Trial oversight groups

Question 1

what different groups are involved in a clinical trial

Answer 1

• trial management group
• data monitoring committee
• trial steering committee
• endpoint review committee

Question 2

What do the TMG do (trial managment group)

Answer 2

oversee the day-to-day management and overall conduct and progress of the trial

Question 3

what does the TMG input into

Answer 3

• trial specific documents: protocol development, CRF, PIS, consent form, SAP,
• monitor protocol deviations
• make decisions about any changes to the protocol

Question 4

what does the trial steering comittee do (TSC)

Answer 4

• supervise overall trial amd make sure its acting in accordance with GCP
• Agree the trial protocol and any protocol amendments
• Agree terms of reference/charter
• Monitor trial progress: recruitment, data completeness,
  losses to FU, adherence to the protocol, participant safety
  and consideration of new information of relevance to the
  trial
• Give advice to the investigators on all aspects of the trial

Question 5

Who makes up the TSC

Answer 5

• independent chair, a majority of independent voting members and a public/patient representative
• CI and one or two other investigators (non-independent)
• Relevant members of TMG may be asked to attend TSC to
  present information

Question 6

when should TSC be timed?

Answer 6

useful to have these after TMG meeting

Question 7

what things are involved in a TSC progress report?

Answer 7

• Summary of trial design
• Sample size
• Summary of site information
  o Local investigator
  o Date of site open
  o Date 1st patient randomised
  o Number of patients randomised
  o Sites to be opened
• Recruitment update
  o By site
  o Projection chart
• Patient flow summary (combined group data only)
• Screening information
• Completeness of data collection
• Trial milestones

Question 8

what is the data monitoring comittee

Answer 8

a group established by the sponsor to assess at intervals the progress of a clinical trial, safety data, and critical efficacy variables

Recommends to sponsor whether to continue, modify or terminate a trial.

Question 9

how many people make up the DMC

Answer 9

few n, (minimum size of 3)

have to be completely independent

Question 10

what is a DMC charter?

Answer 10

something that formalises how DMC should operate

signed by all members of a DMC

one is called “DAMOCLES”

Question 11

What is the DMC report - open

Answer 11

updates of study analysed as a whole

Summary of trial design
* Sample size
* Summary of site information
o Local investigator
o Date of site open
o Date 1st patient randomised
o Number of patients randomised
o Sites to be opened
* Recruitment update - can be visually presented in a recruitment chart / recruitment projection chart
o By site
o Projection chart

Question 12

what is DMC report - closed

Answer 12

update of study analysed per group
* Data by randomised group
* Unblinded?
* Prepared by statisticians
* May or may not include interim analysis
* Based on DMC analysis plan

Question 13

what does the DMC make reccomendations about after folllowing the data at intervals

Answer 13

• Continue / Stop?
• Should we stop trials early?
  – Treatment benefit
  – Safety
  – Lack of treatment effect - futility

Question 14

What is an interim analysis?

Answer 14

• Any analysis intended to compare treatment arms with
  respect to efficacy or safety at any time before the final
  analysis
• Problems
  o Multiple testing (inflate Type I error)
  o Accumulated data
  o Stopping too early
  – For efficacy could overestimate the treatment effect
  (Bassler et al, JAMA 2010)
  – For futility could introduce Type II error

Question 15

statistical methods to stop a trial at interim analyses

Answer 15

> group squential designs
stochastic curtailment

Question 16

when do we stop a study?

Answer 16

• Good practice to pre-specify
• Stopping for benefit needs “proof beyond reasonable
  doubt”
• Decision to stop also requires clinical, ethical judgement

“No single statistical test or monitoring procedure ought
to be used as a strict rule for decision-making, but rather
as one piece of evidence to be integrated with other
evidence.”

Question 17

group sequential designs

Answer 17

repeated analyses throughout the trial (basically a few of interim analyses) and using the results of these we can decide to stop the study when efficacy vs futility it detected

Question 18

group sequential boundaries - different methods and when to stop

Answer 18

so we have the standardized normal statistic (z) and depending on which method we use, different stopping boundaries are used

• Fixed nominal (Pocock) method
  o Stop at i’th interim analysis if |zi| > z-statistics
• O’Brien-Fleming method
  o Stop at i’th of N interim analysis if |zi| > z of x sqr route of 𝑁 / 𝑖
• Haybittle-Peto method
  o Stop at i th of N-1 interim analysis if |zi | > 3 or
  o For the last interim analysis, n, stop at |zn | > z-statistics
• Alpha Spending Function (DeMets & Lan, Stat Med 1994)
  and many more

Question 19

stochastic curtailment

Answer 19

• What is the probability that the study would yield a
  significant result (two-sided) in favour of the intervention
  arm if the study were to be continued for another year,
  given what has been observed to date?
  o Conditional power
  o Bayesian predictive probability
• Can be applied to monitoring for futility