Adaptive designs, and other innovative approaches, provide great benefits but are also more complex to run. In 2019, the Adaptive Designs Working Group started investigating what extra resource Clinical Trials Units (CTUs) might need to support #adaptivedesigns. (1/7)
Funded by the @NIHRresearch@UKCTUNetwork CTU Support Fund and led by Newcastle CTU, the “Costing Adaptive Trials (CAT)’ project set about answering this question. Step 1 was a snazzy logo. (2/7)
We then did a mock costing exercise. Seven CTUs agreed to cost five trial scenarios - each based on a real trial. For each, we outlined a non-adaptive and adaptive version. CTUs returned the staff resource and other costs that they’d put in a funding application. (3/7)
After these were returned, we organised interviews with staff who did the costings. This was to explore reasons in variability between different types of adaptive design and to help us develop guidance. (4/7)
In January we held a meeting with co-investigators and CTU representatives to discuss the results. Now we’re in the final stretch of writing up the results and our recommended guidance (ETA end of March). (5/7)
While platform trials are clearly more prominent in COVID-19 than they were before the pandemic, many other COVID-19 trials use adaptive features. Some examples to follow. #adaptivedesigns
(1/4)
The MATIS trial (NCT04581954) is a multi-arm multi-stage trial with early stopping for lack of benefit which seeks to evaluate treatments to prevent more severe disease. @JMSWason
(2/4)
Similarly, the RECOVERY-Respiratory Support trial (ISRCTN16912075) seeks to identify optimal respiratory support using a multi-arm multi-stage structure. More details are here:
A number of COVID-19 trials use a platform structure and we will look at a few examples in the later stages of development in this thread. #adaptivedesigns
(1/6)
The first treatment to be shown to benefit COVID-19 patients was identified within the ACTT trial. The platform allows to stop the evaluation for benefit and lack thereof as well as adding additional treatments.
The RECOVERY trial (recoverytrial.net) is with 38,000 patients to date the largest COVID trial. This has allowed several questions to be answered including dexamethasone as an effective treatment for severe disease. @PeterHorby@MartinLandray@RichardHaynes3 (3/6)
Throughout this week we have made the case that adaptive designs can be useful to improve efficiency and discussed practical challenges. Two recent papers have made the case that trials that are adaptive can be particularly helpful in the of COVID-19. #adaptivedesigns (1/4)
Different types of adaptations and their utility for studies of COVID-19 treatments have been reviewed in
But the utility of adaptive designs is not limited to studies of COVID-19 treatments. They can also be useful for trials that are impacted by COVID-19 as argued here.
The TAIloR trial (ISRCTN: 51069819) was a multi-arm multi-stage clinical trial that investigated the utility of different doses of telmisartan to reduce insulin resistance in HIV-positive individuals. @thomas_jaki (1/n).
The trial had one interim analysis during which doses that were deemed insufficiently promising were dropped from the study and in the study two of the initial three doses were dropped at this point. (2/n)
Due to the ability to eliminate insufficiently promising doses, fewer patients were exposed to doses that did not provide benefit to patients. (3/n)
The NOTACS trial (ISRCTN: 14092678) an adaptive, multicentre, parallel group, randomised controlled trial comparing the efficacy, cost-effectiveness and safety of 2 types of oxygen therapy in patients at high risk of post-operative pulmonary complication after cardiac surgery 1/n
After a pre-defined number of patients have been recruited and completed follow-up, we will use the data accumulated so far to re-estimate the nuisance parameters and use these to repeat the sample size calculation. (3/n)
We’re now passing over to @DrGWheeler to walk us through software for adaptive designs!
Sorry, attempt 2 (@JMSWason not too experienced with adding gifs!)
Lack of available software is a known barrier to using better trial designs in practice. Q: “How bad is this problem?” 1/9 journals.sagepub.com/doi/full/10.11…
Our research questions were:
1. How many articles proposing new adaptive designs included code with the paper, or a link to code available elsewhere?
2. Which adaptive design approaches and features are well supported by current software, and (importantly), which are not? 2/9