Clinical trial design for beginners: how to plan a successful study

This article is based on a webinar that I co-hosted titled Clinical Trial – Design or Disaster? The webinar was delivered in conjunction with my colleagues at Afortiori Development Ltd on behalf of Medilink East Midlands. You can re-watch the webinar here.

Clinical trial design for beginners: How to plan a successful trial

Trial design is a BIG topic, and quite a scary one too if you’ve never done it before. 

In my career, I’ve designed and analysed lots of clinical trials, in the NHS, academia, and commercial organisations. I thought it would be really helpful for me to share that experience with you. 

Here I’ve given a broad overview of the key points of trial design. I hope it will give you a clear idea of the things you need to consider, especially if you’ve never designed one before. 

What is trial design?

Designing a trial is the first step in working out the best way to collect the data that you want on your product (i.e. the drug, device, or other intervention that you’re testing). The design process is about understanding what you want to prove, and how. It will involve working out all the details about how you’re going to deliver your trial. 

From the outset, it’s important to work with stakeholders so that you’re designing research with the best chance of success. That includes ensuring that your trial will meet your commercial and regulatory requirements. 

involve your stakeholders when designing a clinical trial

Why is design important?

The benefits

Trial design is a vital and non-negotiable part of evidence-based medicine. There are so many benefits to putting the time in at the start to get it right. 

Firstly there’s the admin and documentation side of things. It makes sure that your study documentation and data will meet your regulatory requirements. You’re also less likely to need multiple protocol amendments if you’ve taken the time to do this properly at the start, which in turn will ensure a higher quality study. 

Secondly, in designing your study carefully, you’ll preempt any potential delivery issues. A good study design will save you time and costs because you’ll optimise patient numbers and endpoint measurements. 

Last but not least, it’s only ethical to recruit participants into a high-quality study. Without a rock-solid study design, you can’t be sure that you’re not putting the health and safety of your participants at risk. Keep your participants safe and happy, as well as your stakeholders, by putting in the time at the start of your study to design every last detail of the trial.

The stumbling blocks

If a trial is poorly designed, there are a number of possible outcomes that will negatively impact your results. 

First of all, you could end up with an under-powered study. Not having enough participants in the trial will make it more likely that you’ll end up with a statistically non-significant outcome. What this means is that you won’t have any evidence that your intervention works, and that’s what you want to try and avoid. The whole reason for organising the trial is because we believe the intervention will benefit people. 

You could fail to meet the endpoints of your study. Depending on the aim of your study, it might be difficult to secure regulatory approval for your intervention if this is the case.

You could even invalidate your claim or your study through poor trial design. For example, if you are testing an intervention for its efficacy in the adult population but are recruiting in a university town, you might end up with a very young adult population. This means you won’t have an evidence base for older adults. 

The most common problem that I see resulting from poor study design is inadequate data collection, because there are so many opportunities for error. For example, you could have inadequate CRF design, data collection tools, databases, data entry training, and lots more as well. Not only could this lead to a negative trial, it could even invalidate your whole trial. 

How do you design a trial?

There isn’t really a ‘right’ way to design a trial, nor is there a set path. Even though it can be a bit more like an art than a science, there are lots of useful approaches available. I usually advise taking a combination of the following options to get things started in designing your study.

It’s very rare to do a clinical trial design without some sort of literature review, so make sure you include a comprehensive literature review, and past or current trials review.

It’s a really good idea to do a pilot or a feasibility study as well. These are most useful if you’re doing a large trial, less so for small trials. At this point, you’re not testing your effect size, it’s more about how you would deliver your trial or your intervention.

Finally, you’ll want to do an in-depth review of potential payor considerations. Make sure that, at the end of the trial, you’ll have all the information you need to put your intervention into practice.

How to design a successful clinical trial
Photo by Kaleidico on Unsplash

Who to involve

You also want to source a key opinion leader in the field, and engage trusted clinical experts. They’ll be able to help you really understand the current standard of care and how your trial might integrate with that. 

Where possible, working with patient groups is a really good idea. They’ll be able to help you work through things like recruitment issues and concerns. They can also help to review your study documentation like your CRFs and your patient information leaflet. 

Another person you’d want to talk to in depth is a biostatistician (like me!). We help you to understand the end points, the impact on your sample size, and how representative your population is. 

A project management expert will also be a great help in uncovering potential delivery barriers and enablers based on their experience. Of course, the clinical team will also be able to support you with this.

Things to decide

You might be familiar with the PICOS framework. It’s a really useful tool for framing your research question.

It stands for the following:

Population

Which population are you testing your intervention for? And how many people will you need to recruit in order to achieve the minimum clinically important difference? A statistician will need to calculate this for you using a variety of different models.

Make sure you recruit participants that are representative of your intended claim and keep your study population as generalizable as possible. It’s also important to define your inclusion and exclusion criteria accordingly.

Intervention

You probably already know what this is, because that’s why you’re designing the trial, but there are still other important considerations to be made around it. For example the variant you might be using, and the healthcare setting where you will be running the trial.

Control/Comparator

It’s best practice to include a control in your trial, as it helps us to tell if any improvements are actually due to the intervention. The gold standard is a double-blinded, placebo-controlled study. If that’s not an option then keep your trial as blinded as possible by using a placebo or sham intervention. If you aren’t able to blind certain elements, make sure you can justify that decision. Everything apart from the intervention must be the same between the two groups.

Outcome

All endpoints need to be clinically relevant. That means they are measurable, able to detect a difference, and measured at a relevant time point. If you know that your intervention doesn’t start working for 2 weeks, you don’t want to be measuring your endpoints at 1 week. Conversely, it might be important to include longer-term follow-up, so make sure you include this in your trial design. Your endpoints, whether they are primary, secondary, exploratory, or safety, should also be sensitive enough to be able to detect a change.

Study Type

There are lots of different study types available, including (but not limited to) parallel, cross-over, cluster randomised, first-in-human, and adaptive. The decision of which type to use for your study will depend on a number of factors. For example, the condition you’re treating, the primary endpoint, the population size, and the type of intervention.

points to consider when designing a clinical trial

A final note

Remember, trial design is complex and takes a while to get right. No one person knows everything there is to know about it, so we’re all always learning. 


Want to learn more about how to design a successful clinical trial?

If you’d like to learn more about this and related topics in a supportive community of fellow researchers, why not check out my membership The Health Researcher Clubhouse? 

As a member, you’ll receive regular training, networking opportunities, and co-working invitations to help you work towards your goals. You’ll also be able to get support from myself and my co-founder Isabelle on regular Q&As and masterclasses. 

Join before 1st April 2021 and benefit from early bird pricing and enjoy all the benefits of being a founding member. Click here for more information and the sign up link. 

Related articles

If you found this article helpful, why not check out some more?

Leave a Reply

Your email address will not be published. Required fields are marked *