ÃÛÌÒ´«Ã½app

Elham Rahme, PhD

Associate Professor, Department of Medicine and Associate Member, Department of Epidemiology and Biostatistics, McGill University

Correcting Hazard Ratio Estimates for Outcome Misclassification

ALL ARE WELCOME

SYNOPSIS:

Outcomes assessed from health services administrative databases are subject to misclassification. Misclassification may bias the estimate of the association between a treatment and that outcome with potential negative implications. We assessed through various simulation scenarios the performance of a 2-level multiple imputation technique using internal validation data to account for outcome misclassification in estimating hazard ratios from Cox regression models. We applied this method to a health services administrative database study of the association between statin use and risk of new-onset diabetes in a stratified random sample of 6,247 Quebec individuals among whom about half responded to a survey on diabetes status and about a quarter also provided fasting blood samples for glucose testing.

OBJECTIVES:

Evaluate the misclassification rate in diabetes status in an administrative database compared to patient self-report and blood glucose test results.
Present an internal validation method to correct diabetes status in the administrative database based on patient self-report and blood glucose test results.
Evaluate the performance of this method in terms of bias reduction using simulations.

BIO:

Dr Elham Rahme is an Associate Professor in the Department of Medicine and Associate Member in the Department of Epidemiology and Biostatistics of McGill University. She is Medical Scientist in the Division of Clinical Epidemiology and co-Director of the Metabolism and Experimental Therapeutic Program- Health Services component at the McGill University Health Centre. She is also co-Director of the ‘Methods’ component - Unité de soutien-SRAP du Quebec (CIHR-SPOR). She holds a PhD in statistics and has extensive expertise in pharmacoepidemiology and health services research. Her research interests include the evaluation of the safety, effectiveness, and economic implications of prescribed medications. She has conducted population-based studies on the utilization, adverse events, and cost of NSAIDs; risks and benefits of extended outpatient anticoagulant use following total hip and knee arthroplasty; home care following hospital discharge from hemiarthroplasty, use of serotonin re-uptake inhibitors in association with suicide; rate of acces-related infection in patients undergoing dialysis; estimation of the prevalence of diagnosed and undiagnosed diabetes in Quebec; rates of suicide attempts in Quebec; utilisation, safety and cost of anti-TNF-alpha drugs and assessment and development of biostatistical methods for pharmacoepidemiology research.