Share this page!

Enter URL or ID to Unroll

You can paste full URL like: https://x.com/threadreaderapp/status/1644127596119195649
or just the ID like: 1644127596119195649

How to get URL link on X (Twitter) App

  1. On the Twitter thread, click on or icon on the bottom
  2. Click again on or Share Via icon
  3. Click on Copy Link to Tweet
  4. Paste it above and click "Unroll Thread"!
  5. More info at Twitter Help
Marion Campbell Profile picture
@MarionKCampbell
Apr 17, 2023 • 11 tweets • 5 min read • Read on X
Scrolly
This week some of my discussions have centred on #ClusterTrials. Cluster trials involve the randomisation of intact units (wards, hospitals, GP practices etc) rather than individuals. They have a number of key elements that must be accounted for 1/11
#MethodologyMonday
There are very good reasons for cluster/group randomisation eg when evaluating interventions like clinical guidelines or educational interventions which apply at practice/hospital level; or when there is potential of contamination of the intervention across trial groups 2/11
However, cluster randomisation has some major impacts for design & analysis primarily because observations within a cluster are not independent (outcomes are likely to be more similar within a cluster) 3/11
This intra-cluster correlation - measured by the intracluster correlation coefficient (ICC) - needs to be accounted for or trial results will appear more “significant” than they actually are. 4/11
Sample size needs to be inflated to account for this clustering. In its simplest form a std sample size needs to be inflated by a multiplier - the design effect - which accounts for the cluster size and the ICC 5/11
academic.oup.com/ije/article/44… Image
Estimates of the ICC are needed. There are published estimates to help for different fields. Some are given below 6/11

Implementation research: journals.sagepub.com/doi/abs/10.119…
Primary care: jclinepi.com/article/S0895-…
Schools: jclinepi.com/article/S0895-…
Consent can also be different in cluster trials as many interventions are not targeting patients directly. Some cluster trials therefore seek a waiver of consent. The Ottawa statement on cluster trials is a great read on consent & ethical issues 7/11
journals.plos.org/plosmedicine/a… Image
Selection bias can be an issue in cluster trials as usually treatment allocation is known ahead of patient recruitment. Awareness of allocation can lead to biased recruitment in the different arms of the trial. It is important to mitigate against this 8/11
bmcmedresmethodol.biomedcentral.com/articles/10.11… Image
Process evaluations also need to think about the extra dimension added by the clustering. Aileen Grant et al discuss some of the issues which arise in the following paper 9/11
trialsjournal.biomedcentral.com/articles/10.11… Image
Analysis also needs cluster-tailored methods to ensure the results are adjusted to account for the ICC and don’t over amplify the significance of the results. 10/11
trialsjournal.biomedcentral.com/articles/10.11… Image
There are now neat variants on the cluster design including the cluster-crossover design (CRXO) which is a particularly efficient design in this space 11/11

bmj.com/content/371/bm… Image

• • •

Missing some Tweet in this thread? You can try to force a refresh
　

Keep Current with Marion Campbell

Marion Campbell Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @MarionKCampbell

Marion Campbell Profile picture
May 13
A popular, but often misused design, is the #Crossover trial design. But what are the key things to look out for if you are considering using it? 1/9 #MethodologyMonday #87
In a crossover trial each participant receives two (or more) treatments in a random order. The most common design is an AB/BA design (2 treatment, 2 period design) which randomises half the sample to receive treatment A first then B and the other half to B first then A. 2/9
Because each person acts as their own control this removes a large element of variation, making the design more powerful than a standard parallel group study 3/9
Read 9 tweets
Marion Campbell Profile picture
Mar 25
In clinical trials, we sometimes undertake #SensitivityAnalysis alongside the main primary analysis. But when should we use them and to what purpose? 1/8
#MethodologyMonday #80
Sensitivity analyses can assess the impact of key elements/assumptions of a trial on the result eg impact of any baseline imbalance, impact of the choice of analysis approach etc 2/8
If the different sensitivity analyses provide similar results one is reassured that the trial result is robust and thus the credibility of the trial findings are increased 3/8
Read 8 tweets
Marion Campbell Profile picture
Mar 4
I have spoken about “usual care” or “treatment as usual” as a control arm in trials before, but should you ever protocolise usual care or just measure it as is? 1/8
#MethodologyMonday #77
Whilst “usual care” implies a common package of care being applied across sites, there is often a high degree of heterogeneity in care provided - but many would argue that the heterogeneity will increase the external validity of the trial results 2/8
However, heterogeneity in the usual care control group may affect the internal validity of the trial. It can affect the effect size and can make the trial result hard to interpret. 3/8
Read 8 tweets
Marion Campbell Profile picture
Jun 5, 2023
I have spoken about the importance of minimally clinically important differences (#MCID)
before in relation to sample sizes but how do you decide what it should be? There was an interesting paper published this week adding to this literature 1/8
#MethodologyMonday
The MCID drives the sample size - it is the minimum clinically important difference you set your trial to detect. Set the MCID too small and the sample size will be much larger than needed; but make the MCID too big & your trial will miss clinically important effects 2/8
There are different methods of calculating #MCIDs - the DELTA project is a great resource in this regard. 3/8
DELTA: journalslibrary.nihr.ac.uk/hta/hta18280/#… Image
Read 8 tweets
Marion Campbell Profile picture
May 29, 2023
While 1:1 randomisation to interventions is most common in clinical trials, sometimes #UnequalRandomisation is used. There are a number of factors that influence which randomisation ratio to use 1/9
#MethodologyMonday
One justification for unequal randomisation is when there is a substantial difference in cost of treatments. In this scenario, randomising unequally with fewer to the very expensive arm maximises efficiency when a trial has set resources 2/9
bmj.com/content/321/72… Image
Another is if you are undertaking an early investigation of a new treatment and need to have greater insights into its underlying safety/benefit profile. Here, increasing allocation to the new treatment will provide greater precision around these estimates 3/9
Read 9 tweets
Marion Campbell Profile picture
Apr 24, 2023
One phenomenon that can affect clinical trials is the #HawthorneEffect. This is when purely being involved in a trial can improve performance. 1/9
#MethodologyMonday
The #HawthorneEffect was named after a famous set of experiments at the Hawthorne Western Electric plant, Illinois in the 1920/30s. 2/9
In one experiment lighting levels were repeatedly changed & with each change, productivity increased .. even when reverting to poorer lighting. This was attributed to workers knowing their work was being observed. Productivity returned to normal after the experiments ended 3/9
Read 9 tweets

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us!

Send Email!

:(