Using Quality Improvement to Reduce Racial and Ethnic Disparities in Medicaid Managed Care: Lessons from Oregon Matthew Carlson, Ph.D. Assistant Professor of Sociology Portland State University Charles Gallia, Ph.D. c Acting Research Manager Oregon Department of Human Services Office of Medical Assistance Programs 1
Identifying Disparities Using HEDIS and CAHPS PROS: Relatively simple to do if race data are available. Advantage of familiarity to QI staff and Management. Corresponds to ongoing efforts of QI programs. Barriers: Timing and frequency of measurement. CAHPS composites mask differences in quality/access. Not all QI goals reflected w/hedis (e.g., asthma outcomes, cost, health status). Sampling methods don t t always allow for racial comparisons. Race over-sample necessary. Continuous enrollment criteria may mask disparities. 2
Example Using HEDIS Nerenz & Darling, 2004. HRSA. Nerenz et al. 2002; Commonwealth Fund. 3
100 90 80 70 60 50 40 30 20 10 0 % Always How often did your child s [Dr.] talk with you about how your child is feeling, growing, behaving? % Always How often did your child s [Dr.] treat you with courtesy and respect? % Never Did you have a hard time speaking with your child s [Dr.] because of language? % 10 Cultural Differences in 2004 Oregon CAHPS Measures English Spanish How would rate your child s Dr.? Need to consider cultural differences in survey response. 4
Early Lessons: The CareOregon Pediatric Asthma Program Purpose of the project was to improve asthma care for children: particularly urban minorities Asthma QI team approach: QI coordinator; medical director, social worker, evaluator, analyst Recruited bicultural/bilingual social worker Worked with clinic provider champions in 2 clinics serving predominantly minority children Identified project participants from administrative data and providers Patient Intervention: Social worker provided patient education Clinic Intervention: Provider reports Reviewing patient charts/treatment NHLBI Guidelines 5
Geography is Everything: Patient Demographics Program CareOregon Number 79 Mean Age 7.4 Male 62% Language English 46% 71% Spanish 54% 23% Race/Ethnicity African American 19% 9% Hispanic 52% 30% White 25% 54% Other 4% 7% 6
Percent of Each Group Attending 1 or More Classes (n=79) Language 1 % Attending Spanish English 78% 39% Race 1 Hispanic White Black Other 73% 40% 47% 0% 7
Self-Reported Reported Outcomes for Participants Past 2 Weeks (n=60) Baseline 6-Month Follow-up # Missed school days # Symptom days # Days interfered with social/family activities % following action plan (Past 2 Months) 0.58 6.2 1.10 27%.47 3.2 (p.<.05) 1.30 77% (p.<.01) 8
Percent of Members 0-180 with an ED visit or Hospitalization by Race/Ethnicity Percent with an ED Visit by Year Percent with an Inpatient Visit by Year White Hispanic White African-American 36 30 24 18 12 6 12 10 8 6 4 2 0 Year 1 Year 2 0 Year 1 Year 2 Participating Clinics Only 2002-2003 2003 9
Percent of Members 0-180 Followed-up Within 30 Days of ED or Inpatient Visit 70 60 50 40 30 20 African American White Hispanic 10 0 Year 1 Year 2 Participating Clinics Only 2002-2003 2003 10
Aim: BCAP: Reducing Disparities in Diabetes Care Improve diabetes care for Latinos enrolled in each of 3 participating health plans. Reduce racial and ethnic disparities in diabetes care. Did the team achieve its aim? Plan A: Both Latinos and whites improved in rates of LDL and HbA1c testing. Latinos showed greater improvement than whites. Plan B: Mixed results. disparities were reduced (Whites improved more than Latinos) or worsened (Latinos did worse relative to Whites). The intervention occurred only during the last 3 months of the HEDIS reporting year. The number of Latinos with diabetes also increased during this period, this may have brought down testing rates. 11
Percent Receiving Tests By Ethnicity 2005-2006: 2006: Plan A 70% 60% 50% 40% 30% White Latino 20% 10% 0% % with >1 ED Visit 2005 % with >1 ED Visit 2006 % LDL 2005 %LDL 2006 %HbA1c 2005 %HbA1c 2006 White N=522/582; Latino N=54/59 12
Diabetes Quality Measures By Ethnicity 2005-2006: 2006: Plan B 60% 50% 40% 30% 20% White Latino 10% 0% % with >1 ED Visit 2005 % with >1 ED Visit 2006 % LDL 2005 %LDL 2006 %HbA1c 2005 %HbA1c 2006 White N:3160/3387; Latino N:337/403 13
State s s Role In Getting Health Plans Involved. Articulate interests: Role in Shaping State & Federal Projects. Informing requirements (Medicare & Medicaid). Provide discussion forum. Provide basic demographic information. ID health plan champions. Keep solution focused. Recognize plan s s need to show results of QI efforts. Limit timeframe & spotlight success. Help identify community support and resources. Seek progress not perfection. http://www.chcs.org/publications3960/publications_show.htm?doc_id=250351 d=250351 14
Some Things To Consider When Planning QI Activities The importance of geography: Plan-wide measurement may underestimate disparities that are geographically based. Targeted vs. general population approaches: We provide the same high quality care to ALL our patients, regardless of race or ethnicity. Doing the same thing for everyone can worsen disparities. Some populations may be better able to take advantage of QI activities (education, information etc.) Bilingual, bicultural interventions for example may be more likely to engage Hispanic patients. Patient compliance assumes patients have truly equal access with no cultural or linguistic barriers. 15
Some Things To Consider When Planning QI Activities, Cont. Language and communication. Simple translation of English language educational material or health measures may be ineffective. Oregon s s Medicaid health status health risk assessment. CAHPS Partnering with community organizations Understand unique needs and issues of diverse populations may best be done through community organizations Community organizations may be more trusted by population of interest. 16
African American Health Coalition http://www.aahc-portland.org 17
The Informed, Activated Patient And The Prepared, Proactive Practice Team Patient approaches: Culturally appropriate education delivered by bilingual, bicultural providers; Culturally/Linguistically appropriate care coordination/case management; Community organizations. Provider Approaches: Patient reports. Support with socially complicated patients. Practice guidelines. 18