Evidence Table 3Quality assessment of randomized controlled trials (original and update 1 only)

Author, Year CountryRandomization adequate?Allocation concealment adequate?Groups similar at baseline?Eligibility criteria specified?Outcome assessors masked?Care provider masked?Patient masked?Intent-to-treat (ITT) analysis?Post-randomization exclusions?Reporting of attrition, crossovers, adherence, and contamination?Loss to follow-up: differential/high?Quality Rating
Atmaca, 2002
Yes, closed envelope system without patient stratificationYes-closed envelope system without patient stratificationC Group had higher frequency lesion in the RCA p= <0.02, and T Group had a higher EF <0.04Yes-undergoing elective single vessel PTCA. Inclusion criteria pts with Canadian Cardiac Society Class-II stable angina pectoris and de novo lesions in large native coronary arteries.Yes-but methods not describedYesYesNoSee #3 answer-baseline characteristics were shown after 10 patients were excludedYes/not applicable/yes/not reportedNoFair
Bertrand, 2000
YesYesYesYesYesYesYesYesYes-except for the one that withdrew consentYes/(1 withdrew consent before taking his first study med--not included in data) Not applicable/Not reported/Not reportedNoGood
Bhatt, 2006
Study drug assignment was performed centrally by an interactive voice-response system on the basis of a pre-established randomization scheme, stratified according to site.YesYesYesYesYesYesYesNoYes/not applicable/yes/yesNo. Follow-up with respect to the primary efficacy end points was complete in 99.5% of the C + ASA group and 99.6% of patients in the P + ASA group.Good
CAPRIE Steering Committee, 1996
Cure Investigators, 2001
YesYesYesYesYes-although unclear success of blindingYesYesYesNoYes/not applicable/yes/unsure--reasons for withdrawal not reportedNoGood
Di Pasquale, 2005
Randomization was performed at entry before starting any treatment and carried out using a preliminary computer algorithm, and the assignment of patients was decided at the time of admission by an independentYesYesYesECG and angiographic data were assessed and revised by 2 independent observers in order to reduce bias in the assessment of reperfusion and the result of PCIYesYesNot statedNoNot reported/Not applicable/Not reported/Not reportedNot reported-other than no one diedFair
Diener, 1996
13 countries
ESPRIT Study Group, 2006
14 countries-Europe/Australia
Telephone call, fax, or email to the central trial office.Yes- computer-generated randomization codes stratified by hospital before the start of the trial. The randomization codes and randomization program were generated by a clinical epidemiologist at the Academic Medical Center of the University of Amsterdam who was not otherwise involved in the trial.YesYesTreatment was not blinded. None of the investigators had any knowledge of event rates or complication rates according to treatment allocation.NoNoYes as well as on-treatmentYes--see #11 under Table A1Yes/Not applicable/Yes/YesNoFair
ESPS-2 Authors, 1997
13 countries
Yes-randomized to treatment groups according to a minimization technique which took into account the initial diagnosisYes-randomization was performed by a central computer, accessible to the centers day and night, and requiring the entry by the trialist of inclusion and exclusion criteria before allocating a randomization number to the pt.YesYesYesYesYesYesUnsureYes/Yes/Yes/NoYes-see commentsFair/good
Fiotti, 2003
No-method not reportedNo-sealed envelopeNoYesNoNoNoNoNoYes/Not applicable/Not reported/Not reportedNoFair/poor--not randomized, open-labeled, single centered,
Gorelick, 2003 USAYes −1:1 and the sequence was stratified by site to balance the treatment groups. Local study site personnel called a automated telephone registration system to register a study participantYesYesYesYes-except of 1 statistician who developed the randomization algorithmYesYesYesNoYes/Yes/No/Not reportedYes-15.2% in the Ticlopidine group and 13.3% ASA group lost to f/u or voluntary withdrawalFair/good
Hall, 1996
Italy and Japan
Yes-using a standard list of random numbersMethod not reported-did not indicate whether the standard list of random numbers were unreadable till allocationNo, incidence of total occlusions at baseline angiography was higher in the ASA group (15%) than in the T-ASA group 8%, p<.05. A higher percentage of pts had previous CABG or DM in T+ASA group (11%, 16% respectively) compared with ASA only group (3%, 6%) p= .02 and .01YesNot reportedNoNoYesNoYes/Yes/No/No=NoPoor
Hass, 1989
North America TASS
Randomized by a private independent, nonprofit organization--randomization within each center was stratified on the basis of 3 factors: history of ischemic CV disease, occurrence of a moderate or major stroke >3 months before entry, and the pt’s sex.Not reportedYesYesYesYesYesYesYesYes/Not applicable/Yes/Yes3% ticlopidine (n=46) and 2% assigned to the ASA group, (n=38)Good
Juergens, 2004
Yes-sealed envelope systemNo-sealed envelopeYesYes-not in detail (successful stent deployed)NoNoNoYesUnable to determine-- drug discontinuation occurred more often in the Ticlopidine group--including the composite of drug discontinuation, hemorrhage and vascular complicationsYes/Not reported/Not reported/NoNoPoor-not randomized, open-labeled, single centered, ? Allocation method, use of GP 2B/3An varied not only the agents but the frequency. LD of clopidogrel was 150mg instead of 300mg
Leon, 1998
Yes-used a prespecified randomization sequence to one of the 3 antithrombotic-drug regimens, according to clinical site and history of DMYesYesYesYes-treatment was not blinded, but all end points were adjudicated by a clinical events committee whose members were unaware of the pts’ treatment assignments.NoNoYesYes-3 components were primarily responsible for the differences seen in the incidence of primary event: revascularization of the target lesion (p=0.002), angiographically evident thrombosis (p=0.004), and recurrent MI (p=0.01), there was also significant difference in the incidence of revascularization of the target lesion and angiographically evident thrombosis between the group assigned to ASA and T and either the group assigned to ASA only or the group assigned to ASA and W.Not reported/Not applicable/Not reported/Not reportedNoFair
Mehta, 2001
International PCI-CURE
YesYesYes-although of note, before PCI, fewer pts on clopidogrel than on placebo had MI or refractory ischemia, p=0.008.Yes344/1313 PC pts in the clopidogrel group and 329/1345 PCI patients in the placebo group took open label thienopyridine before PCI. Following PCI, open label continued for 2–4 weeks and then the double-blind therapy was resumed.Yes, except during the open-label time after the PCI procedureYesYesNoYes/No/No/NoNoGood
Mueller, 2003
Germany and Switzerland
Yes-pre-specified randomization sequenceYesYesYes-“consecutive pts with successful stent implantation” were randomizedYes-treatment was not blinded, but all end points were adjudicated by a clinical events committee whose members were unaware of the patients’ treatment assignmentsNoNoYesUnable to determineYes/Not applicable/Not reported/Not reportedNo-Fair/poor-not blinded
Muller, 2000
No-unblindedYes-prespecified randomization sequenceYesYes-successful implantation (<50% residual stenosis without acute complications in the catheter lab resulting in death or emergency CABG)Yes-endpoints were adjudicated by a clinical-events committee whose members were unaware of the pts treatment assignmentsNoNot reportedYesNoYes/Not applicable/Not reported/NoNoFair-unblinded and not powered to show SS difference in cardiac events
Patti, 2006
Italy ARMYDA-2
Randomization blocks were created and distributed to the 2 centersNot reportedAge was significantly higher in the conventional loading dose vs. high loading dose p=0.027YesYesYesYesNoNoYes/Not applicable/Yes/YesNoGood
Piamsomboon, 2001
No- unblindedNot reportedMean lumen diameter in the ticlopidine groups was smaller than the clopidogrel group 2.75 ± 0.33 vs. 3.00 ±0.52, p= 0.01)YesNot reportedNot reportedNot reportedYesNoNot reported/No/Not reported/Not reportedNoPoor
Rupprecht, 1998
Not reportedNot reportedYesYesNoNoNoNoUnable to determineNot reported/Not applicable/Not reported/Not reportedNoPoor
Steinhuble, 2002
North America CREDO
YesYesLess use of statins and calcium channel blockers in the clopidogrel arm 53.5 vs. 57.3, p=.08; 25.5 vs. 29.4, p=.05 respectivelyYesYesYesYesYesNoYes/Not applicable/Yes/YesNoGood
Taniuchi, 2001
Method not reported other than it stated it used a randomized protocolMethod not reportedYes except the C group had more thrombus on angiography than the T group p= 0.009Yes-successful implantation (<20% residual stenosis, with TIMI2 or TIMI 3 flow) of an FDA-approved stent in a native coronary artery or in a CABG)NoNoNoYesCardiac death occurred more frequently in the T group (1.53% vs. 0.61%) resulting in a higher overall rate of major adverse cardiac events (4.60% vs. 3.85%) at 30 day but neither differences reached SS.Yes-1367 screened/1016 randomized; the primary end point, failure to complete 2 weeks of concurrent therapy with ASA was reached in 3.64% (19 pts) in the T group and in 1.62% (8 pts) in C group (p=0.043).NoFair

From: Evidence Tables

Cover of Drug Class Review: Newer Antiplatelet Agents
Drug Class Review: Newer Antiplatelet Agents: Final Update 2 Report [Internet].
Ketchum K, Peterson K, Thakurta S, et al.
Portland (OR): Oregon Health & Science University; 2011 Jun.
Copyright © 2011 by Oregon Health & Science University.

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.