Disparities in Childrens Health Care I

Using Health Information Technology in Primary Care to Mitigate Healthcare Disparities Evaline A. Alessandrini, MD, MSCE Department of Pediatrics University of Pennsylvania School of Medicine iTACH The Children’s Hospital of Philadelphia 1 iTACH • Institute to Transform and Advance Children’s Healthcare • The “i” Stands for Institute and Information • The Goal of iTACH is To Improve the Health and Save the Lives of Children Using Information to Transform Pediatric Healthcare • iTACH is ‐ Multi-disciplinary ‐ Highly collaborative ‐ Focused on improvement and value creation ‐ Able to mitigate health disparities 2 iTACH Objectives 1. Improve data sources 2. Enable performance improvement with better measurement 3. Conduct clinical & business transformation projects that promote quality and value in health care 4. Foster research 5. Influence child health policy Leading to a reduction in disparities 3 Conceptual Framework Health Information Technology: Mitigating Health Disparities Health Status Non-Healthcare Determinants of Health Behaviors & Lifestyle Host Factor, e.g., Race/ Ethnicity Socioeconomic Status Environment Healthcare System Performance Quality Access Resource Use Adoption of Health Information Technology Physician Practice Patterns Knowledge Attitudes Intention Behavior 4 The Impact of Clinical Alerts in an Electronic Health Record on Routine Childhood Immunization in an Urban Pediatric Population Alexander G. Fiks, MD Robert W. Grundmeier, MD Lisa M. Biggs, MD A. Russell Localio, PhD Evaline A. Alessandrini, MD, MSCE Ambulatory Pediatric Association Young Investigator Award In press, Pediatrics 5 Study Objective To test the hypothesis that using health information technology to create a clinical decision support tool for routine pediatric vaccination will ‐ Improve overall immunization rates for children at 24 months of age ‐ Increase the proportion of immunization opportunities captured at health care visits ‐ Reduce or eliminate immunization disparities for racial minorities 6 Background Childhood Immunizations 7 Background Health Information Technology Kaushal, et. al. Health Affairs, 2005 8 Methods: Design and Patients Design • One-year decision support intervention trial with historical controls Patients • Children reaching 24 months of age with at least one health care visit during the control or intervention period Visits • All sick or well office visits for children < 24 months of age with at least one vaccine due 9 Methods: Study Setting The Children’s Hospital of Philadelphia PeRC - Pediatric Research Consortium AHRQ-funded Practice-Based Research Network 29 Urban, Suburban, Rural Kids First Practices 4 Urban Primary Care Centers > 72,000 visits annually > 75% publicly insured > 80% racial minorities EpicCare since 2002 10 Methods: Intervention Intervention period • Electronic health record clinical reminders for vaccinations due appeared prominently whenever a patient encounter opened • Included direct links to order sets • Intervention accompanied by education Control period • No alert, just review of vaccine record • No link to order entry • Intermittent education 11 Health IT Intervention 12 Methods: Outcomes Patient-level • Up-to-date immunizations at 24 months ‐ 4 DTP; 3 PV; 1 MMR; 3HiB; 3 HepB; 1 varicella Visit-level • Proportion of sick or well visits with ‐ Any due immunization given ‐ All due immunizations given 13 Methods: Analyses Patient-level • Risk differences in up-to-date immunizations status pre and post intervention ‐ Marginal standardization using SUDAAN software ‐ Bias corrected confidence intervals • 999 bootstrap resamplings ‐ Accounted for clustering within clinicians Visit-level • Risk differences in captured immunization opportunities pre and post intervention 14 Results: Patient-level 3217 patients (1669 intervention and 1548 control) study participants by race other 9% white 8% West Philadelphia population by race Other 8% White 16% black 83% Black 76% 2000 census 15 Results: Patient-level 2 year Immunization Status 92 % up-to-date at 24 months 90 88 86 84 82 80 78 76 74 pre-intervention post-intervention total white black other Risk Difference 8.4% (5.8, 10.6) 16 Results: Visit-level Control period: 19,909 visits with vaccine due Intervention period: 15, 928 visits with alerts sick visits by race well visits by race other 12% white 11% other 9% white 7% black 77% black 84% 17 Results: Visit-level Rates increase from 11% pre-intervention to 32% post-intervention for immunization at sick visits Education 100 90 80 70 60 50 40 30 20 10 0 Sept '03 Nov Jan Mar May Jul Sept '04 Nov Jan Mar Intervention Percent of Immunization Opportunites Captured Well Visits Sick Visits May Jul Month of Study 18 Results: Well Child Visits % with any immunizations given 95 90 85 80 75 70 65 pre-intervention post-intervention total white black other Risk Difference 12.2% (11.2, 13.1) 19 Results: Sick Visits 40 % with any immunizations given 35 30 25 20 15 10 5 0 pre-intervention post-intervention total white black other Risk Difference 20.7% (19.3, 22.1) 20 Limitations • Historical Controls ‐ Changing trends in immunization practice ‐ Differences in vaccine shortages • Captured immunization opportunities did not reach 100% ‐ Incomplete information on why vaccines not given • Unknown reasons for increased immunization at sick visits for black children • Success relies upon access to medical care • Categorization of race / ethnicity 21 Conclusions The Health IT decision support intervention • Increased up-to-date immunization status at 2 years of age above the Healthy People 2010 goal of 90% ‐ Reduced racial disparities in immunization status • Increased captured immunization opportunities ‐ 10% at well visits ‐ 3 fold at sick visits ‐ Greatest absolute increases for racial minorities at sick visits 22 Ongoing Health IT Projects Cluster randomized controlled studies • Influenza immunization for children > 5 years of age with asthma • Quality and value of otitis media care • Use of asthma care plans and control tools 23 Significance • Intervention to improve quality for all children resulted in a reduction in racial disparities in immunization status at 24 months • After initial investment, the success of the health IT intervention requires little ongoing investment for more easily sustained and disseminated health improvements • Potential inequity in availability of health information technology for all children 24 25 Knowledge-Based Pediatrics • Data integration and synthesis of multi-level data (genes---community) • Clinicians and managers continuously “learn” using clinical and business data • The right knowledge presented to clinicians/patients at the right time • Personalized, predictive, and preventive • Diagnostic aids • Knowledge management and clinical decision support 26 Knowledge-Based Pediatric Healthcare 1. Data Collection: The Biopsychosocial Profile 2. Data Storage and Integration 3. Clinical/Business Analytics 4. Knowledge Deployment 5. Performance and Outcome Evaluation 27 KBPeds--Data Collection: Creating a Biopsychosocial Profile Genomics, Gene Expression, Biomarkers Body Structures and Physiological Functions Community: Physical and Social Environment Conditions: Symptoms and Disorders Self-Assessed Well-Being Family and Relationships Preferences Behavior 28 2. Data Storage and Integration & 3. Clinical/Business Analytics & 4. Knowledge Deployment 2 External Knowledge Databases PubMed ClinicalTrials.gov OMIM Genome Clinical Tracking Systems CardioIMS Neonatology Db Oncology Db Patient-Reported Metrics HRQoL Social Complexity Biomarkers Genomic Profile Proteome Profile Metabolic Profile Knowledge Management/Clinical Decision Support  Improved Diagnosis, Treatment, and Referral Data Storage & Integration APPLICATIONS Epidemiology Biomedical Informatics Health Services & Outcomes Research Clinical Questions Pharmacogenomics Predictive Modeling, Data Mining, Pattern Recognition “Self Service” Portal CHOPlink Pediatric Data Trust Service Pathways New Nosologies & Treatment Pathways Public Reporting Medical Education 3 EHR 4 Children Performance, Quality, and Outcome Metrics Performance Improvement Strategy Management 12 29 Knowledge Deployment • Occurs Where Children Are • 1 Child = 1 Record • Child-centric, which catalyzes collaboration among providers 30 Knowledge-Based Pediatrics Otitis Media – A Collaboration Between Medicine & Surgery • Primary Care/ENT Shared Record • Cluster Services into “Episodes-of-Care” • Clinical Decision Support for Medication Use and Referral • Quality and Cost Feedback • Training 31 well % with any immunizations given delay 35 30 25 total 20 15 10 5 0 pre-intervention post-intervention white black other 32 well 90 % with all immunizations given 80 70 60 50 40 30 20 10 0 pre-intervention post-intervention total white black other 33 sick 35 % with all immunizations given 30 25 total 20 15 10 5 0 pre-intervention post-intervention white black other 34 sick % with any immunizations given delay 18 16 14 12 10 8 6 4 2 0 pre-intervention post-intervention total white black other 35 Percent 70 60 50 40 30 20 10 0 White Black Race Characteristics of PeRC Race Asian or Pacific Islander American Indian or Alaska Native Missing/other PeRC National 36