Use of electronic health data to support a Learning Health System: Lessons from several distributed networks in the US

Transcription

1 Use of electronic health data to support a Learning Health System: Lessons from several distributed networks in the US Jeffrey Brown, PhD Department of Population Medicine Harvard Medical School and Harvard Pilgrim Health Care Institute The Learning Health System in Europe Brussels, Belgium September 25, 2015

2 Some distributed networks in US CDC s Vaccine Safety Datalink (VSD) Health Care Systems Research Network FDA s post-market safety programs Meningococcal Vaccine Safety Study FDA Mini-Sentinel NIH Health Care Systems Collaboratory PCORI National Clinical Research Network (PCORnet) MDPHnet

3 Some distributed networks in US CDC s Vaccine Safety Datalink (VSD) Health Care Systems Research Network FDA s post-market safety programs Meningococcal Vaccine Safety Study FDA Mini-Sentinel NIH Health Care Systems Collaboratory PCORI National Clinical Research Network (PCORnet) MDPHnet

4 Vaccine Safety Datalink Funded by US Center for Disease Control and Prevention Post-marketing safety of vaccines Integrated delivery systems and health insurers Data standardization via a common data model Often rely on chart review Data are updated frequently (can be weekly) Work has informed vaccination policy

5 : Real Time Public Health Surveillance Cambridge Health Alliance 20 Sites 400,000 patients Atrius Health 27 Sites 700,000 patients Launched November 2012 Real time access to ambulatory EHR data to inform public health Automatically identifies reportable and chronic diseases and enables flexible querying Massachusetts Department of Public Health (MDPH) epidemiologists have submitted hundreds of queries Now an operational suite of systems supported by MDPH Mass. League Comm. Health 18 Sites 300,000 patients

6 6

7 Sentinel History 2007: FDA Amendments Act A mandate to create an active surveillance system Access data from 25 million individuals by July 2010 Access data from 100 million individuals by July : FDA launched the Sentinel Initiative 2009: Mini-Sentinel funded under Sentinel Initiative 5-year pilot to create active surveillance system monitoring the safety of FDA-regulated medical products Develop policies, processes, infrastructure, collaboration 2014: Funding awarded for Sentinel 7

8 Collaborators Lead HPHC Institute Data and scientific partners Scientific partners Institute for Health 8

9 Bringing multiple data partners together Centralized vs. Distributed Distributed data system is preferred because Data sits behind data partner s firewall Data remains under local control Only minimally necessary info is shared in a given analysis Preserve patient privacy & proprietary interests 9

10 Distributed Analysis 1- User creates and submits query (a computer program) 2- Data partners retrieve query 3- Data partners review and run query against their local data 4- Data partners review results 5- Data partners return results via secure network 6 Results are aggregated 10

11 Common Data Model 11

12 Impact / Dissemination 4 FDA drug safety communications Tri-valent inactivated flu vaccine and febrile seizures (no increased risk) Rotarix and intussusception (label change) Dabigatran and bleeding (no increased risk) Olmesartan and sprue-like enteropathy (label change) 70 peer-reviewed articles 48 methods reports / white papers Thousands of unique queries and comparisons contributing to over 140 formal assessments 12

13 Label change 13

14 Yih, N Engl J Med. 2014;370:503 14

15 Drugs This assessment [ used ] FDA s Mini-Sentinel pilot... Nov 2,

16 In the months following the approval of the oral anticoagulant dabigatran... in October, 2010, the FDA received through the FDA Adverse Event Reporting System many reports of serious and fatal bleeding events associated with use of the drug. N Engl J Med DOI: /NEJMp

17 Output Figure 1a. New Events of GIH per 100k Days at Risk in the MSDD between October 19, 2010 and December 31, 2011, by Drug, Incidence Criteria, and Washout Period for Individuals with a Pre-Existing Condition of Atrial Fibrillation 4.0 Rate of gastrointestinal 3.0 bleeding per 100k days at risk Incident with respect to Dabigatran and Warfarin (8 scenarios) Incident with respect to Warfarin New GIH Events per 100k Days at Risk Drug Incidence with respect to: Washout Period GIH Events/100k Days at Risk 3.5 Dabigatran Incident with respect to Dabigatran 183-Day Washout 2.0 Warfarin 2.5 Incident with respect to Dabigatran and Warfarin Day Washout Day Washout Day Washout Day Washout Day Washout Day Washout Day Washout Day Washout 365-Day Washout 183-Day Washout 365-Day Washout 183-Day Washout 365-Day Washout 183-Day Washout 365-Day Washout Incident with respect to Dabigatran Incident with respect to Dabigatran and Warfarin Dabigatran Incident with respect to Warfarin Incident with respect to Dabigatran and Warfarin Warfarin Warfarin-GIH-ICH_Part-1.pdf 17

18 18

19 Protocol-based assessment Dabigatran 19

20 Olmesartan label change: sprue-like enteropathy 20

21 ARBs and celiac disease: all users Cases per 100 person years LOSARTAN IRBESARTAN OLMESARTAN TELMISARTAN VALSARTAN Cases New users 235,630 40,071 81,560 24, ,159 ARBs: New users after >365 day washout; Celiac Disease: 1st dx code after >365 day without diagnosis. 21

22 ARBs and celiac disease: all users 2.5 All Cases per million days at risk LOSARTAN IRBESARTAN OLMESARTAN TELMISARTAN VALSARTAN Cases New users 535,045 69, ,630 44, ,618 22

23 ARBs and celiac disease: 2+ years Cases per million days at risk years LOSARTAN IRBESARTAN OLMESARTAN TELMISARTAN VALSARTAN Cases New users 25,045 2,721 4,419 1,124 13,925 23

24 The Mini-Sentinel ' provides an essential public health service. The current configuration the data model, the methods development, and the investigative team represents an impressive achievement.. Psaty. N Engl J Med 2014;370:

25 Key Contributors to Mini-Sentinel s Progress Tightly Coupled network Frequent and coordinated interactions between FDA, data partners, content experts, epidemiologists, and statisticians Clear ownership and goals Established agreements and contracts Distributed data network (no central repository) Public health practice Focused on best understood and useful data for purpose First: Claims and administrative data, plus access to full text records Syntax and semantics are clear and understood Then: electronic medical records, registries, Much more complex to understand, standardize, and use for research Rapid cycle development of capabilities 25

26 Thank You Use of electronic health data to support a Learning Health System: Lessons from several distributed networks in the US Jeffrey Brown, PhD Department of Population Medicine Harvard Medical School and Harvard Pilgrim Health Care Institute The Learning health System in Europe Brussels, Belgium September 25, 2015

27 For More Information PopMedNet Website popmednet.org PopMedNet Wiki popmednet.atlassian.net/wiki 27 27

28 Thank You Links mehi.masstech.org/what-we-do/hie/mdphnet/about nihcollaboratory.org/pages/distributed-research-network.aspx NIH Collaboratory Distributed Research Network Presentations

29 Selected references Vogel J, Brown JS, Land T, Platt R, Klompas K. MDPHnet: Secure, distributed sharing of electronic health record data for public health surveillance, evaluation, and planning. American Journal of Public Health 2014; 0: e1-e6. Klann, JG, Buck MD, Brown J, Hadley M, Elmore R, Weber G, Murphy S. Query Health: standards-based, cross-platform population health surveillance. J Am Med Info Assoc 2014; 0:1-7. Toh S, Platt R, Steiner JF, Brown JS. Comparative-effectiveness research in distributed health data networks.clin Pharmacol Ther 2011; 90(6): Behrman RE, Benner JS, Brown JS, McClellan M, Woodcock J, Platt R. Developing the Sentinel System A national resource for evidence development. New England Journal Of Medicine February 10, Brown JS, Holmes JH, Shah K, Hall K, Lazarus R, Platt R. Distributed health data networks: A practical and preferred approach to multi-institutional evaluations of comparative effectiveness, safety, and quality of care. Medical Care 2010; 48 Suppl: S45-S51. Maro JC, Platt R, Holmes JH, Strom BL, Hennessy S, Lazarus R, Brown JS. Design of a national distributed health data network. Ann Intern Med 2009; 151: Fleurence RL, Curtis LC, Califf RM, Platt R, Selby JV, Brown JS. Launching PCORnet, a national patient-centered clinical research network. J Am Med Inform Assoc doi: /amiajnl [Epub ahead of print] Curtis LC, Brown JS, Platt R. Four Health Data Networks Illustrate The Potential For A Shared National Multipurpose Big-Data Network. Health Affairs, 33, no.7 (2014): Holmes JH, Elliott TE, Brown JS, Raebel MA, Davidson A, Nelson AF, Chung A, La Chance P, and Steiner JH. Data warehouse governance for distributed research networks in the United States: a systematic review of the literature. J Am Med Inform Assoc Mar 28. doi: /amiajnl [Epub ahead of print] Raebel MA, Haynes K, Woodworth TS, Saylor G, Cavagnaro E, Coughlin KO, Curtis LH, Weiner MG, Archdeacon P, and Brown JS. Electronic Clinical Laboratory Test Results Data Tables: Lessons from Mini-Sentinel. Pharmacoepidemiol Drug Saf Feb 18. doi: /pds Curtis LH, Weiner MG, Boudreau DM, Cooper WO, Daniel GW, Nair VP, Raebel MA, Beaulieu NU, Rosofsky R, Woodworth TS and Brown JS. Design considerations, architecture, and use of the Mini-Sentinel distributed data system. Pharmacoepidemiol Drug Saf. 2012;21(S1): Platt R, Carnahan RM, Brown JS, Chrischilles E, Curtis LH, Hennessy S, Nelson JC, Racoosin JA, Robb M, Schneeweiss S, Toh S and Weiner MG. The U.S. Food and Drug Administration's Mini-Sentinel program: status and direction. Pharmacoepidemiol Drug Saf. 2012;21(S1):