Menu
Log in


SA Branch webinar - A LEGENDARY way to do observational data analysis at scale!

  • 19 Aug 2020
  • 6:00 PM - 7:00 PM (UTC+09:30)
  • virtual Zoom meeting

Branch Meeting - Wednesday, 19th August 2020

The South Australian Branch of the Statistical Society would like to invite you to the August meeting of the 2020 program.

Virtual Venue: Join the Zoom meeting using your PC or device https://unisa.zoom.us/j/93416975159?pwd=eWU4aWo5MjZTZ3YwajlRekd6ekhaZz09 
Password: 718170

or Join from dial-in phone line:  Dial: +61 2 8015 2088

Meeting ID: 934 1697 5159


Speaker: A/Prof Nicole Pratt, University of South Australia

Topic: A LEGENDARY way to do observational data analysis at scale!

Abstract

Objectives: Evidence derived from existing healthcare data including administrative claims and electronic health records can fill evidence gaps in medicine, but is often criticized due to the potential for observational study bias, for example due to residual confounding. Other concerns include p-hacking and publication bias. Here we detail a set of principles embodying a new paradigm for observational research aimed at addressing these concerns, and describe a generic implementation of these principles.

Materials and Methods: The Observational Health Data Sciences and Informatics (OHDSI) collaborative launced the Large-Scale Evidence Generation and Evaluation across a Network of Databases (LEGEND) research initiative, aiming to generate evidence on the effects of medical interventions using observational healthcare databases. We define ten principles of LEGEND, prescribing the generation and dissemination of evidence on many research questions at once, for example comparing all treatments for a disease for many outcomes, thus preventing publication bias. These questions are answered using a prespecified and systematic approach, avoiding p-hacking. Best-practice methods address measured confounding, and control questions (questions where the answer is known) quantify potential residual bias. Finally, the evidence is generated in a network of databases to assess consistency, by sharing open source analytics code to enhance transparency and reproducibility, but without sharing patient-level information, ensuring patient privacy.

Discussion: Following guiding principles addressing study bias, p-hacking, and publication bias, LEGEND seeks to generate reliable evidence from existing healthcare data. The principles of LEGEND will be highlighted using an example study on effects of antihypertensives, evaluating internal and external validity of the generated evidence.

Biography

A/Prof Pratt is an expert in biostatistics and pharmaco-epidemiology, specialising in the development of methodologies to study the effects of medicines and medical devices in linked health-care datasets. Nicole made significant contributions to medication safety surveillance, developing analysis software and pioneering a distributed network model to allow it to be implemented globally. Her research includes knowledge generation and quantification of harms from medicines and devices and has led to TGA safety warnings, changes in practice and changed guidance for medicine use. For the last five years Nicole has worked collaboratively with OHDSI (Observational Health Data Sciences and Informatics) and has helped to develop a comprehensive framework for analysing observational healthcare data at-scale. LEGEND (Large-scale Evidence Generation across a Network of Database) aims to generate real world evidence on the effects of medical interventions using best-practice statistical methodology to support clinical decision making.

Feel free to forward this meeting notice to colleagues, all welcome.


Powered by Wild Apricot Membership Software