Skip to main content
. 2014 Oct 31;2014(10):CD003748. doi: 10.1002/14651858.CD003748.pub4

Dawson 2000.

Methods Study design: multicentre, randomised, double‐blind, placebo‐controlled
 Intention to treat: yes; LOCF method
Country: USA
Participants Number randomised: 698 (cilostazol n = 227; pentoxifylline n = 232; placebo n = 239)
Age (mean years ± SD): cilostazol = 66 ± 9; pentoxifylline = 66 ± 9; placebo = 66 ± 9
Sex M/F: cilostazol = 172/55; pentoxifylline = 181/51; placebo = 176/63
 Inclusion criteria: stable, moderate to severe symptoms of intermittent claudication for previous 6 months; confirmed PAD; baseline ICD ≥ 53.6 m (one minute); ACD ≤ 537.6 m (ten minutes)
 Exclusion criteria: patients with Buerger's disease; critical ischaemia; lower extremity surgical or endovascular reconstruction or sympathectomy in previous 3 months; limited exercise capacity due to conditions other than intermittent claudication; medical problems judged likely to preclude study completion; use of pentoxifylline or any investigational drug within 30 days of study enrolment; prior use of cilostazol
Interventions Treatment: cilostazol 100 mg, twice daily with a third placebo for blinding
 Treatment: pentoxifylline 400 mg, three times daily
 Contol: placebo
 Duration: 24 weeks
Outcomes ACD, ICD, resting doppler limb pressures, quality of life questionnaires (SF‐36, WIQ); measured at baseline and weeks 2, 4,8, 12, 16, 20 and 24
Notes Standardised treadmill test, beginning at 0% incline and 3.2 km/h, increasing incline 3.5% every three minutes while maintaining 3.2 km/h speed
Two‐ to three‐week baseline period
Risk of bias
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) Low risk "Randomization of eligible patients was stratified by clinical center, and patients were assigned to one of the three treatment regimens within each center using a permuted‐block design." "Patients were randomly assigned by using an interactive voice randomization system that blinded the investigator, patient and sponsor from treatment assignment"
Allocation concealment (selection bias) Low risk "Patients were randomly assigned by using an interactive voice randomization system that blinded the investigator, patient and sponsor from treatment assignment"
Blinding of participants and personnel (performance bias) 
 All outcomes Low risk "Patients were randomly assigned by using an interactive voice randomization system that blinded the investigator, patient and sponsor from treatment assignment". Study medications were identical in appearance and taken at similar intervals
Blinding of outcome assessment (detection bias) 
 All outcomes Unclear risk Blinding of assessors not adequately discussed
Incomplete outcome data (attrition bias) 
 All outcomes Low risk Missing patients were all accounted for and rates were similar between groups as to those who remained in the study
Selective reporting (reporting bias) Low risk Although no protocol was available all relevant outcomes appear to be reported on
Other bias Unclear risk Study supported by Otsuka America Pharmaceutical Inc