playing with fhir ir
play

Playing with FHIR IR How to Exploit the EHR Mark L Braunstein, MD - PowerPoint PPT Presentation

Playing with FHIR IR How to Exploit the EHR Mark L Braunstein, MD Professor of the Practice School of Interactive Computing Georgia Institute of Technology Visiting Research Fellow E-Health Centre, CSIRO Variable Results Australia ranks


  1. Playing with FHIR IR How to Exploit the EHR Mark L Braunstein, MD Professor of the Practice School of Interactive Computing Georgia Institute of Technology Visiting Research Fellow E-Health Centre, CSIRO

  2. Variable Results “ Australia ranks highest on Administrative Efficiency and Health Care Outcomes, and is among the top-ranked countries on Care Process and Access ” Commonwealth Fund (2017)

  3. Common Trends Aging Population More Chronic Disease Increased Costs https://www2.deloitte.com/content/dam/Deloitte/global/Documents/Life-Sciences-Health-Care/gx-lshc-healthcare-and-life-sciences-predictions-2020.pdf

  4. In Including Australia Australian Institute of Health and Welfare 2018 https://www.aihw.gov.au/reports/australias-health/australias-health-2018/contents/table-of-contents

  5. Systems Not Designed for Chronic Care Medical Journal of Australia https://www.mja.com.au/journal/2007/187/2/care-patients-chronic-disease-challenge-general-practice

  6. Poor Care Coordination https://www.commonwealthfund.org/publications/surveys/2015/dec/2015-commonwealth-fund-international-survey-primary-care-physicians

  7. Lack of Continuity In Practice Nurses or Case 96 92 90 85 Managers 81 66 65 64 60 28 UK NETH NZ SWE AUS US NOR CAN SWIZ GER https://www.commonwealthfund.org/publications/surveys/2015/dec/2015-commonwealth-fund-international-survey-primary-care-physicians

  8. Variable Use of f In Informatics 90 80 80 70 60 60 57 53 50 40 30 30 24 20 11 11 10 0 Email Record Sharing Australia NZ SWIZ US

  9. Healthcare Professionals Want Digital Health Top 5 activities health professionals want to use digital technologies to help better support them to deliver health services Not using, g, but inter eres ested in using g Not interested ed in using ng a Current ently using ng a compute uter, a compute uter, , smart phone e or compute uter, smart phone ne or tablet et Activity smart phone ne or tablet et tablet et for this activity % % % Sharing health records with my patients 25 59 7 Transferring prescriptions to the pharmacy 25 56 8 Providing interactive decision-making support 32 53 6 Communicating with patients before or after consultations 33 49 7 Sharing health records with other practitioners 43 45 4 Courtesy Australian Digital Health Agency

  10. EHRs Have “Issues” https://medicomp.com/whats-the-most-frustrating-about-ehrs/

  11. Goals: IO IOM safe, effective, patient-centered, timely, efficient, equitable Learning Health System https://www.ahrq.gov/professionals/systems/learning-health-systems/index.html

  12. Necessary ry In Informatics Substrate Open EMR Adoption Analytics Access to POC Interoperability

  13. Variable EHR Readiness http://www.oecd.org/els/health-systems/health-statistics.htm

  14. Current EHR Limitations No standards regarding discrete data No integrated communication Not user-friendly Big data but not smart data Lack of regulation Hospital CIOs Remember these are we proceed https://www.beckershospitalreview.com/healthcare-information-technology/the-problem-with-ehrs-5-complaints-from-cios.html

  15. What to Do? Repair? Replace?

  16. In Innovate! “Fostering third party apps creates a market where innovations compete with each other for purchase and use by providers (and patients), thus reducing dependency on updates and specific functions made by an EHR vendor .” -- Ken Mandl, Josh Mandel, Isaac Kohane https://www.sciencedirect.com/science/article/pii/S2405471215000046

  17. How?

  18. It It Works Quantitatively https://research2guidance.com/325000-mobile-health-apps-available-in-2017/

  19. Proof of Effectiveness is Lacking but … 11 of 23 randomized controlled trials showed a meaningful effect on health or surrogate outcomes attributable to apps … the overall evidence of effectiveness was of very low quality … pilot studies … only one has progressed to a large clinical trial. https://www.nature.com/articles/s41746-018-0021-9

  20. … Mostly “ Siloed ” Apps

  21. Ext xtend the Phone App Model to EHRs?

  22. Harvard’s SMART https://apps.smarthealthit.org/

  23. Georgia Tech’s HDAP http://www.hdap.gatech.edu/apps/

  24. Cerner/Epic …

  25. Even the US Government! https://bluebutton.cms.gov/

  26. What Will Happen Next xt? “These apps will give new life to data entered into EHRs and other health IT platforms by providing the ability to visualize risks, trends, and trajectories; mash up clinical records with external data sources; and deliver decision support to clinicians and patients during and between encounters .” -- Ken Mandl, Josh Mandel, Isaac Kohane https://www.sciencedirect.com/science/article/pii/S2405471215000046

  27. A Better EHR?

  28. Ju Juxly Timeline No standards regarding discrete data No integrated communication Not user-friendly Big data but not smart data

  29. Ju Juxly Trends No standards regarding discrete data No integrated communication Not user-friendly Big data but not smart data

  30. Coordinated, , Continuous Care

  31. Within Cerner -

  32. More Complete Patient View EHR Data No standards regarding discrete data No integrated communication Not user-friendly Big data but not smart data Patient Generated Data

  33. Outcome Prediction

  34. How?

  35. HL7 Timeline Messaging (lab test results) Model Driven (patient record summaries) FHIR: V 3.0.1 April 19, 2017 … 2018?

  36. A Common Data Model No standards regarding discrete data No integrated communication Not user-friendly Big data but not smart data

  37. Condition: Human View

  38. Condition: Machine View 73211009 Remember me!

  39. Uniform API

  40. “Just Like” Amazon! https://www.amazon.com.au/s/ref=nb_sb_noss?url=search- alias%3Daps&field-keywords=size+10+ladies+blue+sweater http://hapi.fhir.org/baseDstu3/Condition?code=SNOMED-CT|73211009 Population level query

  41. Condition Specific Charting

  42. Medication Reconciliation One entry Multiple dispensings RxNorm for different names

  43. Clinical In Insights No standards regarding discrete data No integrated communication Not user-friendly Big data but not smart data

  44. Lower Cost No standards regarding discrete data No integrated communication Not user-friendly Big data but not smart data

  45. Precision Medicine No standards regarding discrete data No integrated communication Not user-friendly Big data but not smart data

  46. What About Your Patients?

  47. Patient Controlled Health Record

  48. EHR Connected Mobile Apps

  49. Over 500 Collaborating Hospitals/Clinics (J (June)

  50. Sharing Models Opt Out - Centralized Opt In - Federated

  51. Secondary ry Use?

  52. Chart Review

  53. Search Results

  54. Drill Down: Conditions/Medications

  55. Drill Down: Notes

  56. Research “a historic effort to gather data from one million or more people living in the United States to accelerate research and improve health. By taking into account individual differences in lifestyle, environment, and biology, researchers will uncover paths toward delivering precision medicine .” FHIR App https://allofus.nih.gov/

  57. Emory ry Artificial In Intelligence Sepsis Expert (A (AISE) Trained on 31,000 Emory ICU patients Validated on 52,000 MIMIC III patients Third International Consensus Definitions for Sepsis (Sepsis-3) 65 features (variables) calculated on hourly basis Can predict sepsis 4 hours in advance (ROC of .85)* *https://www.ncbi.nlm.nih.gov/pubmed/29286945

  58. No standards regarding discrete data No integrated communication DeepAISE on FHIR Not user-friendly Big data but not smart data Drag/ Drop Automatic Text messaging Emory AISE Score Improvement Decline or the eICU application Philips DRS Score: Higher predicts readmission Submitted to AMIA 2018

  59. eICU team adjudicates warnings

  60. Medical Education on FHIR? We are partnering with UQ ITEE and Faculty of Medicine to offer an experimental course to explore the potential of using FHIR to digitize case based learning.

  61. Diabetes Case Study

  62. Burn Case Study Nathan and is alert and oriented with a GCS of 15

  63. Nathan’s % total body surface area (%TBSA) of burn is calculated using a Lund-Browder chart to be 62% with 59% full thickness burns, 2% deep dermal and 1% partial thickness burns

  64. Want to Try ry It It Yourself? http://cs6440.gatech.edu/

  65. mark.braunstein@csiro.au

Recommend


More recommend