Prof Andrea Jorgensen PhD MSc BA ATT ATII

Professor, Biostatistics Biostatistics


    Methodological quality and evidence synthesis in pharmacogenetic studies

    Pharmacogenetic studies (PGx studies) investigate how genetic variation influences drug efficacy and toxicity, focusing particularly on genes involved in drug metabolism and transportation. The ultimate goal is to maximise benefits whilst minimising harm from medicines, to bring an era of ‘personalised medicines’.
    Failure to replicate initial significant findings is a notorious problem amongst genetic association studies, a major reason for this being cited as lack of methodological quality. I have developed and published guidelines for assessing methodological quality when undertaking systematic reviews and meta-analyses of PGx studies. These guidelines can also be referred to when designing PGx studies. I have also provided empirical evidence of lack of methodological quality amongst PGx studies in a systematic review of Warfarin PGx.

    In terms of evidence synthesis, as a member of the statistical analysis team of the International Warfarin Pharmacogenetics Consortium (IWPC) I have analysed IPD contributed from twenty pharmacogenetic studies of warfarin worldwide. Using this data, a warfarin dosing model was developed based on a combination of patient demographics and genotype information. Similar methodology has been applied to develop a warfarin dose-revision algorithm, a project on which I collaborated. I am now a member of the statistical group of the International Warfarin Pharmacogenetics in Paediatrics Consortium, which will mirror the work of the IWPC, but this time for the development of a dosing model specifically for use in children.

    I have also undertaken systematic reviews of the PGx of warfarin and clopidogrel, both of which were undertaken in accordance with HuGENet guidelines. I have a particular interest in methodologies that enable maximal use of data identified via PGx systematic reviews within meta-analyses, and I undertook a review of such methods as part of my PhD. These methods have been applied to data identified in my systematic reviews of warfarin and clopidogrel, and ensured that the power to detect genotype effects was maximised. These two projects represent the first application of these methodologies in the field of pharmacogenetics. I also have an interest in the assessment and adjustment for selective reporting in meta-analyses of pharmacogenetic studies.

    Treatment adherence in pharmacogenetic studies

    An issue largely ignored in pharmacogenetic studies published so far is adherence with treatment, the assessment of which should form a fundamental part of analysing pharmacogeneitc data given the more explanatory and less pragmatic questions being addressed compared to those in a clinical trial of drug efficacy. Compliance with treatment has been assessed in a PGx study of the anticoagulant warfarin on which I am study statistician, and I have undertaken PGx analyses of association within this dataset, adjusting for extent of treatment compliance, with interesting findings. I am also co-applicant on an MRC methodological research grant (MRC MRP) in collaboration with Dr Ruwanthi Kolamunnage-Dona which involes developing more advanced methods involving joint modelling for dealing with extent of compliance in the analysis of pharmacogenetic studies.

    Analysing time-to-event outcomes in pharmacogenetic studies

    To date, the focus of developing new methodologies for analysing genetic association studies has been on binary outcomes since this is typically the case in gene-disease association studies which tend to dominate the genetic association literature. However, drug response is typically multi-factorial and complex and as such outcomes in pharmacogenetic studies are often time to event or quantitative. Analysing such outcomes is particularly problematic when dealing with rare or imputed genotypes. I am co-supervisor on a PhD project involving the development of methods and user-friendly software for analysing genetic association where outcomes are time to event, particularly when genotypes are rare or have been imputed.

    Research Group Membership

    Research Grants

    Development of methodology and computationally efficient software for the analysis of PGx exome sequencing studies of complex "time-to-event" outcomes


    April 2018 - March 2022

    R19 Evidence synthesis for biomarker validity to inform biomarker-stratified trials


    September 2017 - August 2020

    Improving the efficiency of biomarker-driven designs by using continuous biomarker information


    January 2018 - November 2019

    Online tool to guide biomarker-guided randomised controlled trials


    March 2016 - August 2017

    Epilepsy Pharmacogenomics: delivering biomarkers for clinical use (EpiPGX)


    October 2011 - September 2016

    Joint modelling of multivariate longitudinal and event time outcomes in clinical research (JoineR-M)


    August 2015 - December 2018

    A pharmacogenomic approach to coumarin anticoagulant therapy (PACT)


    October 2008 - March 2013

    Genetics of non-steroidal anti-inflammatory drugs (NSAIDs)-induced peptic ulceration: a genome wide association study.


    July 2010 - June 2011