Cleveland Clinic
1300 bed main hospital 9 Regional Hospitals 54,000 admissions, 2 million visits Group practice of 2700 salaried physicians and scientists
3000+ research projects Innovative Medical School 30 spin off companies Office of Patient Experience
Lethal Lag Time
It takes an average of 17 years to implement clinical research results into daily practice
Unacceptable to patients
Can Electronic Medical Records and Clinical Decision Support Systems change this?
Electronic Medical Records
Comprehensive medical information
Images Communication with other physicians, medical professionals
Communication with patients
3 million active patients, 10 years
EMR Inputs and Outputs
Inputs • Clinical • Labs • Devices • Remote monitoring • Pt outcomes • Omics • Social media?
EMR Tools • Alerts • Best practices • Smart sets • Workflow • Communication to other providers, patients
Outputs Secondary Use • Data sets • Registries • Quality reports
Clinical Workflow
Workflow
Clinical Decision Support
Process for enhancing health-‐related decisions and actions with pertinent, organized clinical knowledge and patient information
to improve health and healthcare delivery. Information recipients can include patients, clinicians and others involved in patient care delivery http://www.himss.org/ASP/topics_clinicalDecision.asp
Like a GPS, CDS supplies information tailored to the current
situation, and organized for maximum value.
Diagnostic Cockpit
CDS Example: Order Sets
CDS as a Strategic Tool
• CDS should be used as a strategic tool for achieving an organization’s priority care delivery objectives.
• These objectives are driven by external forces such as • payment models • regulations related to improving care quality and safety • internal needs for improving quality and patient safety • reducing medical errors • increasing efficiency
EMR Alert Types Clinical Decision Support
Target Area of Care Example Preventive care Immunization, screening, disease management
guidelines for secondary prevention
Diagnosis Suggestions for possible diagnoses that match a patient’s signs and symptoms
Planning or implementing treatment
Treatment guidelines for specific diagnoses, drug dosage recommendations, alerts for drug-drug interactions
Followup management Corollary orders, reminders for drug adverse event monitoring
Hospital, provider efficiency Care plans to minimize length of stay, order sets Cost reductions and improved patient convenience
Duplicate testing alerts, drug formulary guidelines
Clinical Decision Support Examples
New diagnosis of Rheumatoid Arthritis, prompted to refer to preventive cardiology
Clinical Decision Support Examples
Age > 50 and a fragile fracture diagnosis – order set for bone density scan and appropriate medication regimen
Go to Smart Set
Clinical Decision Support Examples
Solid organ transplant – chemoprevention for skin cancer
The CDS Toolbox (more examples) Drug-‐Drug Interactions
Drug-‐Allergy interactions
Dose Range Checking
Standardized evidence based ordersets
Links to knowledge references
Links to local policies
Rules to meet strategic objectives (core measures, antibiotic usage, blood management)
Documentation templates
Relevant data displays
Point of care reference information (i.e. InfoButtons)
Web based reference information
Diagnostic decision support tools
Virtuous Cycle of Clinical Decision Support
Measure
Guideline
CDS
Practice
Registry
http://www2.eerp.usp.br/Nepien/DisponibilizarArquivos/tomada_de_decis%C3%A3o.pdf
EMRs and Quality of Care
EMR and Quality of Care
Diabetes care was 35.1 percentage points higher at EHR sites than at paper-‐based sites
Standards for outcomes was 15.2 percentage points higher
Across all insurance types, EHR sites were associated with significantly higher achievement of care and outcome standards and greater improvement in diabetes care
Better Health Greater Cleveland
Meaningful Use
The Role of Registries
EMR data available to create a registry for any condition
Study the condition – progression, treatments
Comparative effectiveness of treatments Recruit for clinical trials Develop clinical decision support
Chronic Kidney Disease Registry
Chronic Kidney Disease Registry Established 2009 60,000 patients from the health system Cohort – Adults with two eGFRs less than 60 within 3 months, outpatient results only, or diagnosis of CKD
http://www.chrp.org/pdf/HSR_12022011_Slides.pdf
Validation Results
Our dataset’s agreement with EHR-‐extracted data for documentation of the presence and absence of comorbid conditions, ranged from substantial to near perfect agreement.
Hypertension and coronary artery disease were exceptions
EMR data accurate for research use
Registry Results
2011 5 out of 5 abstracts accepted to American Society of Nephrology annual meeting
Three papers accepted to nephrology journals
NIH grant Partnerships with other research centers
Pediatric Surgical Site Infection Data from the EMR and the operative record When did antibiotics start? Was pre-‐op skin prep done? Was the time-‐out and checklist observed in the OR
Post-‐op care quality
Patient Reported Outcomes
Understanding the outcomes of treatment incomplete without
Patient Reported Outcomes Measurement Information System http://www.nihpromis.org/
Patient-‐Centered Outcomes Research Institute http://www.pcori.org/
Patient Reported Outcomes
Quality of life Activities of daily living Recording weight, diet, exercise using apps Quantified Self
Population Health
New tools to enable the study of disease trends and epidemics
PopHealth -‐ submission of quality measures to public health organizations http://projectpophealth.org
Query Health – standards to enable Distributed Health Queries http://wiki.siframework.org/Query+Health
Predictive Models
Predicting 6-‐Year Mortality Risk in Patients With Type 2 Diabetes
Cohort of 33,067 patients with type 2 diabetes identified in the Cleveland EMR
Prediction tool created in this study was accurate in predicting 6-‐year mortality risk among patients with type 2 diabetes
Diabetes Care December 2008, vol. 31 no. 12: 2301-‐2306
Postoperative nomogram based on 996 patients treated at The Methodist Hospital, Houston, TX, for predicting PSA recurrence after radical prostatectomy.
Kattan M W et al. JCO 1999;17:1499-1499
©1999 by American Society of Clinical Oncology
Nomograms bring into visual perspective the effect exerted by continuous variables against measured end points
Risk Calculators Type 2 Diabetes Predicting 6-‐Year Mortality Risk
Algorithms
clevelandclinicmeded.com/ medicalpubs/micu/
Against Diagnosis
The act of diagnosis requires that patients be placed in a binary category of either having or not having a certain disease.
These cut-‐points do not adequately reflect disease biology, may inappropriately treat patients
Risk prediction as an alternative to diagnosis Patient risk factors (blood pressure, age) are
combined into a single statistical model (risk for a cardiovascular event within 10 years) and the results are used in shared decision making about possible treatments.
Annals of Internal Medicine, August 5, 2008vol. 149 no. 3 200-‐203
Information Overload
New information in the medical literature PubMed adding over 670,000 new entries per year
Information about an individual patient Lab results Vitals Imaging Genomics
Personalized Medicine
The boundaries are fading between basic research and the clinical applications of systems biology and proteomics
New therapeutic models Journal of Proteome Research Vol. 3, No. 2, 2004, 179-‐196.
Example–Parkinson’s Disease
New Cleveland Clinic partnership with 23andMe to collect DNA from Parkinson’s patients
Looking for Genome Wide Associations (GWAS)
23andme.com/pd/
Precision Medicine
”state-‐of-‐the-‐art molecular profiling to create diagnostic, prognostic, and therapeutic strategies precisely tailored to each patient's requirements.”
”The success of precision medicine will depend on establishing frameworks for …interpreting the influx of information that can keep pace with rapid scientific developments.”
N Engl J Med 2012; 366:489-‐491, 2/ 9/2012
Artificial Intelligence in Medicine Developing a search engine that
will scan thousands of medical records to turn up documents related to patient queries.
Learn based on how it is used “We are not contemplating ―
unless this were an unbelievably fantastic success ― letting a machine practice medicine.”
http://www.health2news.com/2012/02/10/the-‐national-‐library-‐of-‐medicine-‐explores-‐a-‐i/
IBM Watson
Medical records, texts, journals and research documents are all written in natural language – a language that computers traditionally struggle to understand. A system that instantly delivers a single, precise answer from these documents could transform the healthcare industry.
“This is no longer a game” http://tinyurl.com/3b8y8os
Digital Humans
Convergence of: Genomics Social media mHealth Rebooting Clinical Trials
Conclusion -‐ 1
EMR as the platform for the future of medicine
Data incoming Clinical Patient Reported Genomic Proteomic Home monitoring
Conclusion -‐ 2
Exploit all uses of the EMR to Improve practice efficiency Ensure patient safety Learn about your patients (registries)
Compare treatments Engage with patients
Conclusion -‐ 3
Understand Personalized and Precision medicine
How will we integrate genomic data in clinical practice in the future?
Conclusion -‐ 4
Predictive models inform care How do we integrate these into practice in the EMR?
Conclusion -‐ 5
How can we reduce the lethal lag time? Getting medical findings into practice more rapidly
How can we engage patients? Real time data on populations New technology for Big Data in health care
Contact me @JohnSharp Ehealth.johnwsharp.com Linkedin.com/in/johnsharp Slideshare.net/johnsharp ______________________ ClevelandClinic.org @ClevelandClinic Facebook.com/ClevelandClinic youtube.com/clevelandclinic