Digital Data Priorities for Continuous Learning in Health and Health Care An Institute of Medicine Workshop: Sponsored by the Office of the National Coordinator for Health Information Technology Supporting Public Health and Surveillance State Level Perspective Marty LaVenture, MPH, PhD Director, Minnesota e-health Initiative and Office of Health Information Technology, Minnesota Department of Health March 23, 2012
Topics for Discussion National State and Local Public Health Context Minnesota e-health Data Quality Framework Public Health Examples of Challenges and Opportunities Considerations and Recommendations 2
USA Context: State Public Health Agencies States: 56 Independent legal entities 50 State Health Departments, 6 Territories City/County: ~ 3,000 local health departments Statutory responsibility to secure, analyze, distribute authoritative information/knowledge & take action to protect the public Limited or no: PH-EHR specifications; or software certification and few vendors Value & need for high quality structured data is documented
EHR Based Data is One Part of Data Needs for Public Health Practice Select MDH Information Systems # Acute disease 8 Maternal and child health 4 Chronic disease 3 Injury 2 Vital statistics 1 Laboratory information 2 management system Total 20
Example Surveillance for Infectious Diseases at MDH Reportable Disease Surveillance Public Influenza & Respiratory Surveillance - SARS - Veterinary/Animal Health & Vectors West Nile Virus Unexplained Deaths Source: Minnesota Department of Health
Public Health Data to Inform Practice and Support Healthier Communities 6
Adoption of Electronic Health Records (EHRs) and Related Health Information Technology (HIT) in Minnesota
MN Ambulatory Clinics Electronically Exchanging Health Information with Partners (N = 750)* 8
Framework for Data Quality in Context of Public Health Timeliness Currency Completeness Data Quality Comprehensive Accuracy Accessibility Needed key attributes for data quality depend on the context of public health activity: Examples Comprehensiveness and completeness for immunization registries Timeliness for new born screening, acute disease surveillance, outbreaks Accuracy for cancer surveillance Access to primary source of data and currency for certain public health programs based on data
Timeliness and Immunization (MIIC) Electronic Data Feeds (Capacity & Capability) Timeliness as good as it can get As the momentum around electronic exchange and standards increases, and as better transport options are adopted, more organizations are expected to move towards real-time
Data Quality in Immunization Registry (MIIC) Current Quality Assurance Efforts Emphasis on completeness and accuracy of data Need to ensure that MIIC gets - All vaccines entered on all patients - Entire immunization history - Historical doses given at different providers Ongoing Quality Assurance efforts - Every new file has to pass QA process - Monthly improbable shots reports - Provider activity report monitored by MDH and MIIC Regional Coordinators - With move to electronic reporting and standards, HL7 files undergo both format testing and content testing - Constant monitoring for duplicates, systematic data input errors etc. Policies/procedures in place - Rejection of unnamed records (vital loads with placeholders like Baby Boy, Baby Girl) - Requiring reporting/data break-down by individual clinic site for data loads (for reminder/re-call purposes) - Discontinued data loads from select providers submitting claims data (mismatch of dates & increased duplicates)
Quality of Data Submitted for Public Health Completeness of automated electronic laboratory reporting: Fields identifying the patient and test results were usually complete ( > 95%) Fields containing patient demographics, patient contact information and provider contacting information were suboptimal (~20% - ~65%) The data was heterogeneous in their completeness by source Source: Dixon BE, McGowan JJ, et al: Electronic Laboratory Data Quality and the Value of a Health Information Exchange to Support Public Health Reporting Processes, AMIA Annu Symp Proc. 2011;2011:322-30. Epub 2011 Oct 22. 12
Data Quality in Context of Immunization Demographic Data for Identification and De-duplication Required: E.g. First name, Last name and date of birth Desired to increase quality of the match E.g. Mother s Birth Name Essential for matching Vaccine Information Date vaccine administered Vaccine manufacturer Vaccine name Essential for forecasting
Data Quality in Immunization Registry (MIIC) Issues to Address Common Data Quality Problems Miscoding of vaccines in EHR Mismapping of vaccines from EHR to extract file Not sending historical data Not sending vaccine lot number and manufacturer Sending ordered versus administered shots Incomplete, invalid, addresses affecting reminder/recall efforts Value Suffers When Data Quality Perception Suffers Not my problem, it must be MIIC! - Issue of Trust / Confidence Leads to use of MIIC for look-up only Lack of interest in advanced MIIC reports & forecasting
Electronic EHR Data Quality Needs Rigorous Quality Thresholds Public Health Lab E.g. Newborns Screening, Reference lab specimens Acute Disease Surveillance Vital Records Birth, Death, Fetal death Cancer Surveillance Systems Other Individually named-based systems Variable Quality Thresholds and Tolerances Syndromic Surveillance
Considerations for Advancement of Quality in Digital Exchange Context Technical Considerations More robust Electronic Health Record (EHR) certification process Ensuring that certified products are able to create standardized file outputs Creation of certification process for products in settings like public health Better alignment of NIST test cases with corresponding implementation guides Create comprehensive test beds that can test both structure and content Prompts closer to point of data capture to ensure completeness of data Improve capture of needed data at point of care Race and ethnicity, mother s birth name are required elements of IIS, but not sent Adopt message (e.g. HL7) testing tools that will make it easier to test and fix errors and check for completeness on part of providers, vendors and receiving entities
Considerations for Advancement of Quality in Digital Exchange Context (contd..) Policy/Programmatic Considerations Increase education/outreach on value of data quality Create incentives for data use to increase value / quality Enable bi-directional flow of data to reap benefits of information Assure adequate vendor support for smaller practices, technically challenged settings with less resources Foster financial incentives that have a data quality focus
Considerations for Advancement of Quality in Digital Exchange Context (contd..) Policy/Programmatic Considerations - continued Incentives / support to modernize State and Local Systems Establish / Adopt common business requirements for Public Health Establish Public Health software certification program Foster greater vendor engagement in public health applications Establish Standard data quality parameters and thresholds Improve feedback via Provider / Data Quality Dashboards to monitor their own data E.g. Types of vaccines by age groups, Timeliness, Rates Enhance value of reports and feedback E.g. MIIC Patient Follow Up (re-vamped reminder/recall)
Final Perspective: Source Collection, Knowledge Discovery, Partnerships, HIT Opportunity MMWR Electronic Surveillance Project-1984 In 1984, CDC, in cooperation with the CSTE and epidemiologists in six states began...the Project. The goal to demonstrate the effectiveness of computer transmission of public health surveillance data between state health departments and CDC.) http://www.cdc.gov/ncphi/disss/nndss/netss.htm
Acknowledgements and Resources Thanks to Bill Brand, Director of Programs, Public Health Informatics Institute (PHII) Emily Emerson, Program Manager, Minnesota Immunization Information Connection (MIIC), MN Dept. of Health Kari Guida, Assessment and Evaluation Coordinator, Office of Health Information Technology, MN Dept. of Health Priya Rajamani, Senior Health Informatician, MDH Office of Health IT (OHIT) and MIIC Programs, MN Dept. of Health Laël Gatewood, Professor, Health Informatics, University of Minnesota For Additional info Contact Marty LaVenture, Director of MN e-health and MDH Office of Health IT Martin.LaVenture@state.mn.us, (651) 201-5950 Minnesota e-health Initiative: http://www.health.state.mn.us/ehealth/