2016-Design and Methodological Issues in Health Studies of Vulnerable Populations


Design and Methodological Issues in Health Studies of Vulnerable Populations 
Chair: John Petkau (University of British Columbia)
Organizer: Lehana Thabane (McMaster University)
[PDF]

LISA LIX, Ottawa University of Manitoba
Methodological Challenges in Aboriginal Health Research  [PDF]
 
Aboriginal populations in Canada, which include First Nations, Inuit, and Metis people, often have poorer health outcomes and a greater frequency of health risk behaviors than non-Aboriginal populations. Comparative information about the health and healthcare use of Aboriginal and non-Aboriginal populations is therefore important in the design and delivery of healthcare services. Misclassification bias in ascertaining ethnicity, small sample sizes, incomplete data about on-reserve populations, and differential item functioning of self-reported health status measures are challenges that can affect the validity of comparative studies. This talk will describe the opportunities for biostatisticians to contribute to high-quality Aboriginal health research in Canada, by addressing these methodological challenges while remaining sensitive to data ownership and control principles and information privacy legislation. 
 
ROBERT W. PLATT, McGill University
Randomized Trials in Obstetric Populations  [PDF]
 
Pregnancy is a vulnerable period, in which both mother and fetus are at risk for benefit and harm. I discuss three problems arising in studies during pregnancy: 1) Treatments can have opposite effects for the fetus and the mother, and some important outcomes are very rare. Composite outcomes are a solution, but raise other problems. 2) Timing during pregnancy of an intervention (e.g., induction of pregnancy) is a common research interest. This raises statistical and logistical challenges; the optimal statistical design can be at odds with feasible care plans. 3) Pregnancy is a highly dynamic process, in which the condition of the mother and fetus can change rapidly. Consideration must be given to timing of recruitment and randomization. 
 
LEHANA THABANE, McMaster University
Planning for the Twilight Years: Why We Need More Statisticians Involved in Trials in the Elderly Living in Long-Term Care Settings  [PDF]
 
Providing optimal healthcare of the elderly living in long-term care (LTC) settings requires deliberate planning and good evidence to support decision-making. Conducting trials in LTC to generate evidence or apply evidence in practice can present many design, implementation, methodological, ethical and analytical challenges. First, the population is quite frail which can present serious challenges; second, the environment is hard to work in because of the overcommitted staff; third, outcome assessment can be challenging; fourth, determining the appropriate unit of randomization, analysis or inference can be complicated. In this presentation, I will use experiences from some of our studies to illustrate the issues and highlight the role of biostatisticians in addressing them.