Transcription
James Doiron, Manager, HRDR , University of Alberta, CanadaShane McChesney, Nooro Online ResearchIASSIST 2014Toronto, Canada
Translating Research in Elder Care(TREC) PI Dr. Carole Estabrooks (T1 CRC in KT); FoN U of A Assessing influence of org. context on use of knowledge (esp research) and, in turn, influence of knowledge use on LTC residents’outcomesPractical objective: contribute to improving the quality of care andquality of life/end of life for residents living in residential LTCsettings and quality of work life for care providersMulti province (4) & sites (90)30 investigators, as well as policy and decision makers, trainees,staffT1.0 Two main projects (quantitative & qualitative), and threeinterrelated pilot projectsT2.0 Three main projects (TMS, Intervention, & SNA)
TREC Measurement System Data Sources1. TREC care provider surveys HCAs (75%) & regulated staff 200 variables—likert, categorical, & open text Alberta Context Tool (ACT) embedded; derived context scales Administered using CAPI (HCAs) and online (regulated staff)2. Facility Profile Survey 180 variables—categorical & open text Administered in paper or electronic format by research managers3. Care Unit Profile Survey 100 variables—categorical & open text Administered in paper or electronic format by research managers4. RAI-MDS 2.0 500 variables Collected continuously through projects (2007-2020) Full Admission, Annual, and Quarterly assessments Quality Indicators, Scales, RUGS Administered by care providers; data acquired from data stewards
tcomes4
1.02.0
TRECs .1.0 & 2.0TREC 1.oTREC 2.0Timeline5 years (2008-2012)7 years (2014-2020)Funding 4.7 million (CAD); CIHR 9 million (CAD);ProvincialProvinces3 (AB, SK, MB)4 (BC, AB, SK, MB)Nursing Homes3690TREC Surveys4000 (2 waves) 15,000 (3 waves)RAI-MDS 2.0 120,000 350,000Intervention FacilitiesN/A45
What is the HRDR? Phase 1: secure VRE for housing and managing health research data and metadata throughout their lifecycleBased in the Faculty of Nursing, University of AlbertaSupports and promotes health research and multi-disciplinarycollaborationOperational phase began Jan 2013Currently 30 active projects and 100 users (local, nat, internat)Provides secure remote access to data and a regular suite ofanalytic softwareOperates on a ‘minimal cost recovery’ basisPromotes secondary use and re-purposing of research data
Automate TheAutomatable RAI-MDS 2.0 Data FileUploads 500 variables per row 90 participating facilities 350K rows expected several versions of instrument several source informationsystems with different exportformats
Real-Time Feedback Replace SFTP Site WithPython-Powered SecureUpload Site real-time feedback to uploader notification of upload to datateam automated scrubs and reports
RejectPrivacyViolationsAt SourcePersonally identifying information or data for non-participatingunits/facilities is rejected immediately and does not get saved todisk.
Inside The HRDRData cleaning as an automated-yet-reproducible series of steps:1. The raw file as uploaded2. The "normalized" file (adapting the 7 incoming formats into a single cleanlayout)3. Unit Reconciliation4. .more stages can be added once they are reduced to practice
Inside The HRDR Data remains in text files for as long as possibleEach version of the file is retained as a historyFolders can be version controlled with MercurialGoals: Auditability Reproducibility Who changed what when? Why?
Free Limited Human Resources To Do Higher-Level Work Browse any row of any file at any stage See flagged issues / warnings Add functionality as new needs / issues / fixes are discoveredAll from having rich human & machine-readable metadata.
Technical Implementation Metadata Technology /Nooro Metadata driven infrastructure Automate ingestion / processing / publications DDI compliant Researchers tools Discovery / Documentation Produce research subsets (data personalization) Support collaborative researcher Business Intelligence Reporting Data mining
Components / Infrastructure
Questions?Contact:James DoironManager, HRDRUniversity of Albertajames.doiron@ualberta.caShane McChesneyNooro Online Researchshane@nooro.com
PI Dr. Carole Estabrooks (T1 CRC in KT); FoN U of A Assessing influence of org. context on use of knowledge (esp research) and, in turn , influence of knowledge use on LTC residents’ outcomes Practical objective: con
