FDA draft guidance published this month says you should. In most cases, adjusting for covariates is not necessary. Randomization generally insurers that covariates are balanced across clinical trial arms. Randomization, however, may not always result in perfectly balanced trial arms. In these cases, the FDA… Read More »Should you adjust for covariates when analyzing data from randomized controlled trials?
A recent JAMA paper by Yadav and Lewis (2021) provide the answer: Bias from immortal time periods is the error in estimating the association between the exposure and the outcome that results from misclassification or exclusion of time intervals Yadav and Lewis (2021) Sounds simple… Read More »What is immortal time bias?
How do you evaluate treatment efficacy and safety outside of the clinical trial setting? This is not just a question of academic interest. In last week’s JAMA, Rubin 2021 writes about some of the challenges of evaluating COVID-19 vaccines outside of a clinical trial setting.… Read More »Use of real-world evidence for regulatory decision-making
[Disclaimer: this post provides the math behind how to answer this question without actually answering the question] If you get a positive test for COVID-19, how likely are you to have the disease? Well, this depends on a number of factors including the baseline prevalence… Read More »If you have a positive COVID test, how likely is it that you actually have COVID?
Is your observational research study following best practices? Is your methodology transparent? To help answer these questions, the Enhancing the QUAlity and Transparency Of health Research (EQUATOR) Network created the STROBE guidelines. The STROBE guidelines–an acronym for The Strengthening the Reporting of Observational Studies in… Read More »Did your real-world study follow STROBE guidelines?
Population-adjusted indirect comparisons (PAICs) include both matching-adjusted indirect comparisons (MAICs) and Simulated treatment comparisons (STCs). The key data requirement for these methods is that they have individual patient data (IPD) from at least one clinical trial. This means the methods are most useful for studied… Read More »When (and how) to use population-adjusted indirect comparisons?
That is the title of a helpful video from Nicholas Lattimer of University of Sheffield. You can view all his videos here.
In many clinical trials, the outcome of interest may be some form of time to event outcome. This could be time until death, time until hospitalization, or time until some other (typically negative) health event. One simple way to model this is to use a… Read More »Can I use a Cox proportional hazard model to treatment effect?
How do you measure the value value of new treatments that improve survival? Clearly one of the key factors for doing so is understanding how much each treatment can improve survival relative to the status quo. In practice, however, estimating this quantity is challenging in… Read More »Estimating survival gains based on clinical trial data