Survey of Clinical Trial Opinion and Preferences

1
Survey of Clinical Trial Opinion and Preferences

• The SurCTOP Study

2
SurCTOP

• Primary aims
• QuestionWhat are the factors leading to the
  low use of data-dependent designs?(Are
  inadequate, inappropriate and/or different
  understanding and opinions leading to the low use
  of data-dependent trial designs?)
• Data collection
• Survey on
• Patients (P)
• Healthy individuals (G)
• Clinicians (C)
• Health professionals (HP)
• Statisticians (S)
• (face-to-face, self-completed questionnaire)
  

• Excel database
• Problems/issues to consider
• Validity reliability of questionnaire
• (Exploring qualitative issues with
  semi-structured questionnaire)
• Generalizability of results
• Comparability of results between different groups
• Survey over the internet

3
In SurCTOP clinical trial designs are divided
into two broad categories
4
SurCTOP (The Questionnaire 1)

• Internal Validity
• The questions and answers
• Content validity
• (all versions)

• Question design
• IIQ(6) (all) - cost of drug development
• IIIQ(6) (S) - DDDs
• Force-answer strategy while allowing people
  to freely express their views on the
  questionnaire.
• Patient (C Brayne, W Hollingworth, G Yip)
• Clinician (S Griffin, J Benson)
• Statistician (T Johnson, T Prevost, E Pinto)
• Pharma (David of Pfizer)
• Regulator (KT Khaw , J Nickson, W Hollingworth)
• Health professional
• (combination of form C and S)

5
SurCTOP (The Questionnaire 2)

• Pilot (10 patients)
• Test-retest reliability (patients clinicians)
• External validity

• Face-to-face in ward D5 (P Weissberg)
• Revised after every two patients, 5 revisions
  made in total.
• Retest samples were from those who expressed an
  interest in getting a brief summary of results.
• 36 patients
• 44 clinicians
• Self-selective samples
• Patients/healthy people
• Clinicians
• Health professionals
• Statisticians

6
SurCTOP (Sampling size Response rates)
SurCTOP Results

• Patients/healthy people
• (patient arm of SurCTOP was then called
• ICTUA, LREC committee/Berrios G)
• Clinicians
• Health professionals
• Statisticians

• Ward D5 (P Weissberg), F6 (ER Chilvers), A4
  (Martin) and Oncolocy clinic (PG Corrie)
  patients/ supportive staff of IPH (FStr).
• RR - 5 (n110)/ 49 (n33)
• Clinicians in Addenbrookes H (H David), all
  levels.
• RR - 42 (n292)
• Health professional working/studying in IPH (R
  Himsworth, C Brayne) (Forvie Strangeway sites)
• RR - 47 (n74/158)
• Statisticians working in IPH (MRC
• Biostatistics Unit mainly) (S Thompson).
• RR - 74 (n37)

7
Past experience in trials (Proportions by
respondent categories)
SurCTOP Results
8
Needs for DDDs(summary of responses to the
preference questions)
SurCTOP Results
9
Needs for DDDs(Common cold)
SurCTOP Results
10
Needs for DDDs(Myocardial infarction)
SurCTOP Results
11
Needs for DDDs(Lung cancer)
SurCTOP Results
12
Needs for DDDs(Patients from cold to lung cancer)
SurCTOP Results
13
Needs for DDDs(Best proven therapeutic method in
trials)
SurCTOP Results
14
Needs for DDDs(Best proven drugs after trials)
SurCTOP Results
15
Reasonable time to market(Clinical phase)
SurCTOP Results
16
Needs for DDDs(Clinicians as clinicians as
patients)
SurCTOP Results
17
Differential Needs for DDDs(My preference is...)
SurCTOP Results
18
Understanding of DDDs
SurCTOP Results
19
Understanding of DDDs(Clinicians)
SurCTOP Results
20
Understanding of DDDs(Health professionals
Statisticians)
SurCTOP Results
21
DDDs to be used more frequently
SurCTOP Results
22
How long drug development takes?
SurCTOP Results
23
How long drug development takes? (Proportions
giving the right answer 15 years)
SurCTOP Results
24
How much it costs to develop a drug?
SurCTOP Results
25
How much it costs to develop a drug?
(Proportions giving the right answer 350 million
pounds)
SurCTOP Results
26
A trial looking for 20 treatment difference at
0.025 alpha level (one-sided) and 97.5 power
usingOBrien Flemming stopping rules with 5
interim analyses vs fixed-sample trial design
27
PROBABILITY MODEL and HYPOTHESES Two arm
study of binary response variable Theta is
difference in probabilities (Treatment -
Comparison) One-sided hypothesis test of a
greater alternative Null hypothesis Theta
lt 0 (size 0.025) Alternative hypothesis
Theta gt 0.2 (power 0.975) (Emerson
Fleming (1989) symmetric test) STOPPING
BOUNDARIES Sample Mean scale
a d Time 1 (N 77.83)
-0.3000 0.5000 Time 2 (N 155.67) -0.0500
0.2500 Time 3 (N 233.50) 0.0333 0.1667
Time 4 (N 311.34) 0.0750 0.1250 Time 5 (N
389.17) 0.1000 0.1000
28
Conclusions and Comments
SurCTOP Results

• There is a need for DDDs. Current 8 years of
  getting through clinical evaluation and
  regulatory approval falls short of the expected 4
  years.
• There is a general lack of awareness and
  understanding of DDDs, and also a lack of
  awareness of the time and cost needed to develop
  a new drug.
• More efforts are needed to bring about awareness
  of ethics and DDDs, agreement between different
  parties in trials, more application of DDDs in
  current and future trial practices, and a shift
  of attitudes towards these innovative trial
  designs.
• An awareness on the time and cost needed in drug
  development is a useful adjunct.

29
in particular need to THANK...

• Colleagues who helped in the content validity
  exercise of SurCTOP questionnaire.
• Colleagues who gave me the permission to approach
  the relevant units and study populations, also to
  the administration colleagues in the LREC
  committee, medical director, medical stuffing
  offices.
• Colleagues who provided generous, extensive and
  valuable feedbacks/comments after working through
  the questionnaire, and just to name a few
• (S Thompson, F Verne, S Bird, D
  Spiegelhalter, T Johnson, I White,
• A Prevost, A Raven, S Griffin, J Benson
  this list can continue and I must say that at the
  time of typing this slide I have remembered only
  the most recent, i.e. the health professional and
  statistician arms).
• Patients and colleagues who responded to SurCTOP
  twice for test-retest reliability study.
• All patients and colleagues participated in
  SurCTOP, also particular thanks to clinician
  colleagues in Addenbrookes H who supported us
  with a 42 response rate (n292).

30
Finally THANKS...

• Carol Brayne for supervising my MPhil, and also
  the many supports thereafter also Tony Johnson
  to allow access to the clinical trial database.
• Chris Palmer for making the past two years
  possible, and the many advice and suggestions
  also Simon Griffin for his valuable suggestions.
• Emily Wu and my relatives back home in Malaysia
  (mom/sister) for taking care of the children to
  allow me this LUXURY.
• Also to the many other friendly and helpful
  colleagues and friends around... (C Fuka, Ben,
  Claire, B McWilliams, B Tom, B Lan, E Pinto, D
  Pencheon, S Hickin, J Johnson, G Yip, J Luan)
  and

31

• also to SmithKline Beecham (now GlaxoSmithKline)
  who provided an unconditional grant to support
  the research.