CPBS 7711: Electronic Health Records / Clinical Research Informatics Michael G. Kahn 6 December 2011
Dec 19, 2015
CPBS 7711: Electronic Health Records / Clinical Research Informatics
Michael G. Kahn6 December 2011
Various phenotypes
Topics
1. Translational research barriers
2. Sources of clinical data
3. The generation of clinical data
4. Administrative data sources
5. National clinical data sources
6. TCH clinical data sources
7. How can I exploit any of this stuff
Submission& ReportingEvidence-based
Review
NewResearchQuestions
StudySetupStudy Design
& Approval
Recruitment& Enrollment
StudyExecution
ClinicalPractice
PublicInformation
T1 Biomedical Research Investigator Initiated T1 T2 Translational ResearchIndustry Sponsored Commercialization
ClinicalTrial Data
BasicResearch Data
PilotStudies
RequiredData Sharing
OutcomesReporting
OutcomesResearch
Evidence-based Patient
Care and Policy
EMRData
A Lifecycle View of Clinical Research
Translational Phases
Westfall JM, Mold J, Fagnan L. Practice-based research – “Blue Highways” on the NIH roadmap. JAMA. 2007 Jan 24;297(4):403-6.
Translational Zones Example Beta-blockers and Myocardial Infarctions
Drolet BC, Lorenzi NM. Translational research: understanding the continuum from bench to bedside. Transl Res. 2011 Jan;157(1):1-5.
Setting the context: Translational Barriers
Bench Bedside /ClinicTranslational Barrier 1
Wide-spreadAppropriateUse in Standard Practice
Translational Barrier 215-20 years
New Terms: Translational Bioinformatics & Clinical Research Informatics
Sakar IN. Biomedical informatics and translational medicine. J Transl Med. 2010 Feb 26;8:22.
Topics
1. Translational research barriers
2. Sources of clinical data
3. The generation of clinical data
4. Administrative data sources
5. National clinical data sources
6. TCH clinical data sources
7. How can I exploit any of this stuff
The Clinical Data Landscape
Slide from Philip R.O. Payne, Ph.D. The Ohio State University Medical Center, Department of Biomedical Informatics
Slide from Philip R.O. Payne, Ph.D. The Ohio State University Medical Center, Department of Biomedical Informatics
Topics
1. Translational research barriers
2. Sources of clinical data
3. The generation of clinical data
4. Administrative data sources
5. National clinical data sources
6. TCH clinical data sources
7. How can I exploit any of this stuff
A Framework for Health Care Data Use
• Internal Data/Information– Patient Care
• Patient specific• Aggregate• Comparative
– General Operations
• External Data/Information– Comparative– Expert/Knowledge-based (Research)– Regulatory
Modified from: Wager KA, Lee FW Glaser JP. Managing Health Care Information Systems. A Practical Approach For Health Care Executives. Jossey-Bass San Francisco 2005.
Patient Encounter Data and InformationPrimary Purpose
Type Clinical Administrative
Patient-specific
Identification sheet (aka “Face Sheet”)Problem listMedication recordHistoryPhysicalProgress notesConsultationsPhysicians’ ordersImaging and X-ray resultsLab resultsImmunization recordOperative reportPathology reportDischarge summaryDiagnoses codesProcedure codes
Identification sheetConsentsAuthorizationsPreauthorization approvalsSchedulingAdmission or registrationInsurance eligibilityBillingDiagnoses codesProcedure codes
Aggregate Disease indexesSpecialized registriesOutcomes dataStatistical reportsTrend analysesAd hoc reports
Cost reportsClaims denial analysesStaffing analysesReferral analysesStatistical reportsTrend analysesAd hoc reports
Modified from: Wager KA, Lee FW Glaser JP. Managing Health Care Information Systems. A Practical Approach For Health Care Executives. Jossey-Bass San Francisco 2005.
Important documents you may not recognize
Problem list Significant active or dormant illnesses and operations. Items may be acute (this encounter only) or chronic (long-duration, chronic or intermittent). Includes entries from all care-givers.Includes both medical and non-medical issues
Medications list A list of all active medications that the patient has been prescribed and supposedly is taken. Patient compliance issues may result in significant deviation from the med list
Medication administration record (MAR)
Detailed record of every patient that the patient received or did not receive while under inpatient care. Reasons for not receiving a medication include: refused, away, NPO
H&P: History and Physical
A comprehensive review of the patients symptoms and signs as understood at the beginning of a treatment episode. Sections include: CC, HPI, PMH, PSH, FH, SH, ROS, PE, Assessment & Plan
Modified from: Wager KA, Lee FW Glaser JP. Managing Health Care Information Systems. A Practical Approach For Health Care Executives. Jossey-Bass San Francisco 2005.
Important documents you may not recognize
Progress Notes On-going reassessment and documentation during the course of treatment made by physicians, nurses, therapists, social workers and other clinical staff. Most popular documentation model used to be SOAP, now being replaced by APSO
Physician orders
Directions or prescriptions given to other members of the health care team regarding medications, tests, diets, treatments, etc.
Discharge summary
For an inpatient encounter, a summative account of the reason for admission, the significant findings from tests, procedures performs, therapies provided, response to treatment, condition at discharge and instructions for home care, including medications, activity, diet and follow-up care.
Modified from: Wager KA, Lee FW Glaser JP. Managing Health Care Information Systems. A Practical Approach For Health Care Executives. Jossey-Bass San Francisco 2005.
Data Creation Flow: Inpatient
Modified from: Wager KA, Lee FW Glaser JP. Managing Health Care Information Systems. A Practical Approach For Health Care Executives. Jossey-Bass San Francisco 2005.
Data Creation Flow: Inpatient
Modified from: Wager KA, Lee FW Glaser JP. Managing Health Care Information Systems. A Practical Approach For Health Care Executives. Jossey-Bass San Francisco 2005.
Data Creation Flow: Outpatient
Modified from: Wager KA, Lee FW Glaser JP. Managing Health Care Information Systems. A Practical Approach For Health Care Executives. Jossey-Bass San Francisco 2005.
Topics
1. Translational research barriers
2. Sources of clinical data
3. The generation of clinical data
4. Administrative data sources
5. National clinical data sources
6. TCH clinical data sources
7. How can I exploit any of this stuff
UB-04
Uniform billing form used by Medicare and adopted by most insurance companies
CMS-1500
Billing form for physician services
Topics
1. Translational research barriers
2. Sources of clinical data
3. The generation of clinical data
4. Administrative data sources
5. National clinical data sources
6. TCH clinical data sources
7. How can I exploit any of this stuff
SeDLAC: A Secondary Database Resource supported by the CCTSI Informatics Core
• Primary data resources are:– National Center for Health Statistics (
www.cdc.gov/nchs/) and – Agency for Healthcare Research and Quality (
www.ahrq.gov)
• Extensive collection of searchable databases:– freely available to all– replicated on SeDLAC servers
Available Databases
• NHIS: population-based interview– National Health Interview Survey
• NAMCS: outpatient provider-based– National Ambulatory Medical Care Survey
• NHAMCS: hospital urgent care-based– National Hospital Ambulatory Medical Care Survey
• MEPS: family-based, repeated measures– Medical Expenditure Panel Survey
• HCUP: inpatient-based– Health Care Utilization Project
Available Databases
• NSFG: women-based (expanding to include men)– National Survey on Family Growth
• BRFSS: population-based (telephone)– Behavioral Risk Factor Surveillance Survey
• NHANES:population-base interview and exam– National Health and Nutrition Examination Survey
• NHCHS: agency-based– National Home Care and Hospice Survey
Topics
1. Translational research barriers
2. Sources of clinical data
3. The generation of clinical data
4. Administrative data sources
5. National clinical data sources
6. TCH clinical data sources
7. How can I exploit any of this stuff
Children’s Hospital Clinical Analytics:Access to Clinical Databases
– Epic -- TCH only. DX, PX, medications/Rx, flow sheets.
– Colorado Hospital Association (CHA): administrative database of inpatient (only) for all Colorado hospitals. Updated quarterly
– CHCA PHIS – Pediatric Health Information Systems- an external database of comparisons 30+ free-standing Children's Hospitals. Updated quarterly.
– NACHRI -- National Association of Children's Hospitals and Related Institutions: more than 70 participating hospitals- Updated quarterly.
Available EPIC DataComprehensive Medical Record- Admit, transfer, discharge-MR#, Account #-Name, address, phone, zip code-Diagnosis, Procedure codes-DRG, MDC-ED transfers-Department(s)-Appointments-ProvidersOutcomes etc..
Demographics-Age, race, gender, county, etc
Utilization, claims & billing-Individual charges-Insurance billing-Insurance payment
Clinical Documentation-Vital signs-Allergies-Detailed flowsheet data-Med orders, med admin, Med Rx- Procedure orders / notes-Physician, nursing, ancillary notes-Laboratory, Microbiology, Radiology, Pathology Results
Behind The Scenes… The Emergency Dept ERD
Examples of Research Participation
•Elaine Morrato•“Erickson M, Miller N, Kempe A, Morrato EH, Benefield E. Benton K. Variability in Spinal Surgery Outcomes among Children s Hospitals in the US”. Presented at the 2009 American Academy of Orthopedic Surgeons (AAOS) Annual Meeting, Las Vegas, NV, February 25-28, 2009.
•Morrato EH, Erickson M, Beaty B; Benton K; Benefield E, Kempe, A. ”Variability in surgical outcomes for spinal fusion surgery in U.S. children’s hospitals”. •Presentation presented at the Health Services Epidemiology Spotlight Session at the Society for Epidemiology Research Conference, June 24-27, 2008, Chicago, IL. Am J Epidemol. 2008; 167(Suppl): S40.• Information CI group provided:
– Patient list for specific surgical procedures (laproscopic vs open)– Diagnosis– Demographics– Outcomes
Examples of Research Participation
•Peter Mourani: “Outcomes of Premature Infants Admitted to PICU with Acute Respiratory Disease”. • Information CI group provided:
– Medication orders, Lab test results– DX codes, ADT events– Outcomes, Analysis
•Sarena Teng: “Retrospective Review of Propofol as Bridge to Extubation in Pediatric Post-operative Cardiac Patients” • Information CI group provided:
– List ICU patients on mechanical ventilation, propofol– Medication review– Patient outcomes
Examples of Research Participation
• Marion Sills: “Emergency Department Overcrowding and Quality of Care for Children”. Information CI group provided:– Medication orders – Lab test results– DX codes– ADT events
• Molli Pietras: “Evaluation of Prolonged Precedex Infusion in Critically Ill Infants and Children”
• Information CI group provided:– List of qualifying patients– Flowsheet data– Medications – Outcomes
PHIS Data Overview
1
INPATIENT Emergency Department
AmbulatorySurgery
ObservationUnit
PHIS Data Repository
Medical Records
System
Billing
Systems
All data submitted electronically (no manual entry) on a quarterly
basis
PHIS By The Numbers*
• Participating Hospitals: 40• Inpatient Cases: 2.2 million• Inpatient Days: 13.1 million• ED encounters: 6.7 million • Total Charges: $90.7 billion• Total ICD-9 Codes: 33.6 million• Pharmacy Transactions: 116.8 million• Physicians: 297,250
* Since 2002, does not include available archived data back to 1992
Seattle
Oakland
Palo Alto
Madera
Los Angeles
Orange
San Diego
Phoenix
Denver
Dallas
Fort Worth
Corpus Christi Houston
Little Rock
New Orleans
Birmingham
Memphis
Nashville
Atlanta
St. Petersburg
Miami
Boston
Hartford
New York
Philadelphia
DC
Norfolk
Pittsburgh
Dayton
Columbus
Cincinnati
Akron
Buffalo
Milwaukee
Chicago
St. Louis
Detroit
Indianapolis
Minnesota
Omaha
Kansas City
PHIS Patient Abstract
• Demographics– Gender– Birthweight (gms)– DOB– Pediatric Age Group– AAP Age Code– Age (based on age at
admission)• Age in Years• Age in Months (if less than
2 yrs)• Age in Days (if less than 30
days)– Race/Ethnicity
• Episode of Care– LOS– Admit Date/Month/Year– Discharge Date/Month/Year– Infection Flag– Surgical and Medical
Complication Flags– Disposition– Pre-Op LOS– Post-Op LOS
PHIS Patient Abstract
• Physician Profiles– Attending Physician– Attending Physician Sub-specialty– Principal Px Physician– Principal Px Physician Sub-specialty
• Dx/Px Profiles– Principal Dx– Principal Px
• Clinical Classification (Groupers)– Major Diagnostic Category
(MDC)– CMS (HCFA) DRG– APRDRG
• Version 15• Version 20• Version 24
Topics
1. Translational research barriers
2. Sources of clinical data
3. The generation of clinical data
4. Administrative data sources
5. National clinical data sources
6. TCH clinical data sources
7. How can I exploit any of this stuff
Kohane Nat Rev Genet 2011 Jun 12(6) 417-28
Kohane Nat Rev Genet 2011 Jun 12(6) 417-28
URL:
www.gwas.net
QRS duration
Dementia
Peripheral vascular disease
Cataracts Type II diabetes
Coordinating center
RFA HG-07-005:Genome-Wide Studies in Biorepositories with
Electronic Medical Record Data
• 2007 NIH Request for Applications from the National Human Genome Research Institute
“The purpose of this funding opportunity is to provide support for investigative groups affiliated with existing biorepositories to develop necessary methods and procedures for, and then to perform, if feasible, genome-wide studies in participants with phenotypes and environmental exposures derived from electronic medical records, with the aim of widespread sharing of the resulting individual genotype-phenotype data to accelerate the discovery of genes related to complex diseases.” (Emphasis added)
EMR-based Phenotype Algorithms
• Typical components– Billing and diagnoses codes– Procedure codes– Labs– Medications– Phenotype-specific co-variates (e.g., Demographics,
Vitals, Smoking Status, CASI scores)– Pathology– Imaging?
• Organized into inclusion and exclusion criteria
EMR-based Phenotype Algorithms
• Iteratively refine case definitions through partial manual review to achieve ~PPV ≥ 95%
• For controls, exclude all potentially overlapping syndromes and possible matches; iteratively refine such that ~NPV ≥ 98%
Phenotype Reuse
T2DM
Diabetic Retinopathy
Primary PhenotypesSite Phenotype Validation
(PPV/NPV)Group Health Dementia 73% / 92%
Marshfield Clinic
Cataracts / Low HDL 98% / 98%82% / 96%
Mayo Clinic PAD 94% / 99%
Northwestern University
Type 2 DM 98% / 100%
Vanderbilty University
QRS Duration 97% / 100%
www.gwas.net
Opportunities for CPBS Collaborations
• NLP/Text Mining electronic records• Novel phenotyping classification algorithms• (Limited) access to genotypes
– Disease-specific– Study-specific– Investigator-specific
CPBS 7711: Electronic Health Records / Clinical Research Informatics
Michael G. Kahn6 December 2011