STREAMLINING OPENCLINICA DATA UPLOADS JACOB ROUSSEAU 1 MARINEL

STREAMLINING OPENCLINICA DATA UPLOADS JACOB ROUSSEAU 1, MARINEL CAVELAARS 2, JEROEN BELIËN 3, JAN-WILLEM BOITEN 4, GERRIT MEIJER 1 1 DEPT. OF PATHOLOGY, THE NETHERLANDS CANCER INSTITUTE, AMSTERDAM, THE NETHERLANDS 2 THE HYVE, UTRECHT, THE NETHERLANDS 3 VU UNIVERSITY MEDICAL CENTER, AMSTERDAM, THE NETHERLANDS 4 LYGATURE, UTRECHT, THE NETHERLANDS

OPENCLINICA DATA UPLOADER Open. Clinica Data Uploader: Web-based and stand-alone tool for data uploads to Open. Clinica 2

CTMM-TRAIT • Tra. IT (Translational research IT) started in October 2011 as a public / private partnership initiative to provide a national translational research infrastructure • It was initiated by CTMM (Centre for Translational Molecular Medicine) • CTMM merged in January 2016 with TI Pharma and operates under the name Lygature - http: //www. lygature. org/ • CTMM-Tra. IT has 32 partners contributing to the Tra. IT services – all Dutch academic medical centers, charities, private partners, research institutes and Open. Clinica LLC • Tra. IT services are used by more than 2700 users from appoximately 250 institutes 3

CTMM-TRAIT APPLICATION SUITE 4

OPENCLINICA. NL ENVIRONMENTS IN TRAIT • Open. Clinica instances • Archive – for large ‘non-active’ studies • Acceptance – for technical testing, integration development and issue solving • Sandbox – for user CRF- and study-development • Production – for clinical data storage 5

OPENCLINICA. NL 6

OPENCLINICA. NL USER SUPPORT • Large number of the Tra. IT applications are centrally hosted at our hosting provider • Tra. IT provides user support for all applications and a dedicated support officer for Open. Clinica. nl • Open. Clinica training 7

NUMBER OF STUDIES Number of active studies since 2014 250 200 150 100 50 2014 2015 au g n ju r ap b fe c de t oc au g n ju r ap fe b 0 2016 8

NUMBER OF USERS Number of active users since 2014 2000 1500 1000 500 2014 2015 au g n ju r ap b fe c de t oc au g n ju r ap fe b 0 2016 9

INTEGRATED APPLICATIONS • Ldot → Open. Clinica Ldot is the study scheduling- and workflow-tool developed for CTMM-Tra. IT by MEMIC, Maastricht University, Uses Open. Clinica SOAP web-services • Open. Clinica → ODM 2 I 2 B 2 → tran. SMART is the central data-sharing and analytics datamart used by CTMM-Tra. IT • Open. Clinica → Imaging applications. ( DICOM-viewers and t. EPIS ) With parameterized links • tran. SMART → Imaging applications Using URL’s • Single Sign-On using SAML 2. 0 • tran. SMART → Galaxy Custom tool • Open. Clinica Data Uploader ( OCDU ) → Open. Clinica 10

DATA UPLOADING TO OPENCLINICA • • Users want to upload data to Open. Clinica from various sources • Excel and csv-files • EHR-systems • Study migrations from other e. CRF applications • Clinical data acquisition devices Open. Clinica extensions to upload data • Open. Clinica Bulk Loader – sgranite • Open. Clinica web-service client (Tra. IT) - Arjan van der Velde; tightly integrated with Mirth. Connect • Open. Clinica Data Importer (Tra. IT) – Cuneyt Parlayan; Windows application / inserts of subjects and events with SQL-scripts • ODIN – Csaba Halmagyi, Dept. Hematology, University of Cambridge – web based, uses webservices 11

DESIGN GOALS • Run either in interactive or in stand-alone mode to allow unattended uploads • Web-based • Easy intergration in different types of data-pipelines • Usable for users with little Open. Clinica know-how • User-interface with step-by-step wizzard-flow • Complete client-side validation with clear error-messages in order to avoid user confusion and subsequent errors generated by Open. Clinica web-services • Mapping functionality from source-data to Open. Clinica CRF-items • Connect to multiple Open. Clinica instances • Handle different initial situations (subject and/or events can be present or absent) 12

HIGH-LEVEL DESIGN • Input: text-files with tab-separated values • Separate input files for subjects, events and CRF-data • Column names should match item-names in the CRF but custom mapping is possible • Reuse mappings previously made for unattended uploads • Clear feedback on errors in the input files • Use Open. Clinica SOAP-Webservice API 13

DEVELOPMENT PARTNERSHIP • The Hyve, Utrecht, The Netherlands • NIHR Rare Diseases Translational Research Collaboration, Department of Haematology, University of Cambridge, United Kingdom 14

FUNCTIONAL SKETCH – INTERACTIVE MODE 15

FUNCTIONAL SKETCH – UNATTENDED MODE 16

USER INTERFACE – WIZARD FLOW

USER INTERFACE – COLUMN MAPPING TO CRFITEMS

TECHNOLOGY STACK • Java 1. 8 with lambda-expressions • Spring-Boot • Spring-MVC and Thymeleaf • Postgres (8. 4) • Unit-testing with JUnit 4 • Continuous integration • Git. Hub • Open. Clinica Data Importer: https: //github. com/CTMM-Tra. IT/Open-Clinica-Data-Uploader • Open. Clinica: https: //github. com/CTMM-Tra. IT/Open. Clinica 19

CHANGES TO OPENCLINICA WEB-SERVICES CODE • Addition of extra information to the web-services • Study → get. Metadata, required status of an item per CRF-version • Study. Subject → list. All. By. Study, study-subject information to include study. Subject. OID • Study. Subject → list. All. By. Study, changes of the Event element in to include: • Repeat occurrence • Event status • List of event. CRF’s with CRF-name, CRF-OID, status and version-information … <Open. Clinica: Item. Present. In. Form. OID="F_MUSTFOR_NON__020" Page. Number="1" PHI="No“ Show. Item="Yes" Order. In. Form= "2" Required="Yes"> <Open. Clinica: Left. Item. Text>male 0_female 1</Open. Clinica: Left. Item. Text> …. 20

APPLICATION CONFIGURATION • Customization of log-in page, welcome message, logo and application name • Selection of Open. Clinica instances to connect to in application. yml file • Scheduling and error reporting for unattended mode 21

DEMONSTRATION • Interactive mode • Definition of mapping • Registration of subjects and scheduling of events • Upload of data • Validation errors 22

FUTURE PLANS Future development on Open. Clinica Data Uploader • Convert to REST-API • Integrate with more types of pipe-lines Open. Clinca web-services enhancements • Addition of web-service dry-run / test-only mode • Introduce single validation module for both manual and web-service channels 23

ACKNOWLEDGEMENTS • • • Csaba Halmagyi - Dept. Hematology, Cambridge University, Great Britain Piotr Zakrzewski - The Hyve, Utrecht, the Netherlands Bo Gao - The Hyve, Utrecht, the Netherlands Ilona Pinter - The Hyve, Utrecht, the Netherlands Jolanda Strubel - The Hyve, Utrecht, the Netherlands Rene Breet - Free University, Medical faculty, Amsterdam, the Netherlands Marieke Vianen - Utrecht University Medical Center, Utrecht, the Netherlands Cuneyt Parlayan - Free University, Medical faculty, Amsterdam, the Netherlands Remond Fijneman - The Netherlands Cancer Institute, Amsterdam, the Netherlands 24

QUESTIONS & COMMENTS 25

1 REVISIONS • Revision history: • • 0. 2 processed comments 0. 3 addition more integrated applications