Associations Betw een Practice- Reported Medical Homeness and - PowerPoint PPT Presentation

Associations Betw een Practice- Reported Medical Homeness and Health Care Utilization Among Publicly Insured Children Presentation at the AcademyHealth Annual Research Meeting Minneapolis, MN June 16, 2015 Anna L. Christensen, PhD • Joseph S. Zickafoose, MD, MS • Brenda Natzke, MPP • Stacey McMorrow, PhD • Henry T. Ireys, PhD

Citation • Publis hed in May 2015 volume of Academic Pediatrics : – Chris tens en AL, Zickafoos e J S , Natzke B , McMorrow S , Ireys HT. As s ociations between practice-reported medical homenes s and health care utilization among publicly ins ured children. Academic Pediatrics . 2015; 15: 267–274. 2

Background • C hildren’s Health Ins urance Program R eauthorization Act (C HIPR A) Quality Demons tration Grant Program – $100 million to improve health care for children – 10 awardees (18 s tates ), 5-year grants s tarting in 2010 – 52 total projects – National evaluation overs een by the Agency for Healthcare R es earch and Quality (AHR Q) • 12 s tates with patient-centered medical home (PC MH) projects 3

Background & Research Question • R elations hip between “medical homenes s ” and children’s health care utilization – R es ults vary by s tudy, outcome (preventive care, E D vis its , hos pitalizations ), and population (general population vs children with chronic conditions ) – Mos t s tudies as s es s parent-reported medical homenes s – Two s tudies of practice-reported medical homenes s s how mixed res ults (Cooley 2009, Paus tian 2013) • Is the “medical homenes s ” of primary care practices as s ociated with health care utilization by publicly ins ured children? 4

Methods • C ros s -s ectional bas eline analys is – 3 s tates : IL, NC, S C – 64 practices (IL = 32, NC = 18, S C = 14) • C hildren (birth – 18 y) in Medicaid – Fee-for-s ervice or primary care cas e-management – E xclus ions : >1-month gap in coverage, partial benefits , waiver program, other ins urance, ins titutionalization • Attribution of children to practices – Majority of well-child vis its – If no majority of well-child vis its , majority of other vis its 5

Methods: Measures • Practice-reported “medical homenes s ” – National Committee for Quality As s urance (NCQA) 2011 medical home s elf-as s es s ment: IL – Medical Home Index (MHI): NC – Medical Home Index- R evis ed S hort Form (MHI-R S F): S C – Tertiles : low, medium, high • Utilization (prior 12 mo.) – WCV: ≥75% of recommended # of well-child vis its – E DV: any non-urgent, potentially avoidable emergency department vis it (NYU algorithm; B en-Is aac 2010) 6

Methods: Analysis • Multi-level logis tic regres s ion – S eparate models for IL and NC/S C • C ovariates – Child-level: age, race/ethnicity, chronic condition/dis ability • Pediatric Medical Complexity Algorithm (Simon, et.al. 2014) • Medicaid eligibility based on disability – Practice-level (NC/S C only): urban/rural, # of providers • S ens itivity tes ts – R e-es timated models with medical homenes s as : • Continuous variable Categorical variable with cut points at 25 th and 75 th percentile • – Inferences did not change 7

Child Characteristics IL (n = 33,895) NC/SC (n = 57,553) Age group, % 0 to 5 years 53 57 6 to 12 years 31 30 13 to 18 years 16 14 Race/ethnicity, % black 45 33 white 31 45 other 24 22 Chronic condition or disability, % 31 34 8

Results: Medical Homeness & Well-Child Visits 100 Regression Adjusted Percent 90 80 70 60 Low MH Medium MH 50 High MH 40 30 20 10 0 IL (NCQA) NC/SC (MHI-RSF) 9

Results: Medical Homeness & Well-Child Visits 100 Regression Adjusted Percent 90 77 76 80 74 70 60 Low MH Medium MH 50 High MH 40 30 20 10 0 IL (NCQA) NC/SC (MHI-RSF) 10

Results: Medical Homeness & Well-Child Visits 100 Regression Adjusted Percent 90 77 76 80 74 69 69 70 63 60 Low MH Medium MH 50 High MH 40 30 20 10 0 IL (NCQA) NC/SC (MHI-RSF) 11

Medical Homeness & Non-Urgent ED Visits 100 Regression Adjusted Percent 90 80 70 * OR = 0.65 (95% CI 0.47-0.92) 60 Low MH Medium MH 50 High MH 40 29 26 30 23 * 20 10 0 IL (NCQA) NC/SC (MHI-RSF) 12

Medical Homeness & Non-Urgent ED Visits 100 Regression Adjusted Percent 90 80 70 60 Low MH Medium MH 50 High MH 40 29 29 29 27 26 30 23 * 20 10 0 IL (NCQA) NC/SC (MHI-RSF) 13

Conclusions • Medical homenes s was not as s ociated with well-child vis its • Higher medical homenes s was as s ociated with fewer non-urgent E D vis its , but only in IL where NC QA medical home s elf-as s es s ment meas ure was us ed • Limitations – Cros s -s ectional – May not be repres entative of Medicaid managed care – Could only attribute children with s ome s ervice us e – Different meas ures vs . different s tates 14

Implications • Meas uring medical homenes s – No s ingle bes t meas ure – Different meas ures capture different proces s es – Differences in definitions and meas ures of medical homenes s may contribute to mixed findings in current literature – Cons ider us ing more than one meas ure 15

For More Information • National E valuation of the C HIPR A Quality Demons tration Grant Program http://www.ahrq.gov/policymakers /chipra/demoeval/index.html • Anna C hris tens en achris tens en@ mathematica-mpr.com • Henry Ireys , Project Director hireys @ mathematica-mpr.com • Acknowledgements : C indy B rach & Linda B ergofs ky (AHR Q), C arl C ooley (C rotched Mountain), J eanne McAllister (Indiana), S arah S cholle (NC QA), Genevieve K enney (Urban Ins titute), C atherine McLaughlin (Mathematica) 16

Extra Slides 17

CHIPRA Quality Demonstration Program Focus: Five Broad Strate gies to Improve Quality • Use CMS ’ core pediatric quality measures (Category A) • Promote Health Information Technology/E lectronic Health R ecords (Category B) • Implement provider-based models (Category C) • Apply model pediatric E HR format (Category D) • Other innovative approaches (Category E ) 18

Demonstration Grantees* and Partnering States, by Grant Category States A B C D E Oregon* x x x Alaska x x x West Virginia x x x Maryland* x x Georgia x x Wyoming x x x Utah* x x x Idaho x x x Florida* x x x x Illinois x x x x Maine* x x x Vermont x x x Colorado* x x New Mexico x x Massachusetts* x x x South Carolina* x x x Pennsylvania* x x x North Carolina* x x x 19

Results: Medical Homeness & Well-Child Visits 100 90 78 80 75 74 Unadjusted Percent 66 70 64 64 60 Low MH Medium MH 50 High MH 40 30 20 10 0 IL (NCQA) NC/SC (MHI-RSF) 20

Medical Homeness & Non-Urgent ED Visits 100 90 80 Unadjusted Percent 70 60 Low MH Medium MH 50 High MH 40 31 31 30 29 27 30 25 20 10 0 IL (NCQA) NC/SC (MHI-RSF) 21