Send the link below via email or IMCopy
Present to your audienceStart remote presentation
- Invited audience members will follow you as you navigate and present
- People invited to a presentation do not need a Prezi account
- This link expires 10 minutes after you close the presentation
- A maximum of 30 users can follow your presentation
- Learn more about this feature in our knowledge base article
Intervention Study Analysis
Transcript of Intervention Study Analysis
Ana Sosa, Connor Hawkins, Michael Peters
Intervention Study Analysis
Is sufficient evidence presented to justify the study?
28 million adults suffering with migranes in the US alone
most use some sort of medication to treat symptoms
1998: drug containing acetaminophen, aspirin and caffeine was the first OTC approved by FDA to treat migranes
There was no prior study published that directly compared the use of OTC (e.g. acetaminophen) vs prescription drugs (sumatriptan) in the treatment of migrane in adults.
It NEEDED to be done.
Were all the relevant outcomes reported?
Yes and no
Related to the intent of study: Yes
With and without aura
Clinically relevant: Perhaps
Vomiting more than 20% of the time
50% time in bedrest
Do the inclusion and exclusion criteria match the goals of the study?
Depends on theoretical interpretation. The authors do not explicitly state that the excluded individuals represent a different diagnosis.
If the reader infers that eg co-morbid emesis represents a sufficiently different diagnosis, then exclusion is justifiable. If your theoretical stance is that it is part of the same diagnosis, then exclusion is not justifiable. Note that different classification systems for migraines exist.
Is there a clear statement of the purpose of the study?
The purpose of the study is to
address the need
and thus conduct a study to directly compare AAC (acetaminophen/aspirin/caffeine - most common OTC drug) and sumatriptan (prescription) for early treatment for migranes.
Statistical tests and Significance
Statistical Significance: P ≤0.05
Chi-squared Test: categorical variable
Wilcoxon Rank Sums & Nonparametric Survival Techniques: Time for meaningful pain relief
Cochran-Mantel-Haenszel Test: Used to compare treatment groups
How Many Tests of Hypothesis Appeared?
Does Size Matter?
Possible Beta Error?
Power of the test= 1-beta
Were the Differences Clinically Significant?
Migraine history of study population consistent with general migraine population
Commonly prescribed treatments
Excluded subjects who vomited more than 20% of the time
50% of time in bedrest
Encapsulation of Sumatriptan
Were the Patients Analyzed so Generalization Can be Made?
Excluded vomiting and bedrest patients
All medications were encapsulated
Was the Intervention Performed in a way that may be Generalized?
Small placebo group
Sumatriptan no better than placebo
Absolute S50 treatment was consistent with literature
Did the authors consider sample size prior to experimentation, and was the sample size sufficient?
70 per active treatment group
80% power to detect a difference of 25% of population
Allowed for two-sided with alpha=0.05
Outcomes Sufficient to Guarantee the Greatest Therapeutic Good?
No magic bullet, so why ask for a guarantee?
"Significantly superior" to S50 and placebo
Funded by Bristol-Myers Squibb
Is there a clear statement of the study hypothesis?
Do the authors account for every patient who is eligible for the study but does not enter it?
Caucasian individuals were overrepresented in the study (82% white).
Is the baseline comparability of the treatment and control groups documented?
Yes! Table 3.
What is the intervention?
One group was given two tablets of a combination of acetaminophen (500mg), aspirin (500mg) and caffeine (130mg), another group was given 2 tablets of sumatriptan (25mg per tablet) and a third is given a placebo. Patients were distributed between the groups in a 2:2:1 ratio.
Participants took the tablets when they started to assess a migraine coming on, and then answered a series of questions at designated time points.
Were contamination and co-intervention considered?
Yes, briefly. The experimenters noted whether or not a patient used "rescue medication" during the migraine.
Were withdrawals, dropouts, cross-overs and poor compliers analyzed in accordance with the aims of the study?
Not completely. As mentioned, 188 subjects were randomized, yet only 171 participated, and the explanation for this discrepancy was not discussed. Also discussed, the experimenters had a very limited ability to assess, and thus analyze for compliance.
What outcome measures were utilized?
Measure at fixed time points:
Subjective report on headache pain intensity (scale of 0-3)
Headache pain relief (0-4)
Functional disability (0-4)
Associated gastrointestinal and neurological symptoms (0=no, 1=yes)
Measured independent of time point:
Onset of meaningful migraine relief
Subject global assessment on medication effectiveness (5 point scale)
Investigator global assessment (same 5 point scale)
Usage of rescue medication
Is the intervention clearly defined and replicable?
The intervention is clearly defined, however one serious question about replicatability is in the subjective nature of pain assessment, especially considering the racial bias in the sample. Cultural factors have been shown to influence pain perception and interpretation (Lasch 2000), so a more diverse group of patients may not have the same results. For example, the concept of pain onset may vary, resulting in a different timing of medication administration.
Was compliance with intervention(s) measured and were non-compliers analyzed correctly?
The only measurement of compliance was completion of the journal. One individual lost their journal, so they were removed from the study. A more appropriate treatment would have been to have them repeat the experiment with a new journal.
As the study was intended to have high external validity, the necessary trade off was a decreased ability to monitor compliance.
Is it clearly outlined whether the study is considering EFFICACY or EFFECTIVENESS?
"We expect that the
of AAC would be similar to that of S50, because in four studies of AAC in treatment of migraine, its absolute 2-hour response rates have been comparable to those that have been reported with S50."
were taken when evaluating the effectiveness of the medication via diaries
headache pain intensity (PI
), 4-point scale
0 = no pain
3= severe pain
headache pain relief (PAR)
, 5-point scale
0 = no relief
4= complete relief
y, 5-point scale
ability to preform daily activities (0)
completely impaired (4)
associated gastrointestinal and neurologic symptoms
No = 0 | yes = 1
Were all patients who entered the study accounted for?
188 patients were randomized, but only 171 participated.
The ones that participated were sufficiently accounted for, including the one who lost their notebook.
A specific explanation for the remaining 17 was not given, but should have been.
"We expect that the efficacy of AAC would be similar to that of S50, because in four studies of AAC in treatment of migraine, its absolute 2-hour response rates have been comparable to those that have been reported with S50."
If it is a controlled trial, is the allocation of subjects TRULY randomized?
It was a "randomized , double-blind, placebo controlled clinical-trial"
Computer generated randomization
subjects randomly assigned to AAC, Sumatriptan or placebo treatments
double-blinded medication (gelatin capsules)
Was prognostic stratification used?
Is the population from which the sample is drawn CLEARLY described?
No. What was described was the size of the sample more than the population from which the sample was drawn.
However it was noted that the sample population studied was consistent with that of the general migraine population
Sample size in a 2:2:1 ratio for each treatment used: AAC, Sumatriptan and Placebo respectively (intended to be 70/70/35)
Are inclusion and exclusion criteria specified and replicable?
Inclusion (eligibility) = subjects that had an average of 1-8 migraine episodes/month (International headache society approved) of at least moderate intensity.
Exclusion = subjects who reported vomiting in more than 20% of migraine episodes or needed bedrest in more than 50% of migraine episodes.
(this bit was actually replicated from studies in acetaminophen)
Not prognostic stratification but some stratification on age, sex.
Our assessment of the intervention study = GOOD