๐ขPublication Alert! A new paper from my PhD is now online at Statistics in Medicine, sharing a simulation study on the performance of mixed effects models and GEEs for partially clustered trials. With @LisaYelland, Tom Sullivan & @JessKasza
https://t.co/GUHWeIdhnF
Summary๐งต1/5
Kylie Lange and colleagues outline a sample size calculator for clinical trials in which some participants are independent and others are clustered, such as multiple births. @K_Lange_@UniofAdelaide@sahmriAU@austrim_cre@SCTorg
@austrim_cre@LisaYelland My gratitude to @austrim_cre for the scholarship support and for providing such a valuable network with which to share and improve my ideas.
Help SAHMRI create healthier futures by applying to join our team!
Current vacancies:
๐ข Statistician
๐ป Lead, Business Transformation
๐ Bioinformatician
Learn more โก https://t.co/FRbwhzPX30
@DogICUma In addition to Shiny CRT already mentioned (which covers various clustered designs) there is also this one specific to cluster randomised designs and includes binary outcomes: https://t.co/t50gbdI8XN
Some great work from @KristyRobledo looking at how multiple births are accounted for in neonatal trials (spoiler: not well!) - partial clustering in action!
We should do better in accounting for multiple births in neonatal randomised trials: a methodological systematic review #newpaper@TrialsCentre
https://t.co/hhO6pCwjc1
But not always! Mixed models and exchangeable GEEs did have some problems for some randomisation methods when the ICC was high. Check out the paper for what to consider when designing and analysing these trials! (5/5)
https://t.co/GUHWeIdhnF
@austrim_cre@sahmriAU@UniofAdelaide
๐ขPublication Alert! A new paper from my PhD is now online at Statistics in Medicine, sharing a simulation study on the performance of mixed effects models and GEEs for partially clustered trials. With @LisaYelland, Tom Sullivan & @JessKasza
https://t.co/GUHWeIdhnF
Summary๐งต1/5
to compare the performance of GEEs and mixed models in estimating the treatment effect on a continuous outcome. We found that the power formulas for exchangeable GEEs are approximately accurate for mixed effects models. And in many settings, both methods perform well. (4/5)
The full program for the Australian Trials Methodology Conference 2023 is now online.
Check out some of our invited speakers
DAY 1
Full details:
https://t.co/8ER2FEsTp7
Tickets still available - attend in person or online!
We're getting pretty excited to see all our invited speakers - just over a month to go!
Contributed sessions will be announced shortly.
Have you got your tickets yet? ๐ซ Join us in Melbourne or online!
https://t.co/LLPaduNkhd
A late update, but I had a great time at #ISCB44#ISCB2023 last week in Milan. I was able to share some my PhD research with a poster presentation and catch other @austrim_cre presentations from Ehsan Rezaei and Kelsey Grantham. I've come home with a head full of new ideas!