The impact of covariate adjustment at randomization and analysis for binary outcomes: understanding differences between superiority and noninferiority trials

Abstract

SUMMARY The question of when to adjust for important prognostic covariates often arises in the design of clinical trials, and there remain various opinions on whether to adjust during both randomization and analysis, at randomization alone, or at analysis alone. Furthermore, little is known about the impact of covariate adjustment in the context of non-inferiority (NI) designs. The current simulation-based research explores this issue in the NI setting, as compared to the typical superiority setting, by assessing the differential impact on power, type I error, and bias in the treatment estimate as well as its standard error, in the context of logistic regression under both simple and covariate adjusted permuted block randomization algorithms. In both the superiority and NI settings, failure to adjust for covariates that influence outcome in the analysis phase, regardless of prior adjustment at randomization, results in treatment estimates that are biased toward zero, with standard errors that are deflated. However, as no treatment difference is approached under the null hypothesis in superiority and under the alternative in NI, this results in decreased power and nominal or conservative (deflated) type I error in the context of superiority, but inflated power and type I error under NI. Results from the simulation study suggest that, regardless of the use of the covariate in randomization, it is appropriate to adjust for important prognostic covariates in analysis, as this yields nearly unbiased estimates of treatment as well as nominal type I error.

A cadaveric precision and accuracy analysis of augmented reality–mediated percutaneous pedicle implant insertion

“…36,37 Additional potential pitfalls include errors with preoperative planning, shifts in the entry point and/or trajectory when traversed soft tissue is excessive, and instruments sliding on angled bony surfaces such as hypertrophied facet joints. [10][11][12][13]38 AR-HMDs with headset-integrated tracking cameras (Fig. 2A) and translucent direct retinal near-eye displays (Fig.…”

Impact of baseline covariate imbalance on bias in treatment effect estimation in cluster randomized trials: Race as an example

“…This goal can be difficult to achieve, and most CRTs do not implement design strategies to address this challenge. Extensive literature explores the random covariate imbalance and adjustment in individual RCTs and CRTs during the statistical analysis phase, especially for nonlinear regression analysis [25][26][27][28][29][30][31]. However, the use of these statistical adjustments is inconsistent.…”

Statistical fundamentals on cancer research for clinicians: Working with your statisticians

“…A non-inferiority study is always one-sided, thus addressing the chance of observing a difference as large as, and in the same direction, as that observed. The margin to be detected is usually also smaller (e.g., 5% in 5-year overall survival [OS] in the recent RTOG-1016 trial) and, therefore, a larger sample size is usually required [18] , [19] . Another example of a non-inferiority trial is NRG HN-002 (NCT02254278) which hypothesized that two treatment arms (reduced dose IMRT with or without weekly cisplatin) were non-inferior to the SOC of high-dose CCRT in low-risk minimal smoking HPV + OPC, where effectiveness was defined as 2-year progression-free survival (PFS) of ≥ 85% with a margin of 6% compared to SOC (assuming 2-year PFS for SOC is 91%), and without unacceptable swallowing toxicity at 1-year.…”