Table of Contents
|
Response from PI to Lancet Peer Reviews (by email)
From: Wayne Taylor
To: Marie-Clare McMenemy (Marie-Clare McMenemy)
Sent: Sunday, June 06, 1999 11:22 PM
Subject: Lancet FAST-TRACK paper 99/5388
Dear Marie-Clare,
Attached is a revised manuscript which incorporates our response to the suggestions made by
the 4 reviewers. I will also fax the manuscript to you along with a copy of this letter.
My response to the 4 reviews follows:
Reviewer 1:
1. We used the modified Rankin scale. The reference has been added, number 11.
2. Since the only 2 events that occurred in the 45 patients with cancelled surgery occurred in
patients on high dose ASA, inclusion of these patients would have actually
strengthened the conclusions. This has been noted at the bottom of page 13.
3. The number of patients in each treatment group with cancelled surgery has been included in
a new table, table 1. This table also responds to the request for a CONSORT diagram.
Because the trial included 4 treatment groups and 2 analysis points (30 days and 3 months)
it was difficult to put the desired information into a diagram.
However it fits nicely into a table.
4. Patient compliance in each of the 4 treatment groups is now shown in table 1, and the
definition of compliance is included in a footnote. Also in the methods I have described the
fact that medication blister packs were collected at each follow-up for assessment of
patient compliance by pill counts.
5. Table 2 has been removed along with the text which described it. We will make this
information available on the web site for those interested in non-endpoint complications.
6. Table 3: Percentages have been added, and the text now includes the p-value and
confidence intervals on the relative risk of hemorrhagic stroke.
7. Table 4: We have not included an analysis restricted to the reduction in mild stroke and MI.
This was not a pre-specified analysis and focusing solely on minor events is not considered to
be good analytic strategy when dealing with composite outcome definitions.
8. Table 5: 95% confidence intervals have been added for the estimates of relative risk.
9. We are not aware of any other randomized trials of ASA in the perioperative period
of carotid endarterectomy, and disagree with the suggestion that these patients are similar
enough to the long term medically managed patients in the UK-TIA and Dutch-TIA trials
for a pooled meta-analysis.
Reviewer 2:
1. The introduction to the abstract has been modified to remove the incorrect inference and to
make it clear that ASA is know to be of benefit in the prevention of stroke following TIA
and previous stroke.
2. Unfortunately I was not able to locate any reference to the Rankin being validated on data
obtained by telephone. However, this is not exactly relevant as telephone follow-ups were
only performed on patients who were event free. This point has been included in the section
titled "Patient Follow-up" on page 10.
3. The "floating reference" has been fixed.
Reviewer 3:
1. The rationale for not including a placebo group has been included in the introduction at the
top of page 3.
2. same as reviewer 2 point 1.
3. The methods section has been modified to clarify the dosage regimen. This appears in the
first paragraph under "Study Treatment" on page 5.
4. same as reviewer 1 point 1.
5. The choice of a 2 day "wash out period" for the efficacy analysis was a compromise.
On the basis of the pharmacology of platelet inhibition with ASA a 7 day period would have
been ideal. However, this would have made the analysis impossible as we would not have
obtained an adequate sample size. As it is the 2 day rule reduced the sample size for the
efficacy analysis to a little less than half of the total number of patients in the study.
This point has been made at the end of the text which describes the need for the efficacy
analysis (top of page 9).
6. same as reviewer 1 point 5
7. This section of the text has been re-written and shortened to make the point more clearly.
The relative risks, and confidence intervals, can be found in the table (5).
8. The fact that heparin was used during surgery and reversed at the end of the procedure
has been clarified in the "Surgical Treatment" section on page 11. The possibility that heparin
use at the time of surgery may limit the generalizability of these findings beyond their clear
applicability to patients undergoing carotid endarterectomy has been made in the discussion
on page 16.
9. Since most of the literature on ASA in stroke prevention deals with the long term management
of stroke threatened patients and there is no randomized data related to ASA for stroke
prevention in carotid endarterectomy there is almost no other way in which we could begin
the discussion of the role of ASA.
10. same as reviewer 1 point 9.
The discussion has been enhanced to stress the dangers of over-generalizing these results
beyond the perioperative period of carotid endarterectomy.
11. and 12. these grammatical points are appreciated and have been addressed in the paper.
Reviewer 4:
1. The CONSORT diagram has been included as a table (see review 1 point 3).
2. same as reviewer 3 point 3.
3. Compliance by treatment group has been included in table 1. The need to in some cases to
use patient self report because blister packs were not returned was similar in all 4 treatment
groups. This has been noted under "Patient Compliance" on page 11.
4. same as reviewer 1 point 8.
5. The discussion related to the lack of a placebo group has been expanded to address the
implications for interpreting the results of this trial (page 17).
Thank you for the opportunity to respond to these suggestions. The reviewers are to be
commended. Their suggestions have led us to make important improvements in the paper.
I look forward to hearing from you.
Sincerely,
Wayne
D. Wayne Taylor
Professor
Clinical Epidemiology and Biostatistics
McMaster University
|