, UCSF/SFGH, Psychiatry Department 2:35 P.M. Confronting Healthcare Disparities in Medi-Cal Managed Care: The Role of Ethnicity, Race, and Primary Language Michael A. Rodriguez, MD, MPH, Professor, UCLA, David Geffen School of Medicine 3:05 P.M. Evaluating the Impact of Medi-Cal s Extended Lock-In Period on Children s Insurance Coverage and Hospitalizations Andrew Bindman, MD, PhD, Professor, UCSF, Chief, Division of General Internal Medicine 3:35 P.M. Reactor: Challenges to Medi-Cal in 2008 and Beyond Toby Douglas, Deputy Director, Health Care Programs, Department of Health Care Services 3:50 P.M. Questions and Comments 4:00 p.m. Adjourn Note: presentations are 15 minutes allowing for up to 15 minutes of questions and dialogue
Impact of Medicare Part D on Access to Antipsychotic Drugs and Hospital Costs in California Martha Shumway, PhD and Sukyung Chung, PhD University of California, San Francisco Funded by the California Program on Access to Care Applied Policy Research Program grant ENN11N and NIMH grant T32-MH018261. 1
Medicare Part D Prescription drug coverage for Medicare beneficiaries since January 2006. Coverage provided through multiple private Prescription Drug Plans. Expanded coverage for most beneficiaries. Restricted coverage for persons dually eligible for Medicare and Medicaid, who previously had comprehensive Medicaid drug coverage. 2
Dual Eligibles ~20% of California's Medicare beneficiaries (~ 1 million people) Compared to other Medicare beneficiaries: have more complex health problems have more co-morbid conditions take more medications have lower incomes 3
Dual Eligibles and Part D Dual eligibles are at risk for medication disruption under Part D because drug plans: do not cover all drugs have limited networks of pharmacies use co-payments and other utilization controls to contain costs 4
Dual Eligibles and Antipsychotic Drugs 20% of dual eligibles take antipsychotic drugs for major mental illnesses. The thought disorders and cognitive impairments that make antipsychotic drugs necessary may increase the difficulty of deciphering drug plan rules and solving coverage problems, leading to increased medication disruption. Disruptions in antipsychotic drugs can quickly lead to relapses that require costly hospital care. 5
Study Overview Conducted in 2006, the first year of Part D implementation. Used available data to estimate Part D's impact on: access to antipsychotic drugs hospital costs Evaluated the impact of policy changes hypothesized to improve access and reduce hospital costs. 6
Methods Since no actual implementation data were available in 2006, we used: decision tree / simulation models data from a variety of sources (Medicaid drug utilization data, Medicaid Statistical Information Statistics, drug formularies of Medicare prescription drug plans, surveys of Medicare beneficiaries, the Medical Expenditure Panel Survey, data from articles published in academic journals) probabilistic sensitivity analyses to examine uncertainties Modified original model to examine the impact of recommended policy changes. Estimated hospital costs associated with each model using published cost values. 7
Decision Tree Showing Medication Disruption Plan known or verified at the pharmacy.7 Plan unknown.3 In-network pharmacy.65 Out-of-network pharmacy.35 Reenroll at the pharmacy.7 Refill denied.3 All Rx covered Incomplete coverage.83.17 Go to an in-network pharmacy.8 Refill delayed.2 [+] Go to another pharmacy & reenroll.7 Refill delayed.3 [+] [*] 1 Get all Rxs with no utilization control.540 PA.102 QL.188 Tier 2+ (no QL,PA).170 Switch to another plan.7 Refill delayed.3 All Rxs covered.83 Incomplete coverage [+] 1.17 0.3.2.1 [+] 1 [+] [*] 8
Findings Over 150,000 dual eligible Californians were taking antipsychotic drugs. 10 drugs accounted for over 90% of antipsychotic drug prescriptions. 5 drugs accounted for 80% of prescriptions. Frequently prescribed drugs were all brand name drugs, thus 86% of prescriptions were subject to utilization controls (e.g., additional copayments, prior authorization, or quantity limits). 9
Findings 21% of dual eligibles' antipsychotic drug prescriptions were likely to be disrupted due to delayed filling of prescriptions or the need to switch medications. Hospital costs associated with drug disruption were estimated to total $61.9 million ($399 per dual eligible person taking antipsychotic drugs). 10
Findings Four potential policy changes were examined: Insure that plan enrollment can be verified or initiated during a pharmacy visit 90% of the time. Simplify prior authorization procedures so that no more than 20% of prescriptions are delayed while waiting for physician documentation. Eliminate prior authorization completely. Eliminate quantity limits on prescriptions. 11
Findings Impact of policy changes on drug disruption risk Initial policies Improve enrollment verification Simplify prior authorization (no more than 20% disruptions) Eliminate prior authorization Eliminate quantity limits All policy changes in combination 21% 20% 18% 17% 19% 14% 0% 5% 10% 15% 20% 25% % prescriptions disrupted 12
Findings Impact of policy changes on hospital costs Initial policies Improve enrollment verification Simplify prior authorization (no more than 20% disruptions) Eliminate prior authorization Eliminate quantity limits All policy changes in combination $61.9 $60.7 $55.3 $51.2 $56.2 $41.1 0 10 20 30 40 50 60 70 80 millions of $ in added hospital costs 13
Policy Implications Part D appears to put dual eligible persons who take antipsychotic medications at considerable risk of medication disruption, which is associated with notable increases in hospital costs. A range of policy changes could potentially reduce disruption risk and hospital costs. 14
Policy Implications Findings from 2006 remain relevant in the third year of Part D implementation: Enrollment issues are of continued concern as plans change. In 2008, as many as half of California's dual eligibles need to change plans to avoid premiums. The majority would be automatically assigned to a new plan. Utilization controls remain in place over time and use of utilization controls has increased in each year of Part D implementation. 15
Policy Recommendations Eliminate or reduce prior authorization requirements for beneficiaries with an established chronic illness. Eliminate or increase quantity limits to reduce co-payment burden on beneficiaries. Continue to improve interagency coordination and communication to guarantee enrollment during pharmacy visits. Obtain and analyze actual data on Part D implementation at State and local levels. 16
Confronting Healthcare Disparities in Medi-Cal Managed Care Michael A. Rodriguez, MD, MPH Associate Professor UCLA Department of Family Medicine March 27, 2008
Acknowledgments Dr. Richard Sun Medi-Cal Managed Care Division Blue Cross of California Central Coast Alliance for Health Contra Costa Health Plan Health Net Inland Empire Health Plan Kern Health Systems LA Care Health Plan San Francisco Health Plan California Program on Access to care
Disparities and Data Health and healthcare disparities are widespread and persistent Need for more data on care by ethnicity/race and language IOM recommends ethnicity/race and language data collection and reporting
Data Collection and Monitoring Identifies magnitude and factors associated with disparities Assists health plans to monitor performance Ensures accountability to enrollees and payors Allows for evaluation of intervention programs Identifies discriminatory practices
Medicaid Largest health care services provider to low income ethnic minority populations in U.S. California has the largest Medicaid population in the United States (Medi-Cal) Half of Medi-Cal enrollees are cared for through Medi-Cal Managed Care (MMCD)
Current Practice in MMCD Collects data on ethnicity/race and language Requires health plans to collect annual HEDIS quality of care data Does not routinely assess care quality by ethnic/racial or primary language groups
Study Objectives Identify and quantify ethnic/racial and primary language disparities in the quality of care for Medi-Cal Managed Care enrollees. Provide recommendations to address findings.
Study Population Eight managed care plans participated 70.4% of total MMCD population 16 counties (from San Diego to Sacramento) All plan models represented: Two-Plan Model (Commercial Plans and Local Initiatives) Geographic Managed Care County Organized Health Systems Source: Management Information System/Decision Support System containing eligibility as it appears on the Medi-Cal Eligibility Data System (MEDS) file. Medical Plan members include all counties of operation
Main Outcome Categories Health plans collected data on 16 measures Child and Adolescent Healthcare Women s Healthcare Chronic and Other Diseases
Child and Adolescent Healthcare Child Immunization Status Appropriate Treatment for Upper Respiratory Infection Well Child Visits (< 15mo) Well Child Visits (3-6yrs) Adolescent Well Care Visits
Women's Healthcare Chlamydia Screening Prenatal Care Postpartum Care Breast Cancer Screening Cervical Cancer Screening
Chronic and Other Diseases Diabetes: Hemoglobin A1c Testing Diabetes: Retinal Eye Exam Diabetes: LDL-c Screening Diabetes: Nephropathy Screening Appropriate Treatment for Asthma Appropriate Treatment for Acute Bronchitis
Ethnicity/Race The Medi-Cal application: Ethnicity/Race is optional Write-in Field
Ethnicity/Race Categories Non-Latino White Latino African American Asian or Pacific Islander Asian or Pacific Islander Amerasian Asian Indian Japanese Korean Hawaiian Guamanian Samoan Chinese Cambodian Filipino Vietnamese Laotian
Primary Language The Medi-Cal application: What Language/Dialect do you speak best?
Language Categories English Spanish Asian or Pacific Islander languages Cantonese Japanese Korean Tagalog Mandarin Cambodian Lao Other Chinese Samoan Vietnamese Mien Hmong Thai
Data Management 2006 HEDIS-like data submitted to MMCD Data were de-identified & combined at MMCD Working with MMCD and the Health Plans, UCLA team reviewed and cleaned the data Stratified analysis
RESULTS
Population Size HEDIS-like Measure Total Observations Child Immunization Status (CIS) 8,925 Appropriate Treatment for Upper Respiratory Infection (URI) 178,903 Well Child Visits, < 15mo (W15) 6,473 Well Child Visits, 3-6yrs (W36) 13,950 Adolescent Well Care Visits (AWC) 9,335 Chlamydia Screening (CHL) 73,121 Prenatal Care (PC) 8,378 Postpartum Care (PPC) 8,376 Breast Cancer Screening (BCS) 24,699 Cervical Cancer Screening (CCS) 27,866 Diabetes Care: Hemoglobin A1c Testing (CDC) 8,564 Diabetes Care: Retinal Eye Exam 8,564 Diabetes Care: LDL-c Screening 8,564 Diabetes Care: Nephropathy Screening 8,564 Appropriate Treatment for Asthma (ASM) 23,977 Appropriate Treatment for Acute Bronchitis (AAB) 10,439
Ethnicity/race among Indicators 100% Ethnic Composition of Study Population per HEDIS-like Measure Missing 80% 60% 40% Asian/Pacific Islander African American Non-Latino White Latino 20% 0% CIS URI W15 W36 AWC CHL PC PPC BCS CCS CDC ASM AAB
Primary Language Among Indicators Language Composition of Study Population per HEDIS-like Measure 100% 80% 60% Missing API Language Spanish English 40% 20% 0% CIS URI W15 W36 AWC CHL PC PPC BCS CCS CDC ASM AAB
Child and Adolescent Healthcare Indicators by Ethnicity 90% 80% 70% Non- Latino White 60% Latino 50% 40% 30% 67.3% 78.4% 62.1% 80.7% 83.2% 77.9% 84.0% 79.6% 55.3% 55.2% 47.3% 57.1% 62.4% 73.6% 64.2% 72.5% 31.7% 35.1% 32.7% 41.4% African American Asian or Other Pacific Islander 20% Child Immunization Status Appropriate Treatment for Upper Respiratory Infection Well Child Visits (< 15mo) Well Child Visits (3-6yrs) Adolescent Well Care Visits
Child and Adolescent Healthcare Indicators by Language 90% 80% 70% English 60% 50% 40% 30% 68.2% 82.8% 88.4% 81.8% 77.2% 79.6% 52.7% 57.7% 74.2% 64.7% 76.6% 76.8% 32.9% 35.6% 44.3% Spanish Asian or Pacific Islander Language 20% Child Immunization Status Appropriate Treatment for Upper Respiratory Infection Well Child Visits (< 15mo) Well Child Visits (3-6yrs) Adolescent Well Care Visits
Women s Healthcare Indicators by Ethnicity 90% 80% 70% Non-Latino White 60% Latino 50% 40% 30% 45.8% 48.1% 53.4% 41.8% 82.9% 83.1% 79.3% 80.8% 57.7% 61.0% 47.1% 64.3% 49.2% 55.3% 42.4% 56.3% 59.5% 70.7% 66.1% 68.2% African American Asian or Other Pacific Islander 20% Chlamydia Screening Prenatal Care* Postpartum Care Breast Cancer Screening Cervical Cancer Screening
Women s Healthcare Indicators by Language 90% 80% 70% 60% English 50% 40% 30% 50.3% 43.8% 38.1% 81.9% 85.0% 83.4% 55.3% 71.4% 70.2% 45.7% 57.2% 57.1% 62.1% 70.5% 72.0% Spanish Asian or Pacific Islander Language 20% Chlamydia Screening Prenatal Care* Postpartum Care Breast Cancer Screening Cervical Cancer Screening
Chronic and Other Illness Indicators by Ethnicity 90% 80% 70% Non- Latino White 60% Latino 50% 40% 30% 70.4% 74.7% 66.0% 76.2% 46.9% 50.2% 45.0% 55.2% 79.9% 83.8% 74.3% 82.8% 43.8% 46.7% 42.0% 48.0% 84.7% 83.9% 82.5% 84.9% 45.5% 44.6% 47.7% 39.2% African American Asian or Other Pacific Islander 20% Diabetes Care: Hemoglobin A1c Testing Diabetes Care: Retinal Eye Exam Diabetes Care: LDL-c Screening Diabetes Care: Nephropathy Screening Appropriate Treatment for Asthma* Appropriate Treatment for Acute Bronchitis
Chronic and Other Illness Indicators by Language 90% 80% 70% 60% English 50% 40% 30% 69.2% 78.0% 79.9% 45.4% 53.2% 56.7% 78.0% 87.2% 88.6% 43.2% 50.6% 43.9% 83.3% 85.4% 84.0% 43.3% 42.6% 37.6% Spanish Asian or Pacific Islander Language 20% Diabetes Care: Hemoglobin A1c Testing Diabetes Care: Retinal Eye Exam Diabetes Care: LDL-c Screening Diabetes Care: Nephropathy Screening Appropriate Treatment for Asthma Appropriate Treatment for Acute Bronchitis
Summary of Results African Americans received lowest quality of care for 10 of 16 measures Non-Latino Whites received lowest quality of care for 3 of 16 measures Asian or Pacific Islander language speakers received highest quality of care for 9 of 16 measures English speakers received lowest quality of care for 13 of 16 measures
Findings and Limitations Questions on ethnicity/race and primary language are optional and not standarized Inconsistent data collection (e.g. applicant vs. intake worker) Variable rates of missing ethnicity and language data; ranging from 4% to 22% for ethnicity (Mean: 8%) and <1% to 15% for language (Mean: 5%) No process for validating ethnicity and language data Beneficiary knowledge, beliefs, attitudes toward healthcare, overall health, SES Population issues (e.g., continuous enrollment, enrollment bias, case mix) No information on physician-patient ethnic or linguistic concordance, patient adherence, provider setting (e.g. CHC)
Recommendations Ensure standardization of ethnicity/race and primary language data collection during the application process Provide standardized options for ethnicity/race and language with specific subcategories for broad groups Implement a method of validating ethnic/racial and primary language data after collection Adopt aggregations of ethnicity and language data that can be used in all internal reports
Recommendations Expand the requirement for MMCD quality reviews to include ethnicity/race and primary language as part of the analysis Develop and evaluate quality improvement projects to reduce and/or eliminate disparities in healthcare Offer technical, administrative and financial incentives to health plans to address disparities Support research to better understand the reasons for the observed disparities
Questions & Comments?
Evaluating the Impact of Medi- Cal s Extended Lock-In Period on Children s Insurance Coverage and Hospitalizations Andrew B. Bindman, MD Professor Medicine, Health Policy, Epidemiology & Biostatistics University of California San Francisco Supported by Commonwealth Fund
Eligibility Re-determination More than half of Medicaid beneficiaries nationwide have interruptions in coverage Medicaid re-enrollment policies affect the number of beneficiaries who have interruptions in coverage Federal requirement of at least annual redetermination State laws vary regarding use of shorter period of eligibility re-determination
Natural Experiment of Interrupted Medi-Cal Coverage California extended Medi-Cal eligibility redetermination period for all children in California from every 3 to every 12 months on January 1, 2001 17 other states had a similar policy
Study Questions How did the reduced administrative burden of re-enrollment in Medi-Cal affect the continuity of Medi-Cal coverage for children? What were the health and cost consequences of this policy change?
Pre/Post Study of Re-Enrollment Policy Change for Children Children 1-17 years with a minimum of 1 month of Medi-Cal coverage in California Outcome = time to a hospital admission for an ambulatory care sensitive condition Main predictor = time period Pre policy change = Jan 1999 to December 2000 Post policy change = Jan 2001 and December 2002
Linked CA Hospital Discharge and Medi-Cal Eligibility Files OSHPD: Hospital Discharge Data 1999 Diagnosis (ICD-9 Code) 2002 Linkage Month/Year of admission DHS: Medi-Cal Enrollment Database 1999 Demographics Monthly enrollment history Aid Category (e.g. TANF or SSI) FFS, managed care Other insurance 2002
Ambulatory Care Sensitive Conditions: AHRQ Prevention Quality Indicators 1. Condition with acute course and window for intervention 2. Condition with chronic course amenable to self-management ACS Conditions Acute Conditions: Dehydration Ruptured Appendicitis Cellulitis Bacterial Pneumonia Urinary Tract Infection Chronic Conditions: Asthma Hypertension COPD Diabetes Mellitus Heart Failure Angina
Children 1-17 Years in California Medi-Cal Before and After Extension of Re-Determination Period 1999-2000 2001-2002 N 3,288,171 3,230,120 Mean Age (yrs) 9 9 % Female 50 51 Ethnicity (%) Hispanic 54 56 Black 13 12 Asian 8 8 Other 25 24 Aid Group (%) TANF 47 50 SSI 3 3 Other 50 47 Managed Care (%) 47 41
Children with Continuous Medi-Cal Enrollment by Time Period Percentage 70% 60% 50% 40% 30% 20% 10% 0% 62 49 Pre: 1999-2000 Post: 2001-2002 Years of Enrollment
Probability of a Hospitalization for an ACS Condition Over Time Cumulative Probability 0.40 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00 Before 2001 Enrollment Extension After 2001 Enrollment Extension 0 3 6 9 12 15 18 21 24 Time (Months)
Children: Adjusted Risk of ACS Hospitalization Relative Hazard P-Value Post policy 0.74 <.0001 Age 0.88 <.0001 Female 0.97 0.0175 Ethnicity Hispanic 3.26 <.0001 Black 4.70 <.0001 Asian 1.10 0.0926 Other 2.97 <.0001 Aid Group TANF 1.47 <.0001 SSI 24.9 <.0001 Managed Care 0.82 <.0001
Comparison Group: Adults in Medi-Cal Medi-Cal eligibility re-determination period did not change during study period for adults in California Adults with Medi-Cal coverage 1999-2000 = 62% 2001-2002 = 60% Adjusted relative hazard of a hospitalization for an ACS condition for adults in post vs pre period= 1.11
Hospital Savings and Medi-Cal Coverage Costs Associated with Enrollment Extension in 2001 2002 Hospital Savings Reduction in number of hospital days 15,635 Average daily Medi-Cal payment for a hospitalization for an ACS condition $1,129 Total hospital savings $17,651,915 Medi-Cal Coverage Costs Increased number of continuously enrolled months of Medi-Cal coverage Average monthly capitation rate Total Medi-Cal Costs for increased coverage TANF/Other SSI TANF/Other SSI TANF/Other SSI 1,395,283 41,524 $96 $370 $134,393,685 $15,350,490 Total Medi-Cal Coverage Costs $149,744,175
Subsequent Health Insurance Coverage for Children with Interrupted Medi-Cal Coverage Among hospitalized children 59% regained Medi-Cal 7% were uninsured 33% had another form of insurance
Limitations Do not have complete information on all health care related and administrative costs of more frequent eligibility re-determination Experiment of lengthening eligibility lock-in period was with children and results may not apply to adults
Probability of Adult ACS Hospitalization Over Time by Medi-Cal Coverage Status Cumulative Probability 25% 20% 15% 10% 5% 0% Interrupted Continuous 0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 Time (Months)
Policy Implications Less frequent eligibility determination reduces an administrative barrier that increases continuity of Medi-Cal coverage Continuity of Medi-Cal coverage can support better health and decrease wasteful spending on hospitalizations that could have been avoided with less costly outpatient care Proposals to increase the frequency of eligibility redetermination as a cost cutting measure should consider the administrative and health related costs
CaMRI California Medicaid Research Institute Collaborative partnership between University of California and California Department of Health Care Services Focus is on health policy research, evaluation, and technical assistance Similar model in several other states