Tania Patrao


Dr Robert Ware, Professor Gail Williams


Injecting illicit drugs is a behaviour strongly associated with other highly marginalised characteristics and is a source of health and social harm out of all proportion to its prevalence in the community. This population is typical of many marginalised populations with the combination of low population prevalence, illicit behaviours, complex aetiology and extreme health hazards making research involving People who Inject Drugs exceptionally challenging. Some of the most challenging aspects of research with People who Inject Drugs is data collection and analysis. Data collection is complicated by the injecting drugs being illicit, and by the propensity for research subjects to be or become homeless, itinerant, mentally ill, intoxicated, imprisoned or simply unwilling or unable to divulge accurate or sensitive information. Similarly, data analysis is hampered by the lack of sampling frames for drug user populations which makes assessment of representativeness difficult and extrapolation problematic.


The aims of this research are to:

  1. Describe the challenges related to data collection and analysis of longitudinal studies with injecting drug users;
  2. Develop novel counterfactual and regression models to assess how recruitment and attrition bias impacts on the generalizability and reliability of study outcomes;
  3. Combine data from prospective cohorts, surveillance systems and other databases to assess the effectiveness of developed recruitment strategies;
  4. Generate theoretical sampling frames for future Injecting Drug Use (IDU) cohorts for intervention trials (eg: HCV, vaccine studies).


The research will be divided into two components addressing longitudinal study design and analysis.


Conduct a systematic literature review of recruitment strategies used in hard to reach populations to establish the most appropriate and cost effective methods. The aim of this systematic review is to establish sound recruitment strategies which will result in appropriate samples sizes and generalizable results.


a) systematic literature review of current recruitment and retention strategies used in longitudinal studies of marginalised populations. 

b) paper on the predictors of attrition in a longitudinal study of ex-prisoners using longitudinal data; 

c) analysis of attrition using routinely collected QH data.


The next phase will concentrate on methods to adjust for missing data which is a common problem in studies involving marginalised populations. Missing data creates the potential for biased attrition and decreased generalizability and statistical power of longitudinal studies.

The datasets used in this analysis will be the Passports to Advantage study (University of Melbourne) and the SuperMIX study (Burnet Institute), both of which are longitudinal studies managed by the most experienced researchers in key related fields of justice health (Dr Stuart Kinner) and injecting drug use (Dr Paul Dietze).


a. Development of novel regression models to assess how attrition bias impacts on generalizability and reliability of study outcomes 

b) Development of a theoretical sampling frame for research studies involving People who Inject Drugs


Recruitment, retention and attrition in longitudinal studies is challenging for most research studies with research in marginalised populations being exceptionally challenging. The outcomes of this PhD will result in refining better sampling strategies and different methods to minimise attrition. The Centre for Research Excellence in Injecting Drug Use under which this PhD is being conducted have a strong focus on enhancing the translation of research outcomes into policy and practice. This research will influence the establishment of future IDU cohorts for intervention trials and will be translatable to other marginalised populations. 


The PhD candidature was commenced in June 2014, the systematic literature review is underway and active collaboration with Chief Investigators for access to the datasets is in progress.


2014 – 2017


For more information relating to this project, please contact Tania Patrao:


  • Thumbnail_logo-university-of-queensland