Transcript NURS2001: Metabolism and its control

Overview of N-of-1 study suite – lessons
learned from three aggregated N-of-1 trials.
Geoff Mitchell*, Jane Nikles, Sue-Ann
Carmont, Janet Hardy, Phillip Good, Meera
Agar, Katherine Clark, Carol Douglas, Rohan
Vora, Hugh Senior, David Currow
Current N-of-1 Trials in PC
• Methylphenidate for fatigue in adv Ca
• Paracetamol for pain in adv Ca
• Pilocarpine for dry mouth in adv Ca
• (also known as single patient trials)
What are N-of-1 Trials
• Multiple Cross-over RCTs
• Provides evidence on treatment effect for an individual
patient
• METHOD: multiple paired treatment periods with random
treatment order, e.g.,
Benefits of N-of-1 trials
• Possible to aggregate a series of n-of-1 trials to get a
population estimate.
• N-of-1 like a trial where participant is in both arms of a
randomised controlled trial
•
•
•
•
Participant is:
Blinded to which arm is going at any one time
Contributes data to each side of an RCT
“control” and “intervention” participation perfectly matched.
When are n-of-1 trials justified?
• Medicine short acting
• Immediate treatment effect and short half life
• Treatment expensive, important side effects or
controversial
• Validated measure of effect available
• Question is important
Usefulness in palliative care
• RCTs very hard to conduct – very expensive
• Many fail because sample size is not met
• Aggregating N-of-1s trials could reduce sample size
dramatically, make the gathering of evidence easier
What Worked Well
• MPH exceeded recruitment target
• Caresearch Database was an efficient method of data
management
• PaCCSC network of sites and skilled research staff
facilitated:
–
–
–
–
–
Trial establishment
Liaison with pharmacy
Recruitment
Protocol Compliance
Data Integrity
What Didn’t Work Well
• Lower than anticipated recruitment rate and study
withdrawals due to:
– Very unwell participants
– Competing trials in a limited population
• Medications are not commercial, so had to be
compounded
• Delays due to complex ethics process, multiple ethics
applications and SSAs, and legal contract review by
lawyers
• Some patients did not like the taste of pilocarpine or it
produced over-salivation
What we would do differently
• We would continue with the Caresearch database and the
PaCCSC network
• Have more realistic timelines on trials that are difficult to
recruit in grant applications
• Ensure eligibility criteria is less restrictive if possible
• Look at different recruitment strategies including through
the community rather than solely hospital based
Results
• Paracetamol for pain pilot trial had major difficulty in
recruiting as participants had to not take paracetamol, 7
patients recruited, data still to be analysed
• Pilocarpine for dry mouth, 22 patients recruited, data still
to be analysed
• MPH for fatigue, see next slide
Methyphenidate vs placebo on fatigue (FACIT-F)
individual participant scores
• (Mean difference (95% credible interval))
• Did not improve fatigue on a population level,
• 8 of 33 participants clinically important improvement in fatigue,
• 1 showed important worsening