My research interests are in statistical methods and its applications. I have made contributions to hierarchical and multilevel models (also known as variance components models), the design and analysis of clinical trials, survival analysis, statistical methods for critical care, health services research and statistical bioinformatics.
In work funded by the ARC, I am developing new
statistical methods for institutional comparisons in collaboration with Associate Professor John Moran
of The Queen Elizabeth Hospital,
Adelaide. Using a three-stage approach and hierarchical generalised linear
mixed models, we have identified unusual performance as measured by significantly high or low mortality in Australian and New
Zealand intensive care units. In 2009 and 2010, 4 ICUs in private hospitals had unusually high mortality
(Kasza et al, Statistics in Medicine, 2013). Using a three-level
hierarchical mixed model for longitudinal ICU mortality data, we have also established that there were 7 ICUs with
poor performance (high mortality) in one or more years over the period 2000 to 2010:
3 were in
private hospitals in Queensland, 1 a private hospital in Victoria, 2 were metropolitan ICUs, and 1 was a large tertiary
ICU in South Australia (Solomon et al, BMC Medical Research Methodology, 2014).
link to open access paper
-->
Sir David Cox and I have written a research monograph
on Components
of Variance, in which
we develop
statistical methods for variance components from first principles, with a
strong focus on both concepts and applications. The monograph is
intended for statisticians as well as users of statistical methods. Applications include the analysis of
two-colour microarray
data (.RData file, save as linked file),
blood
pressure data (text file), and data from
intensive care units in Australia and New Zealand from the Australian and New Zealand Intensive Care (ANZICS)
Database. See the review of our book from the Journal of the Royal Statistical
Society, Series A. You will find a list of errata here.
I have written an overview article on variance components
for the Encyclopedia of Biostatistics (Second Edition), which you may find helpful. Potential Honours, Masters and PhD students interested in working in any of these
or related areas of research are free to
contact me at any time.