How to design a pre-specified statistical analysis approach to limit p-hacking in clinical trials: the Pre-SPEC framework
Kahan BC, Forbes G, Cro S.
This article describes a five-point framework (the Pre-SPEC framework) for designing a pre-specified analysis approach that does not allow p-hacking. The framework is based on the principles in the SPIRIT and ICH-E9 guidelines and is intended to be used in conjunction with these guidelines to help investigators design the statistical analysis strategy for the trial’s primary outcome in the trial protocol.
A four-step strategy for handling missing outcome data in randomised trials affected by a pandemic
Cro S, Morris TP, Kahan BC, Cornelius VR, Carpenter JR
This paper presents a four-step strategy for handling missing outcome data in the analysis of randomised trials that are ongoing during a pandemic. Settings where treatment effects for a ‘pandemic-free world’ and ‘world including a pandemic’ are of interest are considered.
Public availability and adherence to prespecified statistical analysis approaches was low in published randomized trials
Kahan BC, Ahmad T, Forbes G, Cro S.
This article reviews how often a prespecified statistical analysis approach was publicly available for randomized trials indexed in PubMed. For most published trials, there was insufficient information available to determine whether the results may be subject to p-hacking.
Understanding current practice, identifying barriers and exploring priorities for adverse event analysis in randomised controlled trials: an online, cross-sectional survey of statisticians from academia and industry
Phillips R, Cornelius V
This article reports the findings of a survey of statisticians that aimed to gain a better understanding of current adverse event analysis practices and explored statisticians' priorities, concerns and barriers when analysing adverse events.
Current approaches to handling rescue medication in asthma and eczema randomized controlled trials are inadequate: a systematic review
Chis Ster A, Cornelius V, Cro S
This review examines how rescue medication is defined, reported, and accounted for in randomized controlled trials in eczema and asthma populations. We found that while rescue medication is frequently permitted in trials alongside the treatment under evaluation, there is minimal attention or interest in isolating the underlying treatment effect of interest.
Evidence of unexplained discrepancies between planned and conducted statistical analyses: a review of randomised trials
Cro S, Forbes G, Johnson NA, Kahan BC
A review of published randomised trials, identifying a high rate of unexplained discrepancies between the planned and reported statistical methods used to analyse the primary outcome; therefore a need for improved reporting and transparency around statistical methods.
Sensitivity analysis for clinical trials with missing continuous outcome data using controlled multiple imputation: A practical guide. Statistics in Medicine, 2020
Cro S, Morris TP, Kenward MG, Carpenter JR
This tutorial in biostatistics provides an accessible overview of controlled multiple imputation procedures, and a practical guide to their use for missing data sensitivity analysis of a continuous outcome in clinical trials. Worked examples and Stata code are presented.
An evaluation of inverse probability weighting using the propensity score for baseline covariate adjustment in smaller population randomised controlled trials with a continuous outcome. BMC Medical Research Methodology, 2020
Raad H, Cornelius V, Chan S, Williamson E, Cro S
This article explores the use of IPTW for baseline covariate adjustment in small sample randomised trials and subsequently provides a caution on the use of IPTW without small sample modifications to the standard error calculation.