Study	Reason for exclusion
Atarashi 1983	Patients with stroke within one month were excluded from the study
Domzal 1986	Confounded study: vinpocetine (27 participants) was compared with aminophylline (30 participants)
Dzyak 2002	Not a randomised trial
Gliem 1988	The method of allocation was not described No clinical outcomes were reported
Hadjiev 2004	Not a randomised trial
Kalvach 1988	No clinical outcomes reported, and probably a non‐randomised trial
Levic 2001	The study was probably not randomised, it is not clear if treatment was started within 2 weeks of stroke onset, and the study evaluated the late neuropsychiatric consequences of stroke Death and disability are not reported
Lipani 1984	In addition to stroke patients, patients with other diagnostic categories (TIA, dementia, Huntington's chorea) also entered the study The number of participants with stroke was small (6 participants), and some, if not all, of these were in the chronic phase of stroke
Manchev 2003	253 patients treated with 10 to 20 mg vinpocetine for 5 to 30 days iv or 3 to 30 days per os and 40 controls Not a randomised study
Manconi 1986	Patients with diagnostic categories other than stroke also entered the study The term 'previously established cerebral thrombosis' most probably denoted patients a long time after stroke and not patients with acute stroke At the end of the report ‘chronic cerebral dysfunction’ is mentioned referring to the whole study population and data are not presented separately for those with previous stroke
Reneles 1986	Mixed patient group (TIA, acute and chronic stroke) The mean time from stroke onset to randomisation was 5.6 months Of the 16 participants randomised within one month after stroke no data were given for those randomised less than 14 days after stroke onset
Suslina 1999	Non‐randomised study Lipid peroxidation was measured No clinical outcome
Wang 2006	Study performed in acute hemorrhagic stroke Patients with ischaemic stroke were excluded
Wasilewski 1985	There were 50 participants in both the placebo and cinnarizine groups whereas in the vinpocetine group there were only 32 This imbalance in sample size plus the term 'the patient groups were selected at random' suggests that randomisation, if performed at all, was improper We attempted to clarify this with the principal investigator, but no source data could be obtained
Yi 2004	83 patients were screened, 70 were randomised: 35 to vinpocetine and 35 to venoruton Therefore the study is confounded
Zou 1990	Based on the MEDLINE abstract the study was confounded because vinpocetine was compared with acupuncture

iv: intravenous 
 per os: by mouth 
 TIA: transient ischaemic attack