Types of studies

We will include randomized controlled trials (RCT) in which the trial authors compare Neuroaid with placebo or no treatment in people with ischemic stroke.

Types of participants

We will include people with a definite diagnosis of ischemic stroke of any age or sex, regardless of stroke severity on admission, using the World Health Organization (WHO) criteria of stroke, confirmed by computerized tomography (CT) or magnetic resonance imaging (MRI) scan (WHO Task Force 1989; Sacco 2013).

Types of interventions

We will include trials evaluating the efficacy and safety of Neuroaid in people with ischemic stroke, regardless of dosage of treatment, or the length of treatment. Neuroaid is referred to as MLC601 or MLC901. The control interventions will be placebo or no treatment.

We will include studies that compare use of:

• Neuroaid versus placebo or no treatment;

• Neuroaid plus baseline treatment versus baseline treatment alone.

Types of outcome measures

Electronic searches

We will search the Cochrane Stroke Group Trials Register and the following electronic databases:

• Cochrane Central Register of Controlled Trials (CENTRAL; latest issue) in the Cochrane Library (date searched);

• MEDLINE Ovid (1948 to present);

• Embase Ovid (1980 to present);

• Science Citation Index Web of Science (1900 to present);

• CINAHL EBSCO (Cumulative Index to Nursing and Allied Health Literature; 1982 to present)

• AMED Ovid (the Allied and Complementary Medicine Database; 1985 to present);

• China Biological Medicine Database (CBM; 1978 to present);

• China National Knowledge Infrastructure (CNKI; 1979 to present);

• Chinese Science and Technique Journals Database (VIP; 1989 to present);

• Wanfang Data (1984 to present).

We developed the MEDLINE search strategy with the help of the Cochrane Stroke Group Information Specialist and will adapt it for searching the other databases (Appendix 1).

Searching other resources

We will also search the following ongoing trials and research registers:

• US National Institutes of Health Ongoing Trials Register (ClinicalTrials.gov);

• World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP);

• Stroke Trials Registry;

• ISRCTN Registry.

In an effort to identify further published, unpublished, and ongoing studies, we will search all reference lists of retrieved articles. We will use Science Citation Index Cited Reference Search for forward tracking of relevant articles. When necessary, we will attempt to contact the relevant manufacturers and study authors to obtain further data.

Selection of studies

Two authors (CZ, JZ) will independently screen the titles and the abstracts of the references identified from the searches of the electronic databases, and will exclude obviously irrelevant reports. We will retrieve the full‐text articles of the remaining records, and the same two authors will independently select trials according to our predetermined inclusion criteria. We will resolve any disagreements through discussion, or when necessary, we will consult a third review author (ML). We will collate multiple reports of the same study so that each study, not each reference, is the unit of interest in the review. We will record the selection process and complete a PRISMA flow diagram (Liberati 2009).

Data extraction and management

Two authors (CZ, JZ) will independently extract data from the included studies using a standardized data extraction form. We will record the following: characteristics of participants, interventions, treatment duration, and all assessed outcomes after the treatment and at the end of the follow‐up period. If the two authors cannot reach consensus, the third author (ML) will make a final decision. For dichotomous outcomes, we will extract the number of participants experiencing the event and the total number of participants in each arm of the trial. For continuous outcomes, we will extract the mean value and standard deviation for the changes in each arm of the trial, along with the total number in each group. Wherever possible, we will use outcomes from the intention‐to‐treat (ITT) population, and if not possible, we will extract per protocol outcomes.

Assessment of risk of bias in included studies

Two authors (CZ, JZ) will independently assess the risk of bias in included studies using Cochrane's tool for assessing risk of bias, as described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). The two authors will resolve any disagreements through discussion, with a third author (ML) if necessary. Assessment of risk of bias will address the following seven domains.

• Random sequence generation

• Allocation concealment

• Blinding of participants and personnel

• Blinding of outcome assessors

• Incomplete outcome data

• Selective outcome reporting

• Other possible bias

We will categorize the risk of bias for each domain above as 'high', 'low', or 'unclear', and provide information from the study report, together with a justification for our judgment in the 'Risk of bias' tables.

Measures of treatment effect

For dichotomous outcomes, we will express results as risk ratios (RRs) with 95% confidence intervals (CIs). For continuous outcomes, we will present results as mean difference (MD) with 95% CIs (if the same scale for each trial is available), or standardized mean difference (SMD) with 95% CIs (if different scales were used). Also for continuous outcomes, we plan to compare the change between the baseline and end of treatment, and between the baseline and end of follow‐up.

Unit of analysis issues

We will assess the level at which randomization occurred for all studies. For studies with non‐standard designs (e.g. cross‐over trials, cluster‐randomized trials), we will manage the data according to theCochrane Handbook for Systematic Reviews of Interventions (Higgins 2011).

Dealing with missing data

We will contact the trial authors for missing data. If some data remain unavailable, we will consider both best‐case and worst‐case scenarios.

Assessment of heterogeneity

We will use the I² statistic to measure heterogeneity among the studies in each analysis. We will consider I² greater than 50% to be moderate to substantial heterogeneity, and we will seek the potential sources of the heterogeneity (clinical and methodological heterogeneity). We plan to perform meta‐analysis using the random‐effects model, regardless of the level of heterogeneity.

Assessment of reporting biases

We will use funnel plots to check for reporting biases (Egger 1997).

Data synthesis

We will perform a meta‐analysis by pooling the appropriate data using RevMan 5 (RevMan 2014). If there are no suitable studies, we will present results in a narrative 'Summary of findings' table format.

Subgroup analysis and investigation of heterogeneity

If the appropriate data are available, we intend to undertake subgroup analysis according to:

• different kinds of Neuroaid: MLC 601 versus MLC 901;

• different stroke severity: mild stroke versus moderate to severe stroke;

• different times to the start of treatment: within one month versus after one month from stroke onset;

• different ischemic stroke subtype classification: TOAST subtype.

Sensitivity analysis

We will perform sensitivity analyses with data by excluding studies:

• with inadequate concealment of allocation;

• in which outcome evaluation was not blinded;

• in which loss to follow‐up was not reported, or was more than 10%.