Transcript Evaluating_Evidence

MIGRAINE IN PRIMARY CARE ADVISORS
Guildford, 1 May 2003, 2-6 pm
Understanding the evidence in evaluating
acute migraine medications in clinical
practice
Introduction and objectives
Dr Andrew Dowson
Kings’ Headache Service, London
Programme
• Dr Andrew Dowson: Introduction and
objectives
Discussion:
• Comparing acute medications: clinical trials
experience
• Comparing acute medications: experience
from clinical practice
• Practical advice for general use in the clinic
• Dr Andrew Dowson: Conclusions
Objectives
• Review clinical trial evidence for the clinical
profiles of triptans and other acute
medications for migraine
• Compare these data with the situation in
primary care
• Provide the practising clinician with rational
methods of evaluating clinical trial data
– Publications
– Detail aids
– Integrating clinical trial data into the real world
setting
Outputs
• MIPCA newsletter for GPs
– How to evaluate detail aids and trial data
• Slide set for educational use
• Article on understanding the evidence
for GPs
– Critical review of the evidence
Comparing acute medications: clinical
trials experience
Overview
• Clinical trial endpoints
• Post hoc endpoints
• Clinical trial data on the triptans and
other acute medications
• Misunderstandings and pitfalls
• Understanding the evidence
Clinical trial endpoints
Primary endpoint for clinical trials
• Headache relief: Improvement from
severe or moderate headache to mild or
no headache at 2 hours
– ‘Glaxo’ endpoint
• Improvement from severe or moderate
to no headache at 2 hours
– ‘IHS’ recommended endpoint
Pilgrim AJ. Eur Neurol 1991;31:295-9.
IHS Clinical Trials Subcommittee. Cephalalgia 2000;20:765–86.
Efficacy outcome
Pain relief (‘Glaxo’)
3
2
1
0
Severe
Moderate
Mild
None
Pain-free (‘IHS’)
Pilgrim AJ. Eur Neurol. 1991;31:295-299.
Secondary endpoints for clinical
trials
•
•
•
•
•
Headache relief at other time points
Headache free at various time points
‘Meaningful’ relief
Return to normal functioning
Relief of nausea, vomiting and photophobia /
phonophobia
• Sustained pain relief
• Headache recurrence
• Adverse events
Pilgrim AJ. Eur Neurol 1991;31:295-9.
Discussion: perspective on
endpoints
• In general, pain relief parallels that reported for other
symptoms
– Headache relief as a proxy for migraine relief?
• Stopwatch endpoints are used in pain studies, but not in
migraine studies
• If the patient treats the headache when mild, non-headache
symptoms may not have developed
– Potential problems with IHS diagnostic criteria
• Adverse events vs safety
– Tolerability (AEs) variable between patients
– Drug-related AEs – tolerability vs safety
– Side effects cloud use of drugs
• Patients may not differentiate migraine from other headaches
– Spectrum study indicates migraine may present with symptoms of
TTH
Discussion: study design
• Data need to be accurate
– Demonstrating reliability, power and validity
• Large study size increases power and clinical
relevance of the results
• Placebo responses may be different in different
settings
– Hospital vs primary care
• Manipulation of patient populations may be used to
obtain desired effects
– Relative numbers of patients and centres
• Different study designs may lead to differences in
results
– cf ‘Glaxo’ type sumatriptan studies with those used in
comparators with domperamol and tolfenamic acid
Post hoc endpoints
Post hoc endpoints
•
•
•
•
•
Meta-analyses
Therapeutic gain (TG)
Number needed to treat (NNT)
Number needed to harm (NNH)
Ad hoc analyses
– e.g. nausea
Goadsby PJ. Can J Neurol Sci 1999;26(Suppl 3):S20-6.
Efficacy: Therapeutic Gain (TG)
Therapeutic
gain
=
Proportion
of patients
benefiting
on treatment
—
Proportion of
patients
benefiting
on placebo
Goadsby PJ. Can J Neurol Sci. 1999;26(Suppl 3):S20-S26.
Efficacy: Number Needed to Treat
(NNT)
100
Number
Needed
to Treat =
Proportion
of patients
benefiting
on treatment
—
Proportion of
patients
benefiting
on placebo
Goadsby PJ. Can J Neurol Sci. 1999;26(Suppl 3):S20-S26.
Tolerability: Number Needed to Harm
(NNH)
100
Number
Needed to
Harm =
Proportion
of patients
with adverse
events
on treatment
—
Proportion of
patients with
adverse events
on placebo
Goadsby PJ. Can J Neurol Sci .1999;26(Suppl 3):S20-S26.
Insights into endpoints
Dr Shaun Kilminster
[email protected]
GCPL Pharmacology
: Royal Surrey County Hospital
Pain clinic
: Royal Surrey County Hospital
Lawrencian Clinical:
1) Short Pain Inventory
2) objective marker of pain
Society of Pharmaceutical Medicine.
Chairman
Headache measures vary in
accuracy
• Accuracy = Reliability x Power x
Validity
– Outcome measures need to exhibit these
features
• Importance on coefficient of variance
Face analogue scale
•Five-point scale favoured in most
therapy areas
•cf migraine 4-point scales
0
1
2
3
4
Visual analogue & Likerts
•Difficult to deduce clinical significance
from changes observed
•Need for simple, intuitive measures
NONE
EXTREME
PAIN
NNT based upon
% of patients achieving a
criterion
e.g.
% of patients with 50% reduction
in pain or no pain
Placebo response & NNT
•NNT varies with the placebo response
•Problematic in areas where a variable placebo
rate is likely, e.g. migraine
NNT
Mean pl ot
25
NNT75%
NNT25%
NNT55%
15
5
-5
-15
-25
.1
.2
.3
.4
PLACEBO RESPONSE
.5
Discussion: post hoc endpoints
• Post hoc endpoints (TG, NNT, NNH)
may not provide clinically relevant
information
• Placebo response affects NNT
• High therapeutic response → little effect
• Low therapeutic response → significant
effect
Discussion: 7 points for the critical appraisal
of clinical trial data and detail aids
What to look out for
•
•
•
•
•
At least 50 patients per treatment group
Patient numbers equal in active and placebo arms
NNT must have 95% confidence interval
Drop-outs from trial ≤ 10%
Drop-outs from trial must be same in active and
placebo arms
• Outcome measures use Likert, not visual analogue
scales
• Speed of onset, duration of analgesia and normal
coping / functioning well are needed for clinical
relevance.
Clinical trial data on the triptans
and other acute medications
Quality of clinical evidence
• Grade A: Consistent evidence from
multiple, well-designed, randomised
clinical trials, systematic meta-analyses
• Grade B: Poorer quality clinical trial
evidence; non-systematic metaanalyses
• Grade C: Opinion and practice in the
absence of objective evidence
Matchar DB et al. Neurology 2000;54.
Patients (%)
Comparing the oral triptans:
headache relief (Grade A)
90
80
70
60
50
40
30
20
10
0
Lowest
Highest
Suma Suma Zolmi Riza Ele 40 Almo Nara Frova
100 50 mg 2.5 mg 10 mg mg
12.5 2.5 mg 2.5 mg
mg
mg
Dowson AJ et al. Curr Med Res Opin 2002;18:414-39
Comparing the ODT and oral
triptans: headache relief (Grade A)
90
80
Patients (%)
70
60
50
Lowest
40
Highest
30
20
10
0
Zolmi 2.5 mg Zolmi 2.5 mg
oral
ODT
Riza 10 mg
oral
Riza 10 mg
ODT
Dowson AJ et al. Curr Med Res Opin 2002;18:414-39
Comparing the non-oral triptans:
headache relief (Grade A)
90
80
Patients (%)
70
60
50
Lowest
Highest
40
30
20
10
0
Suma 6 mg SC
Suma 20 mg NS
Zolmi 5 mg NS
Dowson AJ et al. Curr Med Res Opin 2002;18:414-39
Headache response at 2 h (% patients)
Head-to-head comparator trials with
oral triptans (Grade A)
Zolmitriptan versus sumatriptan
70
NS
NS
NS
60
50
40
30
20
10
0
Zolmi 2.5 mg
n = 500
Zolmi 5 mg
Suma 50 mg
n = 514
n = 508
Gruffydd-Jones K et al. Eur J Neurol 2001;8:237-45.
Headache response at 1 h (% patients)
Head-to-head comparator trials with
oral triptans (Grade A)
40
Rizatriptan versus sumatriptan (n = 1,268)
**
35
** p=0.01
30
25
20
15
10
5
0
Riza 10 mg
Suma 100 mg
Tfelt-Hansen P et al. Headache 2000;40:748-55.
Head-to-head comparator trials with
oral triptans (Grade A)
Pain free at 2 h (% patients)
50
45
40
Rizatriptan versus naratriptan (n = 522)
***
*** p<0.001
35
30
25
20
15
10
5
0
Riza 10 mg
Nara 2.5 mg
Bomhof M et al. Eur Neurol 1999;42:173-9.
Head-to-head comparator trials with
oral triptans (Grade A)
Headache relief at 2 h (% patients)
Almotriptan versus sumatriptan
70
NS
60
50
40
30
20
10
0
Almo 12.5 mg
n = 591
Suma 50 mg
n = 582
Spierings EL et al. Arch Neurol 2001;58:944-50.
Headache relief at 2 h (% patients)
Head-to-head comparator trials with
oral triptans (Grade A)
80
70
Eletriptan versus sumatriptan
NS
a. (n = 692)
***
60
b. (n = 2.113)
*** p<0.001
50
40
30
20
10
0
Ele 40 mg
Suma 100 mg
Ele 40 mg
Suma 100 mg
a. Goadsby PJ et al. Neurology 2000;54:156-63.
b. Mathew NT et al. Headache 2003;43:214-22.
Perspective on eletriptan
comparator studies
• Sumatriptan encapsulated in these
studies
– But marketed as film-coated tablet
• Studies show that sumatriptan
absorption is compromised on
encapsulation1
• Because of this, ABPI does not allow
Pfizer to use these comparative data in
its UK detail aids
1. Fuseau E et al. Clin Ther 2001;23:242-51.
Head-to-head comparator trials with
oral triptans (Grade A)
Headache relief at 2 h (% patients)
Triptans versus ergotamine plus caffeine
70
***
a. (n = 580)
60
b. (n = 733)
***
50
*** p<0.001
40
30
20
10
0
Suma 100 mg
Erg 2 mg
Ele 40 mg
Erg 2mg
a. Study Group. Eur Neurol 1991;31:314-22.
b. Diener HC et al. Eur Neurol 2002;47:99-107.
Perspective on triptan-Cafergot
comparator studies
• Oral ergotamine formulations have
poor bioavailability and suboptimal
efficacy
• Parenteral formulations of ergotamine
are superior to oral
– Rectal
– Injection
Matchar DB et al. Neurology 2000;54.
Head-to-head comparator trials with
oral triptans (Grade A)
Headache relief at 2 h (% patients)
Triptans versus aspirin plus metoclopramide
60
NS
a. (n = 358)
50
40
b. (n = 666)
NS
30
20
10
0
Suma 100 mg
A+M 900+10 Zolmi 2.5 mg A+M 900+10
mg
mg
a. Study Group. Eur Neurol 1992;32:177-84.
b. Geraud G et al. Eur Neurol 2002;47:88-98.
Conclusions from Grade A clinical
data - 1
• All oral triptans are effective and well
tolerated
– Naratriptan and frovatriptan seem to be less
effective than the other triptans
– Sumatriptan, zolmitriptan, rizatriptan, almotriptan
and eletriptan have similar efficacy profiles
• Differences are numerically small and of uncertain
clinical significance
• Subcutaneous and nasal spray formulations
are more effective and faster-acting than the
oral formulations
Conclusions from Grade A clinical
data - 2
Conventional clinical trial endpoints:
• Do not effectively distinguish triptans
from some other acute medications
– e.g. Aspirin plus metoclopramide
• Use artificial criteria that may not be
applicable in clinical practice
– Should / will patients wait till the headache
is moderate to severe before treating?
Caveats from Grade A evidence
• Suppressing of data
– Only positive studies published
• Manipulation of formulations
– Encapsulation issues in eletriptan studies
• Dose-response effects
– Naratriptan and frovatriptan not used at doses equivalent to
other triptans
– Some patients take up to 4 naratriptan tablets per attack to
achieve a satisfactory response
•
•
•
•
Suboptimal dosing schedules
Suboptimal endpoints
Diagnosis issues (migraine or CDH?)
Idiosyncratic responses
– Different results with rizatriptan in eastern and western
Europe
Post hoc comparisons between the
triptans (Grade B)
Meta-analyses
• Tfelt-Hansen et al
• Belsey
• Ferrari et al
• Economic analyses (Williams et al)
Comparing the triptans: Therapeutic
gain (Grade B)
Patients (%)
60
50
40
30
20
10
0
Suma Suma Zolmi Riza Almo Ele 40 Nara Frova
6 mg 50 mg 2.5 mg 10 mg 12.5 mg po 2.5 mg 2.5 mg
SC
po
po
po mg po
po
po
Tfelt-Hansen P et al. Drugs 2000;60:1259-87.
Efficacy: Therapeutic Gain for painfree 2 hours post-dose (Grade B)
Rizatriptan 10 mg
Eletriptan 80 mg
Zolmitriptan 5 mg
Eletriptan 40 mg
Rizatriptan 5 mg
Almotriptan 12.5 mg
Sumatriptan 100 mg
Sumatriptan 50 mg
Zolmitriptan 2.5 mg
Naratriptan 2.5 mg
Eletriptan 20 mg
0%
5%
10%
15%
20%
25%
30%
35%
Therapeutic gain
Belsey JD. J Clin Res. 2001;4:105-27.
Efficacy: Number Needed to Treat
for pain-free response (Grade B)
Rizatriptan 10 mg
Eletriptan 80 mg
Zolmitriptan 5 mg
Eletriptan 40 mg
Rizatriptan 5 mg
Almotriptan 12.5 mg
Sumatriptan 100 mg
Sumatriptan 50 mg
Zolmitriptan 2.5 mg
Naratriptan 2.5 mg
Eletriptan 20 mg
1
2
3
4
5
6
7
8
9
10
NNT
Belsey JD. J Clin Res. 2001;4:105-27.
Tolerability: Number Needed to Harm for
any adverse event (Grade B)
Zolmitriptan 5 mg
Eletriptan 80 mg
Sumatriptan 100 mg
Rizatriptan 10 mg
Zolmitriptan 2.5 mg
Rizatriptan 5 mg
Eletriptan 40 mg
Sumatriptan 50 mg
Eletriptan 20 mg
Almotriptan 12.5 mg
Naratriptan 2.5 mg
1
11
21
31
41
51
61
71
NNH
Belsey JD. J Clin Res.2001;4:105-27.
Integrating efficacy and tolerability
data (Grade B)
0%
Naratriptan
2.5 mg
Tolerability
(Therapeutic penalty)
Almotriptan
12.5 mg
5%
Eletriptan
20 mg
Sumatriptan
50 mg
10%
Zolmitriptan
15%
2.5 mg
10%
15%
Eletriptan
40 mg +
Rizatriptan
5 mg
20%
Rizatriptan
10 mg
25%
20%
30%
Sumatriptan
100 mg
35%
Efficacy
(Therapeutic gain)
25%
30%
Eletriptan
80 mg
Zolmitriptan
5 mg
35%
Belsey JD. J Clin Res. 2001;4:105-27.
Headache relief at 2 h (%)
Headache relief at 2 hours (Grade B)
80
70
60
50
40
25
50
100 2.5
Suma
5
Zolmi
2.5
Nara
5
10
Riza
20
40
80 12.5
Ele
Almo
Ferrari MD, et al. Lancet. 2001;358:1668-75.
Pain-free at 2 hours (Grade B)
Pain-free at 2 h (%)
50
40
30
20
10
0
25
50
100 2.5
Suma
5
Zolmi
2.5
Nara
5
10
Riza
20
40
80 12.5
Ele
Almo
Ferrari MD, et al. Lancet. 2001;358:1668-75.
Sustained pain-free (Grade B)
Sustained pain-free (%)
30
25
20
15
10
5
0
25
50
100 2.5
Suma
5
Zolmi
2.5
Nara
5
10
Riza
20
40
Ele
80 12.5
Almo
Ferrari MD, et al. Lancet. 2001;358:1668-75.
Any adverse event (%) - Placebo
Incidence of any adverse event:
Placebo subtracted data (Grade B)
40
30
20
10
0
-10
-20
25
50
Suma
100 2.5
5
Zolmi
2.5
Nara
5
10
Riza
20
40
Ele
80 12.5
Almo
Ferrari MD, et al. Lancet. 2001;358:1668-75.
Overall comparison (Grade B)
Drug Studied vs
Sumatriptan
2-Hour Pain
100 mg
Relief
Suma 50 mg
Suma 25 mg
Zolmi 2.5 mg
Zolmi 5 mg
Nara 2.5 mg
Riza 5 mg
Riza 10 mg
Ele 20 mg
Ele 40 mg
Ele 80 mg
Almo 12.5 mg
=
=
=
=
+
=/+
+(+)
=
24-Hour
Sustained
Pain-Free
(1 pill only)
Consistency
Across
Migraine
Attacks
Tolerability
=
=/=
=
=
+
=/+
+
+
=/=
=
=
++
=
=
+
=
+
=
=
++
=
=
=
=
++
Based on the results of the present meta-analysis and the direct comparator trials. = indicates no difference when
compared with sumatriptan. + indicates better when compared with sumatriptan. - indicates inferior when compared with
sumatriptan
Ferrari MD, et al. Lancet. 2001;358:1668-75.
Cost-effectiveness analysis
Cost-effectiveness ratio ($US)
180
Cost-effectiveness ratio per attack for sustained pain relief and
no adverse events
160
140
120
100
80
60
40
20
0
Almo 12.5 mg
Riza 10 mg
Suma 50 mg Suma 100 mg
Williams P, Reeder CE. Am J Manag Care 2003; in press.
Williams P, Reeder CE. Clin Ther 2003; in press.
Overview of meta-analyses of the oral
triptans (Grade B)
• All oral triptans were effective and well
tolerated
– More similarities than differences between
the drugs
• Rizatriptan 10 mg exhibited the best
efficacy
• Naratriptan 2.5 mg and almotriptan 12.5
mg exhibited the best tolerability
Discussion: perspective on the metaanalyses
• The different meta-analyses do not provide
consistent data
– Selection of studies, patient populations and
endpoints differ
– Bandolier could not repeat results from Ferrari et
al meta-analysis
• Timing of dosing important
• Power and timing of endpoints important
– Data dredging?
• ‘Economy with the truth’?
Acute treatments in clinical
practice versus clinical trials
Overview
•
•
•
•
Efficacy in clinical practice
Early treatment
Dose optimisation
Sensitivity of endpoints
Efficacy in clinical practice
• In general, triptans seem to be more effective
in clinical practice than in clinical trials
– Different populations of patients
• Greater proportion of women
• CDH patients may be included in trials
– Country differences
– Primary care versus secondary care
• Need for long-term naturalistic studies
conducted with prescription formulations
and doses
Sumatriptan in clinical practice
4-h response (% Attacks)
90
No. of patients = 338
80
70
60
50
40
30
20
10
0
Suma 25 mg
n = 285
Suma 50 mg
n = 2,053
Suma 100 mg
n = 1,522
Dowson AJ et al. IJCP 1999;Suppl 105:25-33.
Zolmitriptan in clinical practice
2-h response (% Attacks)
90
No. of patients = 2,499
80
70
60
50
Pain-free
Headache relief
40
30
20
10
0
Zolmi 2.5 mg
Zolmi 5 mg
Tepper SJ et al. Curr Med Res Opin 1999;15:254-71.
Rizatriptan in clinical practice
2-h response (% Attacks)
80
No. of patients = 3,953
70
60
50
Symptom-free
Headache relief
40
30
20
10
0
Riza 10 mg tab
Riza 10 mg
MLT
Usual therapy
Jamieson D et al. Headache 2003;43:223-30.
Early treatment: Optimising timing of
dosing
• Recent evidence indicates that triptans
work optimally if taken when the
headache is mild early in the attack
– Sumatriptan
– Zolmitriptan
– Rizatriptan
– Almotriptan
• Clinical and cost-effectiveness data
Zolmitriptan: Pain-free response at 2
hours treating mild headache
Zolmitriptan 2.5 mg
Placebo
Pain-free response
(% patients)
60
***
43
40
18
20
0
n=136
*** p<0.0001 versus placebo
n=no. of patients evaluated
n=141
Dose optimisation
• Patients may require lower or higher
than recommended doses to achieve
an optimal response
– Sumatriptan
– Zolmitriptan
– Naratriptan / frovatriptan
• Marketed doses may not be comparable due to
dose-response effects
Patient selection of optimal dose:
Sumatriptan
70
Patient preference after treating 6 attacks (n = 338)
Patients (%)
60
50
40
30
20
10
0
Suma 25 mg
Suma 50 mg
Suma 100 mg
Dowson AJ et al. IJCP 1999;Suppl 105:25-33.
Patient selection of optimal dose:
Zolmitriptan
70
Patient choice of 2.5 mg and 5 mg doses (n = 2,499)
Attacks (%)
60
50
40
30
20
10
0
Zolmi 2.5 mg
Zolmi 5 mg
Tepper SJ et al. Curr Med Res Opin 1999;15:254-71.
Sensitivity of endpoints
• For regulatory purposes, clinical trial
endpoints are designed to show
significant differences between active
drug and placebo
– They may be relatively insensitive in
distinguishing between two active drugs
• They are probably not appropriate for
use in clinical practice, where more
sensitive endpoints may be required
Alternative endpoints
•
•
•
•
Patient preference
Impact questionnaires
Cost effectiveness
Qualitative endpoints
Patient preference
• Global measure of efficacy and
tolerability
• In studies, patients consistently prefer
oral triptans over non-triptan acute
medications
• Patients also prefer individual oral
triptans over other oral triptans
– But responses may be idiosyncratic
Sumatriptan versus ‘usual’ nontriptan therapies
80
No. of patients = 402
70
Patients (%)
60
50
40
30
20
10
0
Sumatriptan 50 mg
Non-triptan
No preference
Kwong WJ et al. Cephalalgia 2001;21:411.
‘Triptans’ versus ‘usual’ non-triptan
therapies
60
No. of patients = 663
50
Patients (%)
40
30
20
10
0
Triptans
Non-triptans
Both
No preference
Robbins L. Cephalalgia 2001;21:406.
Sumatriptan 50 mg versus
zolmitriptan 2.5 mg tablets
50
Patients (%)
45
40
No. of patients = 94
35
30
25
20
15
10
5
0
Suma 50 mg
Zolmi 2.5 mg
No preference
Pascual J et al. Cephalalgia 2001;21:680-4.
Sumatriptan 50 mg tablets versus
rizatriptan 10 mg ODT
60
No. of patients = 374
**
50
Patients (%)
40
** p<0.01
30
20
10
0
Suma 50 mg
Riza 10 mg
Loder E et al. Headache 2001;41:745-53.
Zomitriptan ODT versus sumatriptan
tablet
70
Patients (%)
60
*
60.1
Zolmitriptan 2.5 mg ODT
50
40
39.9
Sumatriptan 50 mg tablet
30
20
10
0
* p<0.05 versus sumatriptan tablet
Zolmitriptan ODT versus rizatriptan
ODT
Respondents (%)
80%
Prefer Zomig
Rapimelt
70%
60%
40%
Prefer
Rizatriptan MLT
27%
No preference
20%
3%
0%
Source: CIMA Patient Preference Study 2001 (West Pharmaceutical Services)
Overall reasons for preference
• Rapid relief
• Effective relief
Discussion: patient preference
• May not be sensitive to subtleties of
pain
• Some lapsed consulters said they
preferred OTC drugs to triptans
Impact questionnaires
• MIDAS
• HIT
• Only MIDAS has so far shown
sensitivity to change
• SPI
Efficacy of zolmitriptan 2.5 mg tablet
assessed with MIDAS
40
No. of patients = 1,972
MIDAS score
35
30
25
20
15
10
5
0
Baseline
After treatment
Torres G et al. Poster at 53rd AAN, 2001.
Change in MIDAS following primary
care interventions
16
No. of patients = 19
MIDAS score
14
*
12
10
8
* p<0.05
6
4
2
0
Baseline
6-mo care
Main A et al. Curr Med Res Opin 2002;18:471-8.
Short Pain Inventory©
• Developed by Dr Shaun Kilminster
• A 17-item self rating questionnaire
• ‘measuring the whole pain disturbance’
• Potential new endpoint for headache
studies
SPI© subscales
•
•
•
•
•
•
Pain severity
Social interaction
Anxiety
Anger
Sadness
Sedation
• TOTAL PAIN DISTURBANCE 17 items
• TOTAL MOOD DISTURBANCE 14
items
SPI© subscales that correlate directly
with level of headache severity
Mood changes
• Social interaction
• Sedation
• Sadness
• Anger
• Anxiety
• Total mood disturbance (TMD)
SPI©: correlation of Total Mood
Disturbance and Headache Pain (n = 75)
SPI: TOTAL MOOD DISTURBANCE (TMD)
55
+1.96*SE
50
+SE
Mean
45
-SE
40
-1.96*SE
35
30
25
20
15
10
0
1
2
3
HEADACHE SEVERITY
4
SPI©: Conclusions
• How the patient feels about their headache
impacts on mood and vice versa
• SPI is a reliable, valid and discriminatory
measure of headache severity
• SPI has high potential utility in headache
research
– Likely to be a highly-sensitive endpoint for clinical
trials
– Naturalistic studies in clinical practice underway
Cost-effectiveness
• Compared to placebo and non-triptan
acute medications, triptans:
– Improve quality of life
– Increase workplace productivity
– Reduce healthcare costs
Rizatriptan versus usual therapies:
improvements in QOL
24-h Migraine-Specific QOL Questionnaire (n = 265)
16
14
*
*
**
***
**
Rizatriptan 10 mg
Usual medications
12
10
* p≤0.05
8
** p<0.01
6
*** p<0.001
4
2
0
Work
funct
Social
funct
Energy
Symptoms
Feelings
Gerth WC et al. Clin Drug Invest 2001;21:853-60.
Workplace productivity loss following
treatment of migraine with sumatriptan or
placebo
Time/day (min)
80
70
*** p<0.001
60
50
40
***
30
20
10
0
Sumatriptan 6 mg sc
Placebo
(n = 76)
(n = 40)
Schulman EA et al. Mayo Clin Proc 2000;75:782-9.
Cost-effectiveness analysis in Canada
• Sumatriptan tablets versus ergotamine
plus caffeine tablets
• Economic benefit in favour of
sumatripan
– $98 saved per attack aborted
– Cost-utility ratio = $29,366 per QALY
Evans KW et al. Pharmacoeconomics 1997;12:565-77.
Qualitative endpoints
Michele Peters
University of Surrey, Guildford
Qualitative research
• Objective: to increase the
understanding of patients and
physicians
• Processes through interviews
– Decision trees
– Working through attacks
– Patient-doctor behaviour and its evolution
over time
– Explaining the variations in patient
responses
Qualitative research in clinical trials
programmes
• MRC advises the use of qualitative
research in clinical trials
• Phase I (modelling, not preclinical)
• Pre- classic Phase II, III and IV clinical
trials
• Study of behaviours:
– Doctor-nurse
– Doctor-patient
MRC. A framework for development and evaluation of RCTs for
complex interventions to improve health; 2000.
What can qualitative research
investigate?
• Why does an intervention have
therapeutic benefit?
– Testing underlying assumptions
– Effect of inappropriate beliefs
– Active and non-active elements of
intervention
– Groups of patients likely or not likely to
respond
• Mitigates against no effect results
MRC. A framework for development and evaluation of RCTs for
complex interventions to improve health; 2000.
Qualitative research: methodology
•
•
•
•
Group interviews (focus groups)
Individual in-depth interviews
Observational research
Organisational case studies
• Quantitative surveys may also be used
MRC. A framework for development and evaluation of RCTs for
complex interventions to improve health; 2000.
Qualitative endpoints
• Pilot studies with small numbers of
people can increase understanding of
issues
• Interviews can give insight into validity
of likely trial results
– Patient compliance
– Importance of placebo response
• Cf:
– Interviews are qualitative
– Factor analysis is quantitative
Qualitative endpoints
• New way of analysing headache issues
• Investigating patient feelings
• Studies needed on how to deliver care
– PCT clinics
– GPs with a special interest in headache
(GPWSIH)1
1. Department of Health. www.doh.gov.uk/pricare/gpspecialinterests/headache.pdf
Summary
• Clinical trials may not reflect clinical
practice due to different:
– Populations of patients
– Treatment modalities used
– Timings of treatment
– Primary versus secondary care sites
• Clinical trial endpoints often do not
differentiate between treatments
Future needs
• Qualitative research to set the agenda
– Patient-doctor behaviour
– Doctor-nurse behaviour
• Study the whole range of management
options
– Pharmacological and non-pharmacological
• Conduct naturalistic studies
• Develop more sensitive endpoints
• Develop an objective test for pain
– Blood test?