Analysis and prediction of recruitment and retention at the design stage of scientific studies
Background: Scientific progress is driven by numerous studies in which data are collected and analyzed using statistical methods that require an adequate number of observations to make reliable statements. The ability of a study to successfully collect the required number of observations depends on a realistic study design based on accurate predictions of recruitment and retention. Such parameters are difficult to predict at the design stage [1,2], and inaccurate predictions at this stage inevitably lead to inappropriately designed studies, small sample sizes, unreliable statistical inference, and study discontinuations.
Aim: Your project, either a master's thesis or a semester project, focuses on one or several tasks listed below and clarifies which frequentist and Bayesian statistical methods provide realistic recruitment and retention predictions before a study started.
- Aleatory and epistemic stochasticity
- Mathematical derivations
- Prior elicitation
- Design of simulations that compare deterministic and probabilistic statistical methods
- Extension of statistical methods to cover several recruiting centers
- Shiny web application
- Proper scoring rules
- Illustrative examples
- Graphical visualizations
- Documentation
Impact: The results of your project contribute to a better understanding of which frequentist and Bayesian statistical methods are useful for analysis and prediction of recruitment and retention at the design stage of scientific studies. In the long term, they help to increase the number of realistically designed scientific studies that successfully collect a sufficient number of observations and provide reliable statistical inference, thus supporting scientific progress worldwide.
Contact: malgorzata.roos@uzh.ch
References:
[1] Heesen P, Roos M. Freely accessible software for recruitment prediction and recruitment monitoring of clinical trials: A systematic review. Contemporary Clinical Trials Communications. 2024;39:101298.
https://doi.org/10.1016/j.conctc.2024.101298
[2] Rojavin M. Patient recruitment and retention: From art to science. Contemporary Clinical Trials. 2009;30:387.
https://doi.org/10.1016/j.cct.2009.06.002