Cholinesterase inhibitors for vascular dementia and other vascular cognitive impairments: a network meta‐analysis

Ceri E Battle; Azmil H Abdul-Rahim; Susan D Shenkin; Jonathan Hewitt; Terry J Quinn

doi:10.1002/14651858.CD013306.pub2

. 2021 Feb 22;2021(2):CD013306. doi: 10.1002/14651858.CD013306.pub2

Cholinesterase inhibitors for vascular dementia and other vascular cognitive impairments: a network meta‐analysis

Ceri E Battle ^1,^✉, Azmil H Abdul-Rahim ², Susan D Shenkin ³, Jonathan Hewitt ⁴, Terry J Quinn ⁵

Editor: Cochrane Dementia and Cognitive Improvement Group

PMCID: PMC8407366 PMID: 33704781

Abstract

Background

Vascular cognitive impairment (VCI) describes a broad spectrum of cognitive impairments caused by cerebrovascular disease, ranging from mild cognitive impairment to dementia. There are currently no pharmacological treatments recommended for improving either cognition or function in people with VCI. Three cholinesterase inhibitors (donepezil, galantamine, and rivastigmine) are licenced for the treatment of dementia due to Alzheimer's disease. They are thought to work by compensating for reduced cholinergic neurotransmission, which is also a feature of VCI. Through pairwise comparisons with placebo and a network meta‐analysis, we sought to determine whether these medications are effective in VCI and whether there are differences between them with regard to efficacy or adverse events.

Objectives

(1) To assess the efficacy and safety of cholinesterase inhibitors in the treatment of adults with vascular dementia and other VCI.

(2) To compare the effects of different cholinesterase inhibitors on cognition and adverse events, using network meta‐analysis.

Search methods

We searched ALOIS, the Cochrane Dementia and Cognitive Improvement Group's register, MEDLINE (OvidSP), Embase (OvidSP), PsycINFO (OvidSP), CINAHL (EBSCOhost), Web of Science Core Collection (ISI Web of Science), LILACS (BIREME), ClinicalTrials.gov, and the World Health Organization International Clinical Trials Registry Platform on 19 August 2020.

Selection criteria

We included randomised controlled trials in which donepezil, galantamine, or rivastigmine was compared with placebo or in which the drugs were compared with each other in adults with vascular dementia or other VCI (excluding cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL)). We included all drug doses and routes of administration.

Data collection and analysis

Two review authors independently identified eligible trials, extracted data, assessed risk of bias, and applied the GRADE approach to assess the certainty of the evidence. The primary outcomes were cognition, clinical global impression, function (performance of activities of daily living), and adverse events. Secondary outcomes were serious adverse events, incidence of development of new dementia, behavioural disturbance, carer burden, institutionalisation, quality of life and death. For the pairwise analyses, we pooled outcome data at similar time points using random‐effects methods. We also performed a network meta‐analysis using Bayesian methods.

Main results

We included eight trials (4373 participants) in the review. Three trials studied donepezil 5 mg or 10 mg daily (n= 2193); three trials studied rivastigmine at a maximum daily dose of 3 to 12 mg (n= 800); and two trials studied galantamine at a maximum daily dose of 16 to 24 mg (n= 1380). The trials included participants with possible or probable vascular dementia or cognitive impairment following stroke. Mean ages were between 72.2 and 73.9 years. All of the trials were at low or unclear risk of bias in all domains, and the evidence ranged from very low to high level of certainty.

For cognition, the results showed that donepezil 5 mg improves cognition slightly, although the size of the effect is unlikely to be clinically important (mean difference (MD) −0.92 Alzheimer's Disease Assessment Scale‐Cognitive Subscale (ADAS‐Cog) points (range 0 to 70), 95% confidence interval (CI) −1.44 to −0.40; high‐certainty evidence). Donepezil 10 mg (MD −2.21 ADAS‐Cog points, 95% CI −3.07 to −1.35; moderate‐certainty evidence) and galantamine 16 to 24 mg (MD −2.01 ADAS‐Cog point, 95%CI −3.18 to −0.85; moderate‐certainty evidence) probably also improve cognition, although the larger effect estimates still may not be clinically important. With low certainty, there may be little to no effect of rivastigmine 3 to 12 mg daily on cognition (MD 0.03 ADAS‐Cog points, 95% CI −3.04 to 3.10; low‐certainty evidence).

Adverse events reported in the studies included nausea and/or vomiting, diarrhoea, dizziness, headache, and hypertension. The results showed that there was probably little to no difference between donepezil 5 mg and placebo in the number of adverse events (odds ratio (OR) 1.22, 95% CI 0.94 to 1.58; moderate‐certainty evidence), but there were slightly more adverse events with donepezil 10 mg than with placebo (OR 1.95, 95% CI 1.20 to 3.15; high‐certainty evidence). The effect of rivastigmine 3 to 12 mg on adverse events was very uncertain (OR 3.21, 95% CI 0.36 to 28.88; very low‐certainty evidence). Galantamine 16 to 24 mg is probably associated with a slight excess of adverse events over placebo (OR 1.57, 95% CI 1.02 to 2.43; moderate‐certainty evidence).

In the network meta‐analysis (NMA), we included cognition to represent benefit, and adverse events to represent harm. All drugs ranked above placebo for cognition and below placebo for adverse events. We found donepezil 10 mg to rank first in terms of benefit, but third in terms of harms, when considering the network estimates and quality of evidence. Galantamine was ranked second in terms of both benefit and harm. Rivastigmine had the lowest ranking of the cholinesterase inhibitors in both benefit and harm NMA estimates, but this may reflect possibly inadequate doses received by some trial participants and small trial sample sizes.

Authors' conclusions

We found moderate‐ to high‐certainty evidence that donepezil 5 mg, donepezil 10 mg, and galantamine have a slight beneficial effect on cognition in people with VCI, although the size of the change is unlikely to be clinically important. Donepezil 10 mg and galantamine 16 to 24 mg are probably associated with more adverse events than placebo. The evidence for rivastigmine was less certain.

The data suggest that donepezil 10 mg has the greatest effect on cognition, but at the cost of adverse effects. The effect is modest, but in the absence of any other treatments, people living with VCI may still wish to consider the use of these agents. Further research into rivastigmine is needed, including the use of transdermal patches.

Plain language summary

Medicines to treat people with vascular dementia and other vascular cognitive impairments

Review question

What is the evidence for cholinesterase inhibitors (medicines designed to improve memory and thinking in people with dementia), when used with people who have vascular dementia?  Background

Vascular dementia (or vascular cognitive impairment) is a term used when a person has problems with memory and thinking that are caused by a disruption of blood supply. There are few drug treatments for vascular dementia.

In this review, we evaluated three drugs from the cholinesterase inhibitor family, donepezil, rivastigmine, and galantamine. These medications are widely used in Alzheimer's dementia but may also be useful in people with vascular dementia. Previous reviews of these cholinesterase inhibitor drugs could not draw definitive conclusions for people with vascular dementia.  Purpose of this review

We wanted to learn whether cholinesterase inhibitors benefit people with vascular dementia. We were interested in their effects on memory, thinking, and daily functioning. We wanted to learn of any harms associated with these drugs.

As some time has passed since the previous reviews, we wanted to update them by searching for new studies. We combined the three previous reviews on donepezil, rivastigmine and galantamine into one review.  What we did

We searched for studies that described the effects of donepezil, rivastigmine, and galantamine for people with vascular dementia. We searched databases of scientific studies and contacted drug manufacturers and experts in vascular dementia. Our search is current to 19 August 2020.

To be included in our review, studies had to randomly assign people with vascular dementia to treatment with a cholinesterase inhibitor, or a dummy pill (placebo) and then compare the two groups. Studies comparing one cholinesterase inhibitor against another were also included. We combined the results of the included studies for each medicine to estimate how effective they were and how likely they were to cause side effects. We assessed how well the studies were conducted and how credible the results were.

We did not find studies which compared different cholinesterase inhibitors with each other. To see whether the different cholinesterase inhibitor drugs differed in their effects, we used a technique called network meta‐analysis, which can provide an idea of how the medicines might perform if they were compared head‐to‐head.  What we found

We found 8 studies including a total of 4373 people with vascular dementia (or vascular cognitive impairment). The studies tested the drug donepezil at two different doses (5mg and 10mg daily), against each other and against placebo. Rivastigmine and galantamine were tested against placebo only. Rivastigmine is available as a skin patch, but the studies only tested the pill version. All eight studies evaluated participants when they first started taking the medicine or placebo and again six months later. Different tests were used to measure the effects. All studies included tests of memory, thinking and reported side effects.

People taking donepezil or galantamine had better scores on memory and thinking tests than people taking placebo, but the benefits were modest and may not be large enough to be evident in daily life. There was no evidence of a difference for rivastigmine, but the evidence was less certain, and the doses taken by some participants may have been too low to show an effect. We found evidence that when compared to placebo, side effects such as nausea and diarrhoea, were more common in people taking donepezil 10mg and galantamine, but probably not donepezil 5mg. We were unable to draw conclusions about side effects of rivastigmine from the studies.

No vascular dementia trials comparing the different cholinesterase against each other have been conducted. Using the information from the individual studies, we made indirect assessments of how the drugs would perform if tested head‐to‐head. The results suggested that donepezil 10 mg had the greatest effect on memory and thinking, but caused more side effects than donepezil 5 mg or galantamine.

There were only a small number of studies for each drug. Certainty in the results varied between drugs and between outcomes, from high to very low certainty. The studies showed only a small benefit at most; however, in the absence of any other treatments, people living with dementia may still wish to consider use of these drugs.

Summary of findings

Background

Description of the condition

Vascular cognitive impairment (VCI) describes a broad spectrum of cognitive impairments caused by cerebrovascular disease, ranging from mild cognitive impairment to dementia (Dichgans 2017; van der Flier 2018). In 2017, the Vascular Impairment of Cognition Classification Consensus Study (VICCCS) led to a revised conceptualisation of VCI, in which VCI is divided into mild and major subtypes according to the level of impairment (Skrobot 2017). Mild VCI is not subdivided, but major VCI (or vascular dementia) has four subdivisions: post‐stroke dementia, subcortical ischaemic vascular dementia, multi‐infarct (cortical) dementia, and mixed dementias. For the purposes of this review, we treated VCI as an umbrella term that incorporates vascular dementia and other cognitive syndromes with a presumed vascular basis (i.e. all categories listed in the VICCCS definition, including mild VCI and all subdivisions of major VCI). Two criteria must be met for a diagnosis of VCI: firstly, a cognitive deficit demonstrated through neuropsychological testing, and secondly, the presence of cerebrovascular disease. VCI is further classified as ‘probable’ or ‘possible’, according to the level of evidence that there is a causal relationship between the cognitive impairment and the vascular disease (Dichgans 2017).

The clinical presentation of VCI depends on the type, extent, and location of the underlying cerebrovascular pathology. Possible symptoms of VCI are numerous and include memory problems, mental slowness, and problems with executive function (such as planning, sequencing, and problem solving). Patients often report difficulties with higher‐order cognitive functions, such as planning, organising, and monitoring behaviour. Behavioural symptoms and psychological symptoms, including emotional lability, anxiety, depression, and apathy, are also commonly reported. Other neurological signs and symptoms often occur, including reflex asymmetry, dysarthria (difficulty with speech), gait disorders, and problems with balance, parkinsonism, rigidity, or urinary incontinence (O'Brien 2003; van der Flier 2018). VCI due to a single stroke presents abruptly, whilst symptoms and signs due to subcortical damage, such as lacunae and white matter disease (a progressive age‐related decline in nerves that connect areas of brain to each other), typically develop more insidiously (Erkinjuntti 2004).

As life expectancy increases, VCI has become a growing public health issue. Approximately 36 million people have dementia worldwide, and this number is expected to reach 66 million by 2030 and 115 million by 2050 (Wortmann 2012). Vascular dementia is the second most common form of dementia after Alzheimer’s disease and accounts for at least 20% of dementia cases (Wu 2016). The prevalence of VCI is strongly age‐related. In participants aged 65 to 84 years, the prevalence of mild forms of VCI that do not reach the criteria required for a diagnosis of dementia is higher than that of vascular dementia. Rates of conversion to dementia, institutionalisation, and mortality are significantly increased in these patients, suggesting that people with mild VCI are an important target population for the prevention of poor outcomes (Dichgans 2017).

Description of the intervention

Accurate assessment and management of vascular risk factors are a key priority in the treatment of VCI, particularly early in the disease when preventive strategies may prove to be most effective (Ritter 2015). Although primary prevention trials have suggested that treatment of hypertension, adherence to a Mediterranean diet, physical activity, and smoking cessation may reduce the risk of cognitive decline, there is limited evidence regarding these interventions for improving cognition in established VCI (Ritter 2015). There are currently no specific pharmacological treatments recommended for improving either cognition or function in VCI. Management strategies used for patients with VCI are similar to those for other forms of dementia. Key principles include treating psychological and behavioural comorbidities, providing information and support to the patient and caregivers, and maximising the patient's independence (Dichgans 2017).

Cholinesterase inhibitors are medicines recommended as options for managing mild‐to‐moderate dementia due to Alzheimer's disease in several clinical guidelines (e.g. Hort 2010; NICE 2018). Alzheimer's disease is the most common cause of dementia and is found in approximately 70% of autopsies of people with dementia (Qiu 2009). The three cholinesterase inhibitors currently marketed for the treatment of Alzheimer's disease are donepezil, rivastigmine, and galantamine. Cholinesterase inhibitors are taken orally once or twice a day; rivastigmine can also be applied transdermally. Tacrine is no longer licenced. Although memantine is often considered alongside cholinesterase inhibitors, it is a different drug class.

Previous Cochrane Reviews have reported modest cognitive benefit from cholinesterase inhibitors in mild‐to‐moderate dementia due to VCI (Malouf 2004; Craig 2006; Birks 2013), but a number of harms related to use of cholinesterase inhibitor treatment have also been reported, with evidence of more adverse events overall in people treated with a cholinesterase inhibitor than with placebo. Nausea, vomiting, and diarrhoea in particular were reported significantly more frequently in the cholinesterase inhibitor groups than in the placebo groups (Birks 2006). Serious adverse events have also been reported, including stroke, pneumonia, and myocardial infarction (Birks 2006).

How the intervention might work

Cholinesterase inhibitors inhibit the activity of the enzyme acetylcholinesterase, and increase acetylcholine levels by decreasing the rate at which the substance is broken down. The aim of prescribing cholinesterase inhibitors for Alzheimer's disease is to compensate for the loss of cholinergic brain cells and to boost cholinergic neurotransmission in forebrain regions (Colović 2013). Reductions in acetylcholine and acetyltransferase activity (markers of cholinergic neurotransmission) are common to both Alzheimer's disease and VCI, raising the possibility that these drugs may be beneficial for both conditions (Toghi 1996; Perry 1997). Donepezil is a second‐generation cholinesterase inhibitor that is a non‐competitive, reversible antagonist of cholinesterase and is highly selective for acetylcholinesterase compared to butyryl‐cholinesterase (Dawbarn 2001). Rivastigmine is a 'pseudo‐irreversible' inhibitor of acetylcholinesterase and also of butyryl‐cholinesterase, which is a non‐specific cholinesterase enzyme. Galantamine is a reversible, competitive inhibitor of acetylcholinesterase with minimal butyryl‐cholinesterase inhibitory activity (Lilienfeld 2002).

Why it is important to do this review

To date, the US Food and Drug Administration and the European Medicines Agency have not approved any pharmacological treatments for VCI or vascular dementia symptoms. Three previous Cochrane Reviews have investigated the efficacy and safety of individual cholinesterase inhibitors for VCI. The review of donepezil for VCI reported some improvements in cognitive function, activities of daily living, and global measures of change (Malouf 2004). The review investigating galantamine in VCI concluded that there were some advantages over placebo in the areas of cognition and global clinical state (Craig 2006). Similarly, rivastigmine had some benefit on cognitive function at 24 weeks in people with VCI (Birks 2013). However, these reviews did not investigate potential differences in efficacy between these medications, therefore a review that covered the evidence for all three cholinesterase inhibitors would be helpful to clinicians.

A number of years have passed since the publication of the original reviews. This new, overarching review ensured that any new trials were included. It also allowed the use of contemporary approaches to evidence synthesis (e.g. use of GRADE methods to assess evidence quality) that were not in use when the previous reviews were written. For the first time, we have included all cholinesterase inhibitors in a network meta‐analysis (NMA) in order to address the question of which cholinesterase inhibitor, if any, is the most efficacious and safest in the management of VCI (Salanti 2012).

Objectives

To assess the efficacy and safety of cholinesterase inhibitors in the treatment of adults with vascular dementia and other VCI.
To compare the effects of different cholinesterase inhibitors on cognition and adverse events in adults with vascular dementia and other VCI, using network meta‐analysis.

Methods

Criteria for considering studies for this review

Types of studies

We included parallel‐group, randomised controlled trials (RCTs) in which participants with VCI are assigned to treatment with a cholinesterase inhibitor or placebo, or to alternative cholinesterase inhibitors. We included any identified trial regardless of publication status. Restricting eligibility to RCTs meant that there were no differences in study design between interventions, satisfying the transitivity assumption for the NMA.

Types of participants

We included participants diagnosed as having vascular dementia on the basis of any validated and internationally recognised diagnostic framework for dementia, including the Diagnostic and Statistical Manual of Mental Disorders (DSM), APA 2013, and the ICD‐11 Classification of Mental and Behavioural Disorders: Clinical Description and Diagnostic Guidelines (WHO 1992), and any classification systems specific to VCI, such as the National Institute of Neurological Disorders and Stroke and the Association International pour la Recherche et l' Enseignement en Neurosciences (NINDS/AIREN) (Roman 1993). Diagnosis of VCI with no dementia (sometimes labelled VCIND) was based on cognitive test scores and a clinical diagnosis to ensure the distinction between vascular and non‐vascular impairment. We considered that all participants identified by these criteria would be equally likely to be treated with any of the cholinesterase inhibitors, thus satisfying the transitivity assumption for the NMA.

We made an a priori decision that we would not include monogenic conditions such as cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) in this review. This early‐onset familial form of dementia may not be a good model of the common forms of VCI seen in older age. CADASIL is included in another Cochrane Review covering cholinesterase inhibitors for rarer dementias (Li 2015).

If studies were conducted in a population with a mixture of dementia subtypes, we included the data if the proportion of participants with VCI was 80% or more. We excluded studies in which the population was described as having undifferentiated dementia, or where dementia subtype was not described, because based on general population frequencies it would be unlikely that more than 80% would have VCI.

Types of interventions

We included any cholinesterase inhibitor licenced for the treatment of Alzheimer's disease or another form of dementia, that is donepezil, galantamine, and rivastigmine.

These medications can be administered orally or, in the case of rivastigmine, transdermally. All routes of administration were eligible for inclusion.

The licenced cholinesterase inhibitors are available in a range of doses. The drugs usually have a dose‐titration period. We considered the final dose achieved in our analyses. Reviews in non‐vascular dementias suggest that doses may differ in efficacy and adverse events. All doses were eligible for inclusion, but we intended to consider studies using a final dose within the manufacturer's recommended range separately from other studies. For donepezil, we planned to include studies where the final dose was a licenced oral dose of 5 mg, 10 mg, or 23 mg daily; we planned to consider each of these doses separately. For rivastigmine, we planned to assess the manufacturer's recommended final dose of 6 to 12 mg daily for the oral preparation, or 4.6 mg/24 hours or 9.5 mg/24 hours for the transdermal preparation; any other doses that were studied would be considered separately. For galantamine, we planned to assess the manufacturer's recommended oral dose of 16 to 24 mg (standard or modified‐release); any other doses that were studied would be considered separately.

Eligible comparator interventions were placebo, or, for the network meta‐analysis only, an alternative cholinesterase inhibitor.

Types of outcome measures

Primary outcomes

Cognitive function (e.g. the cognitive part of the Alzheimer's Disease Assessment Scale, ADAS‐Cog; Syndrom‐Kurz test)
Clinical global impression (e.g. Clinician's Interview‐Based Impression of Change scale, CIBIC‐Plus; Clinical Global Impression of Change, CGIC; Clinical Global Impression, CGI (which measures of symptom severity, treatment response, and treatment efficacy); Sandoz Clinical Assessment Geriatric Scale, SCAG)
Functional performance in activities of daily living (e.g. Alzheimer's Disease Cooperative Study‐Activities of Daily Living, ADCS‐ADL; Behavioural Rating Scale for Geriatric Patients, BGP)
Number of adverse events (including nausea and/or vomiting, diarrhoea, dizziness, loss of appetite and/or anorexia, headache and hypertension). If the number of adverse events was not presented, we used the number of participants with any adverse events (one or more) in a study. We accepted adverse events as defined in the included studies.

Secondary outcomes

Serious adverse events (SAEs) (including death, transient ischaemic attack or stroke, focal motor seizures, pneumonia, and myocardial infarction or congestive heart failure)
Incidence of development of new dementia: if any studies were concerned exclusively with vascular mild cognitive impairments or related syndromes, then we described rates of incident dementia as an outcome. This outcome was considered separately to the other outcomes of interest to the dementia population.
Behavioural disturbance (e.g. Neuropsychiatric Inventory, NPI)
Carer burden (e.g. Carer ‐ Dementia Quality of Life Instrument)
Institutionalisation
Quality of life (e.g. Dementia Quality of life Instrument, Alzheimer's Disease‐Related Quality of Life Measure)
Death

NMA outcomes

We used cognition to represent benefit outcomes and adverse events to represent harms in the NMA. We selected these outcomes as they were the most consistently reported in the included studies.

Search methods for identification of studies

Electronic searches

We searched ALOIS (www.medicine.ox.ac.uk/alois) ‐ the Cochrane Dementia and Cognitive Improvement Group’s (CDCIG) specialised register on 19 August 2020.

ALOIS is maintained by the Information Specialists for the Cochrane Dementia and Cognitive Improvement Group, and contains studies that fall within the areas of dementia prevention, dementia treatment and management, and cognitive enhancement in healthy elderly populations. The studies are identified through searching:

a number of major healthcare databases: MEDLINE, Embase, CINAHL (Cumulative Index to Nursing and Allied Health Literature), and PsycINFO;
a number of trial registers: US National Institutes of Health Ongoing Trials Register ClinicalTrials.gov and the World Health Organization International Clinical Trials Register Platform (WHO ICTRP), which covers ISRCTN; the Chinese Clinical Trials Register; the German Clinical Trials Register; the Iranian Registry of Clinical Trials; and the Netherlands National Trials Register, plus others;
the Cochrane Central Register of Controlled Trials (CENTRAL) (the Cochrane Library);
grey literature sources: ISI Web of Science Core Collection.

To view a list of all sources searched for ALOIS, visit the ALOIS website (www.medicine.ox.ac.uk/alois).

Details of the search strategies run in healthcare bibliographic databases, used for the retrieval of reports of dementia, cognitive improvement, and cognitive enhancement trials, can be viewed on the Cochrane Dementia and Cognitive Improvement Group’s website (dementia.cochrane.org/searches).

We ran additional searches in MEDLINE, Embase, CINAHL, PsycINFO, LILACS (Latin American and Caribbean Health Science Information database), ClinicalTrials.gov, and the WHO ICTRP from inception to ensure that the searches for this review were as comprehensive and as up‐to‐date as possible. The search strategies used are shown in Appendix 1.

Searching other resources

We checked the reference lists of eligible studies and previous systematic reviews to identify additional studies. We contacted pharmaceutical companies (Eisai and Pfizer for donepezil (Aricept); Shire for galantamine (Reminyl); Lundbeck for rivastigmine (Exelon)) and searched their press releases pertaining to cholinesterase inhibitors. We requested all conference posters presented by relevant authors and those sponsored by the pharmaceutical companies. We sought other grey literature through handsearching of reference lists of retrieved relevant trials and systematic reviews. We also handsearched relevant conference abstracts that are not covered in ALOIS, specifically; International Stroke Conference 2017 to 2019 (published in Stroke); European Stroke Organisation Conference 2017 to 2019 (published in European Stroke Journal); and Alzheimer’s Association International Conference 2017 to 2019 (published in Alzheimer’s & Dementia).

Data collection and analysis

Selection of studies

We used Cochrane’s Screen4Me workflow to help in the assessment of the search results. Screen4Me comprises three components: known assessments – a service that matches records in the search results to records that have already been screened by Cochrane Crowd (Cochrane’s citizen science platform) and have been labelled as an RCT or not an RCT; the RCT classifier – a machine learning model that distinguishes RCTs from non‐RCTs; and Cochrane Crowd ‐ Cochrane’s crowd‐sourcing platform where contributors help to identify RCTs and other study types.

For more information about Screen4Me and the evaluations that have been done, visit the Screen4Me web page on the Cochrane Information Specialists Portal (community.cochrane.org/organizational-info/resources/resources-groups/information-specialists-portal). More detailed information regarding evaluations of the Screen4Me components can also be found in the following publications: McDonald 2017; Thomas 2017; Marshall 2018; Noel‐Storr 2018.

After the results had been through the Screen4Me workflow, two review authors (CEB and AHAR) independently assessed the remainder. Any disagreements were resolved through discussion and consultation with a third review author (TJQ). We created a PRISMA flow diagram to map out the number of records identified, included, and excluded (Figure 1). We listed all studies excluded after full‐text assessment and their reasons for exclusion in a 'Characteristics of excluded studies' table.

Data extraction and management

Two review authors independently extracted data using a standardised data extraction form.

We extracted results for the primary outcome measures at the following time points, where reported: up to 3 months, from 3 months to 6 months, from 6 months to 18 months, and more than 18 months. We extracted data from more than one time point, if such information was available.

We extracted the following data that may act as effect modifiers from each included study:

population: diagnostic criteria; baseline mean age; male‐to‐female ratio; comorbidities; concurrent medications; ethnicity and socioeconomic status;
interventions: duration of the intervention, including duration of any wash‐out, run‐in, or titration period; dosage regimen, including during any titration period; route of administration;
outcome measures: measure used, time point completed;
'Risk of bias' domains (see Assessment of risk of bias in included studies);
funding sources.

Assessment of risk of bias in included studies

Two review authors (CEB and AHAR) independently assessed the risk of bias in each study. Any disagreements were resolved by discussion to reach consensus, involving a third review author (TJQ) when necessary. Using the Cochrane 'Risk of bias' tool (Higgins 2017), we assessed the risk of bias of each included study based on the following domains: random sequence generation, allocation concealment, blinding of participants, blinding of outcome assessment, incomplete outcome data, and selective outcome reporting. We judged the level of risk of bias within each study explicitly for each domain as being at 'low', 'high', or 'unclear' risk of bias. We described all judgements fully and presented our conclusions in the 'Risk of bias’ tables.

We judged studies as being at low risk of bias for the incomplete outcome data domain when numbers of and causes for dropouts were balanced between arms. For continuous outcomes, we considered the following factors: the level of missing data, the difference between groups, and the reasons for missingness. We also took into account whether the approach to missing data (e.g. observed case (OC) or last observation carried forward (LOCF)) gave different effect estimates. For dichotomous outcomes, we compared the proportions missing in each group with each other and with the adverse event risk, by visual inspection. If there was a substantial difference in missing data between groups, or the proportion of missing data was comparable with the adverse events risk, we rated the risk of attrition bias as high. We assessed selective outcome reporting by comparing the outcomes the trialists intended to analyse against the published study results. Where no trial protocol was available, we assigned a judgement of high risk of bias when study results did not include the primary outcome measures set out in their methods section.

Measures of treatment effect

For binary outcomes, we used odds ratios (ORs) with 95% confidence intervals (CIs) as the measure of treatment effect. For continuous outcomes, we used mean differences (MDs) with 95% CI. If different instruments were used to measure the same continuous outcome, we used the standardised mean difference (SMD) with 95% CIs. If outcomes were reported both as binary and continuous outcomes, we analysed binary outcomes in one analysis and continuous outcomes in another analysis. For time‐to‐event outcomes, we planned to use hazard ratios (HRs) and their 95% CIs.

We presented results from the network meta‐analysis (NMA) as summary relative effect sizes for each possible pair of treatments. For each study intervention, we also estimated the ranking probabilities for all treatments of being at each possible rank. We then obtained a treatment hierarchy using the surface under the cumulative ranking curve (SUCRA) and mean ranks (Salanti 2011).

Unit of analysis issues

We did not find any cluster‐randomised or cross‐over trials for inclusion in the review.

For multi‐arm trials, we included all intervention groups meeting the criteria for inclusion in pairwise comparisons.

Dealing with missing data

We reported the amount of missing data for each study in the 'Characteristics of included studies' tables.

We used observed case (OC) data wherever possible. If OC data were not available, we used data imputed using the last observation carried forward (LOCF). This was made explicit in the accompanying text. We assessed the impact of this approach in a sensitivity analysis, by comparing the results of analyses based on the two main approaches (OC and LOCF). Where mixed methods or area under the curve methods were reported by study authors, we extracted the results from these analyses, only if OC results were unavailable.

Assessment of heterogeneity

Assessment of clinical and methodological heterogeneity within treatment comparisons

To evaluate the presence of heterogeneity deriving from different trial designs or different clinical characteristics of study participants, we generated descriptive statistics for trial and study population characteristics across all eligible trials that compared each pair of interventions. Two review authors assessed the presence of clinical heterogeneity within each pairwise comparison by comparing these characteristics. We assessed statistical heterogeneity using the I² statistic and its 95% CI, which measures variability that cannot be attributed to random error. We considered an I² of more than 50% as indicative of substantial heterogeneity (Higgins 2017), and reported this in the results.

Assessment of transitivity across treatment comparisons

We expected that the transitivity assumption would hold, with the assumption that all pairwise comparisons would not differ with respect to the distribution of effect modifiers (e.g. rivastigmine, galantamine, and donepezil would have been administered in a similar way across all included trials).

We evaluated the assumption of transitivity by comparing the clinical and methodological characteristics (potential effect modifiers) across the different pairwise comparisons.

Assessment of reporting biases

In view of the difficulty of detecting and correcting for publication bias and other reporting biases, we aimed to minimise the potential impact of these biases by ensuring a comprehensive search for eligible studies and being alert to duplication of data. If there were 10 or more studies in the pairwise meta‐analysis, we used a funnel plot to explore the possibility of small‐study effects (a tendency for estimates of the intervention effect to be more beneficial in smaller studies) and account for the fact that studies estimate effects for different comparisons. The funnel plots would be aggregate combining all relevant studies.

Data synthesis

Methods for direct treatment comparisons

Our primary analyses compared each cholinesterase inhibitor separately with placebo.

We conducted separate analyses for different daily doses as follows: donepezil 5 mg, 10 mg, and 23 mg; galantamine 16 to 24 mg and other doses; rivastigmine 6 to 12 mg orally or 4.6 mg or 9.5 mg transdermally and other doses. We planned that for rivastigmine, we would combine both oral and transdermal routes, and compare their efficacy and tolerability in subgroup analyses. In practice, we identified no eligible trials of transdermal rivastigmine, and it was not possible to separate doses of oral rivastigmine < 6 mg/day from other doses, so all doses of oral rivastigmine were pooled (see Differences between protocol and review).

We conducted separate analyses for treatment durations up to and including 3 months, 3 to 6 months, 6 to 18 months, and > 18 months.

We performed standard pairwise meta‐analyses in Review Manager 5 using a random‐effects model, due to variable levels of heterogeneity, for every treatment comparison where the summary analysis included at least two studies (Review Manager 2014).

Methods for indirect and mixed comparisons

Network meta‐analysis (NMA) is a method used to synthesise information from a network of trials that address the same question, but involve different interventions. NMA combines direct and indirect evidence across a network of randomised trials into a single effect size, and, under certain assumptions, can increase the precision of the estimates whilst respecting randomisation. The model enabled us to estimate the probability that each intervention is the best for each outcome, given the relative effect sizes as estimated in NMA.

Each cholinesterase inhibitor was considered as a separate (intervention) node in the analysis, with donepezil included as two separate nodes for the two different doses investigated (5 mg and 10 mg daily). Data on different doses were not available for galantamine or rivastigmine, so both of these drugs were included as a single node encompassing all doses investigated. For galantamine, the range of doses was within the manufacturer’s recommended range (for treatment of Alzheimer's disease dementia); however, this was not the case for rivastigmine. The decision set in the NMA was donepezil 5 mg daily, donepezil 10 mg daily, rivastigmine 3 to 12 mg daily, and galantamine 16 to 24 mg daily. The supplementary set was placebo.

We assumed that the three cholinesterase inhibitors (donepezil, rivastigmine, and galantamine) are directly comparable treatments. In other words, we assumed that the distribution of important characteristics (effect modifiers) is the same across all treatment comparisons (Salanti 2012). We defined the placebo node as any drug intervention that did not contain an active ingredient, or any trial arm that contained no investigator‐intended treatment. We used Bayesian analysis for the NMA for its flexibility and more natural interpretation. The Bayesian approach allows ranking of the treatments according to their comparative effectiveness (Mills 2013; Kibret 2014; Rücker 2015).

We performed NMA for each primary outcome measure using MetaInsight (bespoke NMA tool developed by the University of Leicester) (Owen 2019). The free web‐based tool (Owen 2019) is supported by the National Institute for Health Research Complex Reviews Support Unit (NIHR‐CRSU). We also received support in the design, analyses and interpretation of the NMA from NIHR‐CRSU.

Subgroup analysis and investigation of heterogeneity

Assessment of statistical heterogeneity

Assumptions when estimating the heterogeneity

As there were only two to four studies in each direct comparison, in standard pairwise meta‐analysis we assumed a common heterogeneity variance for all direct comparisons. In NMA we assumed a common estimate for the heterogeneity variance across the different comparisons.

Measures and tests for heterogeneity

We based the assessment of statistical heterogeneity in the entire network on the magnitude of the heterogeneity variance parameter (T²) estimated from the NMA models. For dichotomous outcomes, we compared the magnitude of the heterogeneity variance with the empirical distribution as derived by Turner (Turner 2012). We also estimated a total I² value for heterogeneity in the network as described elsewhere (Jackson 2014).

Assessment of statistical inconsistency

Local approaches for evaluating inconsistency

To evaluate the presence of inconsistency locally, we used the loop‐specific approach. This method assesses the consistency assumption in each closed loop of the network separately as the difference between direct and indirect estimates for a specific comparison in the loop (inconsistency factor) (Veroniki 2013). The magnitude of the inconsistency factors and their 95% CIs can then be used to infer information about the presence of inconsistency in each loop. We assumed a common heterogeneity estimate within each loop. We presented the results of this approach graphically in a forest plot using MetaInsight (University of Leicester) (Owen 2019)

Global approaches for evaluating inconsistency

To check the assumption of consistency in the entire network, we used the ‘design‐by‐treatment’ model, as described by Higgins 2017. This method accounts for different sources of inconsistency that can occur when studies with different designs (two‐arm trials versus three‐arm trials) give different results, as well as disagreements between direct and indirect evidence. Using this approach, we have drawn inferences about the presence of inconsistency from any source in the entire network based on a Chi² test. Inconsistency and heterogeneity are interwoven; in order to distinguish between these two sources of variability, we employed the I² or between‐study standard deviation for inconsistency, as it measures variability that cannot be attributed to random error or heterogeneity (within comparison variability). We also sought guidance from the NIHR‐CRSU to address any inconsistencies.

Investigation of heterogeneity and inconsistency

Due to insufficient studies we were unable to perform network meta‐regression or subgroup analyses.

Sensitivity analysis

We conducted sensitivity analyses to compare results obtained with OC data or intention‐to‐treat/LOCF data when both were reported. We identified too few studies within each comparison to conduct planned sensitivity analyses related to risk of bias and severity of VCI.

Summary of findings and assessment of the certainty of the evidence

We summarised our results in 'Summary of findings' tables using the online GRADEpro GDT application (GRADEpro GDT). We reported the estimated treatment effects for the primary outcomes in the tables, using the GRADE approach to rate our confidence that the estimate was correct (Schünemann 2011). GRADE categorises quality using four possible ratings: high, moderate, low, and very low. Results of RCTs rated high certainty will generally have no limitations. Evidence from RCTs is downgraded due to several factors, including: imprecision of effect estimates, risk of bias in included studies, inconsistency of results, indirectness of evidence, and publication bias (Schünemann 2011). These factors are described in the footnotes below the 'Summary of findings' tables.

Results

Description of studies

See Characteristics of included studies and Characteristics of excluded studies.

Results of the search

The search identified a total of 12,214 citations. These citations were assessed initially in Cochrane’s Screen4Me workflow to identify potential reports of randomised trials and then by the Information Specialist of the Cochrane Dementia and Cognitive Improvement Group to exclude obviously irrelevant titles. A total of 736 citations were then passed to the review authors for further assessment. We excluded 689 of these on the basis of titles and abstracts, identifying 47 citations that were related to potentially eligible trials. We assessed these in full‐text where available. These 47 citations described nine unique trials. We included eight trials (46 citations) and excluded one trial (one citation) because participants had CADASIL (see Characteristics of excluded studies table). We identified no further trials through scanning reference lists of the included studies. We identified no ongoing studies. The study identification process is illustrated in Figure 1.

Included studies

We included eight trials in the review (for details see Characteristics of included studies). Three trials investigated donepezil (Black 2003; Wilkinson 2003; Roman 2010); three trials investigated rivastigmine (Mok 2007; Ballard 2008; Narasimhalu 2009); and two trials investigated galantamine (Erkinjuntti 2002; Auchus 2007).

In the three donepezil trials, 2193 participants provided data for one or more of the outcomes. All participants had probable or possible vascular dementia according to the NINDS‐AIREN criteria. Their mean age was 73.9 years. All trials were of 24 weeks' duration. Two trials compared 5 mg and 10 mg of donepezil to placebo and to each other (Black 2003; Wilkinson 2003), and one trial compared 5 mg of donepezil to placebo (Roman 2010).

In the three rivastigmine trials, 800 participants provided data for one or more of the outcomes. Two trials included participants with probable vascular dementia (Mok 2007; Ballard 2008), and one trial included participants with cognitive impairment, no dementia (CIND) following a cerebrovascular accident (Narasimhalu 2009). Mean age of participants was 72.2 years. Two trials ran for 24 weeks (Ballard 2008; Narasimhalu 2009), whilst the third trial ran for 26 weeks (Mok 2007). All trials compared oral rivastigmine with placebo, but final doses of rivastigmine varied across the trials: 3 to 12 mg/day in Ballard 2008; mean of 6 mg/day in Mok 2007; and maximum 9 mg/day in Narasimhalu 2009. No trial reported the proportions of participants receiving different doses. No trial studied transdermal rivastigmine.

In the two galantamine trials (Erkinjuntti 2002; Auchus 2007), 1380 participants provided data for one or more outcomes. Both trials included individuals with probable vascular dementia as defined by the NINDS‐AIREN criteria. The mean age of participants was 73.7 years. Both trials lasted 26 weeks. The final daily doses of galantamine were 16 to 24 mg/day in Auchus 2007 and 24 mg/day in Erkinjuntti 2002.

All included trials were sponsored by pharmaceutical companies.

Excluded studies

We excluded one trial of cholinesterase inhibitors for people with CADASIL (see Characteristics of excluded studies) (Dichgans 2008).

Risk of bias in included studies

Risk of bias in the included studies is summarised in Characteristics of included studies and Figure 2 and Figure 3. We considered only two studies to be at low risk across all domains (Ballard 2008; Narasimhalu 2009). We judged the remaining trials to be at unclear risk of bias across one or more domains.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Allocation

Donepezil

One trial that evaluated donepezil 5 mg did not provide sufficient information to assess sequence generation (unclear risk) (Roman 2010). The other two trials, which evaluated donepezil 5 mg and 10 mg (Black 2003; Wilkinson 2003), reported adequate methods for sequence generation (low risk).

No trial provided sufficient information to enable assessment on allocation concealment (unclear risk).

Rivastigmine

All trials that evaluated rivastigmine reported adequate methods for random sequence generation (low risk) (Mok 2007; Ballard 2008; Narasimhalu 2009).

One trial did not provide sufficient information to enable assessment on allocation concealment (unclear bias) (Mok 2007), whilst allocation concealment for the other two trials was well‐described (low risk).

Galantamine

We assessed one trial that evaluated galantamine as at unclear risk of bias for random sequence generation and allocation concealment (Auchus 2007), whilst the other study had an unclear risk of bias for allocation concealment (Erkinjuntti 2002).

Blinding

Donepezil

All three trials that evaluated donepezil were described as double‐blinded. Two trials reported adequate blinding of participants and personnel (performance bias) (low risk) (Black 2003; Wilkinson 2003). The two trials also provided information regarding blinding of some (but not all) of the outcome measures, therefore we judged the risk of detection bias to be unclear.

We considered one trial to have an unclear risk of performance bias (Roman 2010); the same trial also provided no details of blinding on any of the trial outcome measures and was therefore judged as at unclear risk of detection bias.

Rivastigmine

All three trials that evaluated rivastigmine were reported as double‐blinded. Two trials reported adequate blinding of participants, personnel, and outcome assessment, and were thus assessed as at low risk of performance and detection bias (Ballard 2008; Narasimhalu 2009).

We judged one trial to be at unclear risk of performance and detection bias due to limited information on blinding of trial personnel and no description of blinding of outcome assessors (Mok 2007).

Galantamine

Both trials that evaluated galantamine were reported as double‐blinded. One trial described adequate blinding of study participants, personnel, and outcome assessments, and was thus assessed as at low risk of performance and detection bias (Auchus 2007). However, the other trial provided insufficient information regarding blinding of the study team and outcome measures, and was therefore judged as at unclear risk of performance and detection bias (Erkinjuntti 2002).

Incomplete outcome data

Across the included studies, the percentage of participants who completed the trials ranged from 75.3% to 98%.

Donepezil

Only one of the three trials that evaluated donepezil was at low risk of attrition bias (Roman 2010). The other two studies reported a higher rate of discontinuation in the treatment arms and were thus considered as at unclear risk of attrition bias (Black 2003; Wilkinson 2003).

Rivastigmine

All three trials that evaluated rivastigmine met the criteria for low risk of attrition bias due to balanced numbers across intervention groups with similar reasons for loss to follow‐up.

Galantamine

One trial that evaluated galantamine met the criteria for low risk of attrition bias due to balanced numbers across intervention groups with similar reasons for loss to follow‐up (Auchus 2007). The other study reported a higher rate of discontinuation in the treatment arms and was thus considered as at unclear risk of attrition bias (Erkinjuntti 2002).

Selective reporting

Donepezil

One of the three trials that evaluated donepezil did not report final data for one of the primary outcome measures and was judged as at unclear risk of reporting bias (Roman 2010).

Rivastigmine

All three trials that evaluated rivastigmine reported the benefits and harms of the interventions in the manner specified in the methods section of the trial publications. We judged all three trials as at low risk of reporting bias.

Galantamine

Both trials that evaluated galantamine reported the benefits and harms of the interventions in the manner specified in the methods section of the trial publications. We judged both trials as at low risk of reporting bias.

Other potential sources of bias

We identified no other sources of bias in the included studies, which we judged to be at low risk of bias for this domain.

Effects of interventions

See: Table 1; Table 2; Table 3; Table 4

Summary of findings 1. Donepezil 5 mg compared to placebo for vascular dementia and other vascular cognitive impairments.

Donepezil 5 mg compared to placebo for vascular dementia and other vascular cognitive impairments
Patient or population: vascular dementia and other vascular cognitive impairments Intervention: donepezil 5 mg Comparison: placebo Setting: outpatients
Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with placebo	Risk with donepezil 5 mg
Cognitive function: ADAS‐Cog at 24 weeks Scale from: 0 to 70	Mean change from baseline ranged across control groups from −0.58 to 0.34. (Higher score indicates greater impairment.)	MD −0.92 (−1.44 to −0.40)	‐	1601 (3 RCTs) Black 2003 Roman 2010 Wilkinson 2003	⊕⊕⊕⊕ HIGH	Small benefit of donepezil 5 mg, may not be clinically important
Clinical global impression: CIBIC‐Plus (Improvement) at 24 weeks 7‐point Likert scale	Study population		OR 1.58 (1.10 to 2.27)	712 (2 RCTs) Black 2003 Wilkinson 2003	⊕⊕⊕⊕ HIGH	Small benefit of donepezil 5 mg, may not be clinically important
	283 per 1000 (number demonstrating an improvement with placebo)	381 per 1000 (303 to 472)
Functional performance and activities of daily living: ADFACS at 24 weeks 16‐item score ranging from 0 to 54	Mean change from baseline ranged across control groups from 0.76 to 1.44. (Higher score indicates greater impairment.)	MD −0.73 (−1.52 to 0.06)	‐	798 (2 RCTs) Black 2003 Wilkinson 2003	⊕⊕⊕⊝ MODERATE¹	Probably no effect of donepezil 5 mg
Adverse events (all reported adverse events grouped) at 24 weeks	Study population		OR 1.22 (0.94 to 1.58)	1772 (3 RCTs) Black 2003 Roman 2010 Wilkinson 2003	⊕⊕⊕⊝ MODERATE¹	Probably no difference between donepezil 5 mg and placebo
	830 per 1000	856 per 1000 (821 to 885)
Serious adverse events (excluding death) at 24 weeks	Study population		OR 0.94 (0.72 to 1.22)	1772 (3 RCTs) Black 2003 Roman 2010 Wilkinson 2003	⊕⊕⊕⊝ MODERATE¹	Probably no difference between donepezil 5 mg and placebo
	163 per 1000	155 per 1000 (123 to 192)
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; MD: mean difference; OR: odds ratio; RCT: randomised controlled trial; ADAS‐COG: Alzheimer's Disease Assessment Scale ‐ Cognitive subscale; CIBIC‐Plus: Clinician's Interview‐Based Impression of Change Plus Caregiver Input; ADFACS: Alzheimer's Disease Functional Assessment of Change Scale.
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Open in a new tab

¹Downgraded once due to imprecision: the 95% CI includes a result that would not be considered clinically important and a result that would be considered important.

Summary of findings 2. Donepezil 10 mg compared to placebo for vascular dementia and other vascular cognitive impairments.

Donepezil 10 mg compared to placebo for vascular dementia and other vascular cognitive impairments
Patient or population: vascular dementia and other vascular cognitive impairments Intervention: donepezil 10 mg Comparison: placebo Setting: outpatients
Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with placebo	Risk with donepezil 10 mg
Cognitive function: ADAS‐Cog at 24 weeks Scale from: 0 to 70	Mean change from baseline ranged across control groups from −0.58 to 0.34. (Higher score indicates greater impairment.)	MD −2.21 (−3.07 to −1.35)	‐	608 (2 RCTs) Black 2003 Wilkinson 2003	⊕⊕⊕⊝ MODERATE¹	Probably a small benefit of donepezil 10 mg, may not be clinically important
Clinical global impression: CIBIC‐Plus (Improvement) at 24 weeks 7‐point Likert scale	Study population		OR 1.15 (0.78 to 1.70)	699 (2 RCTs) Black 2003 Wilkinson 2003	⊕⊕⊕⊝ MODERATE¹	Probably no effect with donepezil 10 mg
	283 per 1000	312 per 1000 (235 to 401)
Functional performance and activities of daily living: ADFACS at 24 weeks 16‐item score ranging from 0 to 54	Mean change from baseline ranged across control groups from 0.76 to 1.44. (Higher score indicates greater impairment.)	MD −0.95 (−1.73 to −0.17)	‐	813 (2 RCTs) Black 2003 Wilkinson 2003	⊕⊕⊕⊝ MODERATE²	Probably a small benefit of donepezil 10 mg, may not be clinically important
Adverse events (all adverse events) at 24 weeks	Study population		OR 1.95 (1.20 to 3.15)	813 (2 RCTs) Black 2003 Wilkinson 2003	⊕⊕⊕⊕ HIGH	A slight excess of adverse events with donepezil 10 mg over placebo
	875 per 1000	932 per 1000 (894 to 957)
Serious adverse events (excluding death) at 24 weeks	Study population		OR 1.15 (0.81 to 1.64)	813 (2 RCTs) Black 2003 Wilkinson 2003	⊕⊕⊕⊝ MODERATE¹	Probably no difference between donepezil 10 mg and placebo
	179 per 1000	200 per 1000 (150 to 263)
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; MD: mean difference; OR: odds ratio; RCT: randomised controlled trial; ADAS‐COG: Alzheimer's Disease Assessment Scale ‐ Cognitive subscale; CIBIC‐Plus: Clinician's Interview‐Based Impression of Change Plus Caregiver Input; ADFACS: Alzheimer's Disease Functional Assessment of Change Scale.
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Open in a new tab

¹Downgraded once due to imprecision: the 95% CI includes a result that would not be considered clinically important and a result that would be considered important. ²Downgraded once due to imprecision: different outcome measures used to assess functional performance.

Summary of findings 3. Rivastigmine compared to placebo for vascular dementia and other vascular cognitive impairments.

Rivastigmine compared to placebo for vascular dementia and other vascular cognitive impairments
Patient or population: vascular dementia and other vascular cognitive impairments Intervention: rivastigmine (3 to 12 mg/day, oral preparation) Comparison: placebo Setting: outpatients
Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with placebo	Risk with rivastigmine
Cognitive function: ADAS‐Cog at 24 to 26 weeks Scale from: 0 to 70	Mean change from baseline ranged across control groups from −2.8 to 0.4. (Higher score indicates greater impairment.)	MD 0.03 (−3.04 to 3.1)	‐	748 (2 RCTs) Ballard 2008; Narasimhalu 2009	⊕⊕⊝⊝ LOW^1,2	There may be no effect with rivastigmine.
Functional performance and activities of daily living at 24 to 26 weeks: ADCS‐ADL and IADL scales	On the ADCS‐ADL, the mean change from baseline ranged across the control groups from −0.7 to 5.2. The inverted mean change from baseline on the IADL: 0.1	SMD 0.02 (−0.12 to 0.16)	‐	800 (3 RCTs) Ballard 2008 Mok 2007 Narasimhalu 2009	⊕⊕⊝⊝ LOW^2,3	There may be no effect with rivastigmine.
Adverse events (all adverse events) at 24 to 26 weeks	Study population		OR 3.21 (0.36 to 28.88)	831 (3 RCTs) Ballard 2008 Mok 2007 Narasimhalu 2009	⊕⊝⊝⊝ VERY LOW^1,4	Very uncertain evidence of no difference in number of adverse events with rivastigmine compared to placebo
	287 per 1000	563 per 1000 (126 to 921)
Serious adverse events (excluding death) at 24 to 26 weeks	Study population		OR 1.42 (0.90 to 2.25)	622 (2 RCTs) Ballard 2008; Narasimhalu 2009	⊕⊕⊝⊝ LOW⁴	There may be no difference in number of serious adverse events with rivastigmine compared to placebo.
	118 per 1000	160 per 1000 (107 to 231)
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; MD: mean difference; OR: odds ratio; RCT: randomised controlled trial; SD: standard deviation; SMD: standardised mean difference; ADAS‐COG: Alzheimer's Disease Assessment Scale ‐ Cognitive subscale; ADCS‐ADL: Alzheimer's Disease Cooperative Study ‐ Activities of Daily Living Scale; IADL: Instrumental Activities of Daily Living Scale.
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Open in a new tab

¹Downgraded once due to inconsistency in point estimates. ²Downgraded once due to imprecision: the 95% CI includes a result that would not be considered clinically important and a result that would be considered important. ³Downgraded once due to imprecision: different outcomes measures used to assess functional performance and activities of daily living. The ADCS‐ADL is a 54‐point scale, in which a lower score indicates a greater impairment. The IADL is a 0‐to‐8‐point scale, in which a higher score indicates greater impairment. Consequently, the inverted mean change was used for IADL so that both scales would have the same direction for analysis. ⁴Downgraded twice due to imprecision: the 95% CI includes a result that would not be considered clinically important and a result that would be considered important.

Summary of findings 4. Galantamine compared to placebo for vascular dementia and other vascular cognitive impairments.

Galantamine compared to placebo for vascular dementia and other vascular cognitive impairments
Patient or population: vascular dementia and other vascular cognitive impairments Intervention: galantamine (16 to 24 mg/day) Comparison: placebo Setting: outpatients
Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with placebo	Risk with galantamine
Cognitive function: ADAS‐Cog at 26 weeks Scale from: 0 to 70	Mean change from baseline ranged across the control groups from 0.3 to 1. (Higher score indicates greater impairment.)	MD −2.01 (−3.18 to −0.85)	‐	1188 (2 RCTs) Auchus 2007 Erkinjuntti 2002	⊕⊕⊕⊝ MODERATE¹	Probably a small benefit of galantamine, may not be clinically important
Clinical global impression: CIBIC‐Plus (Improvement) at 26 weeks	Study population		OR 1.32 (1.03 to 1.70)	1326 (2 RCTs) Auchus 2007 Erkinjuntti 2002	⊕⊕⊕⊕ HIGH	Small benefit of galantamine, may not be clinically important
	268 per 1000	326 per 1000 (274 to 384)
Functional performance and activities of daily living at 26 weeks ADCS‐ADL and DAD	On the ADCS‐ADL, the mean change from baseline in the control group was 1.3 (lower scores indicate greater impairment). The mean change from baseline on the DAD score was −4.4 (lower scores indicate greater impairment).	SMD 0.11 (−0.24 to 0.46)	‐	1174 (2 RCTs) Auchus 2007 Erkinjuntti 2002	⊕⊕⊝⊝ LOW^2,3	May be no effect of galantamine
Adverse events (all adverse events) at 26 weeks	Study population		OR 1.57 (1.02 to 2.43)	1378 (2 RCTs) Auchus 2007 Erkinjuntti 2002	⊕⊕⊕⊝ MODERATE³	Probably a slight excess of adverse events with galantamine over placebo
	701 per 1000	787 per 1000 (705 to 851)
Serious adverse events (excluding deaths) at 26 weeks	Study population		OR 1.12 (0.78 to 1.59)	786 (1 RCT) Auchus 2007	⊕⊕⊕⊝ MODERATE⁴	Probably no difference between galantamine and placebo
	185 per 1000	202 per 1000 (150 to 265)
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; MD: mean difference; OR: odds ratio; RCT: randomised controlled trial; SD: standard deviation; SMD: standardised mean difference; ADAS‐COG: Alzheimer's Disease Assessment Scale ‐ Cognitive subscale; CIBIC‐Plus: Clinician's Interview‐Based Impression of Change Plus Caregiver Input; ADCS‐ADL: Alzheimer's Disease Cooperative Study ‐ Activities of Daily Living Scale; DAD: Disability Assessment for Dementia Scale
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Open in a new tab

¹Downgraded once due to inconsistency in point estimates. ²Downgraded once due to imprecision: two scales were combined for analysis for functional performance and activities of daily living. The ADCS‐ADL is a 54‐point scale, in which a lower score indicates a greater impairment. The DAD is a 100‐point scale, in which a lower score indicates a greater impairment (downgraded once). ³Downgraded once due to imprecision: the 95% CI includes a result that would not be considered clinically important and a result that would be considered important. ⁴Downgraded once due to imprecision: only one trial included in result.

See Table 1 for donepezil 5 mg results; Table 2 for donepezil 10 mg results; Table 3 for rivastigmine results; and Table 4 for galantamine results. The 'Summary of findings' tables display estimates of treatment effects for the primary and secondary outcome measures, for each of the four interventions versus placebo, after 24 or 26 weeks of treatment.

No data were available for secondary outcomes: incidence of development of new dementia, carer burden, institutionalisation and quality of life.

Donepezil 5 mg and 10 mg

Primary outcomes

Cognition

The ADAS‐Cog (range 0 to 70) was used to assess changes in cognition from baseline to 24 weeks. Donepezil 5 mg improved cognition slightly, although the size of the change is unlikely to be clinically important (3 trials, 1601 participants: mean difference (MD) −0.92, 95% confidence interval (CI) −1.44 to −0.40; using observed case (OC) or last observation carried forward (LOCF); high‐certainty evidence) (Analysis 1.1). Donepezil 10 mg probably improved cognition slightly more, although the size of the change may still not reach clinical importance (2 trials, 608 participants: MD −2.21, 95% CI −3.07 to −1.35; OC data only; moderate‐certainty evidence) (Analysis 1.2). Heterogeneity was low (I² = 0%) in both analyses.

1.1 — Comparison 1: Cognition, Outcome 1: Donepezil 5 mg ADAS‐Cog (OC or LOCF)

1.2 — Comparison 1: Cognition, Outcome 2: Donepezil 10 mg ADAS‐Cog (OC)

Clinical global impression

The CIBIC‐Plus (7‐point Likert scale) was used to assess changes in clinical global impression from baseline to 24 or 26 weeks. Participants were categorised as improved or stable/worse. Donepezil 5 mg improved clinical global impression slightly (2 trials, 712 participants: odds ratio (OR) 1.58, 95% CI 1.10 to 2.27; using OC or LOCF; high‐certainty evidence) (Analysis 2.1). Donepezil 10 mg probably had little of no effect on clinical global impression (2 trials, 699 participants: OR 1.15, 95% CI 0.78 to 1.70; using OC or LOCF; moderate‐certainty evidence) (Analysis 2.2). Heterogeneity was moderate for donepezil 5 mg and 10 mg (I² =24% and 30%, respectively).

2.1 — Comparison 2: Clinical global impression, Outcome 1: Donepezil 5 mg CIBIC‐Plus (Improvement) (OC or LOCF)

2.2 — Comparison 2: Clinical global impression, Outcome 2: Donepezil 10 mg CIBIC‐Plus (Improvement) (OC or LOCF)

Functional performance in activities of daily living

The Alzheimer's Disease Functional Assessment and Change Scale (ADFACS; range 0 to 54) was used to assess changes in functional performance in activities of daily living, from baseline to 24 or 26 weeks. Donepezil 5 mg probably had little to no effect on functional performance (2 trials, 798 participants; MD −0.73, 95% CI −1.52 to 0.06; using LOCF; moderate‐certainty evidence) (Analysis 3.1). Donepezil 10 mg probably resulted in a slight improvement in functional performance, although the size of the change is unlikely to be clinically important (2 trials, 813 participants: MD −0.95, 95% CI −1.73 to −0.17; using LOCF; moderate‐certainty evidence) (Analysis 3.2). Heterogeneity was low (I² = 0%) in both analyses.

3.1 — Comparison 3: Functional performance in activities of daily living, Outcome 1: Donepezil 5 mg ADFACS (LOCF)

3.2 — Comparison 3: Functional performance in activities of daily living, Outcome 2: Donepezil 10 mg ADFACS (LOCF)

Adverse events

Adverse events reported in the studies included nausea and vomiting, diarrhoea, dizziness, headache, and hypertension. There was probably little to no difference in the number of adverse events between donepezil 5 mg and placebo (3 trials, 1772 participants: OR 1.22, 95% CI 0.94 to 1.58; using LOCF; moderate‐certainty evidence) (Analysis 4.1). Donepezil 10 mg resulted in a slight excess of adverse events compared with placebo (2 trials, 813 participants: OR 1.95, 95% CI 1.20 to 3.15; using LOCF; high‐certainty evidence) (Analysis 4.2). Heterogeneity was low (I² = 0%) in both analyses.

4.1 — Comparison 4: Adverse events, Outcome 1: Donepezil 5 mg (LOCF)

4.2 — Comparison 4: Adverse events, Outcome 2: Donepezil 10 mg (LOCF)

Secondary outcomes

Serious adverse events

There was probably little to no difference in the number of serious adverse events reported between donepezil 5 mg and placebo (3 trials, 1772 participants: OR 0.94, 95% CI 0.72 to 1.22; using LOCF; moderate‐certainty evidence) (Analysis 5.1) or between donepezil 10 mg and placebo (2 trials, 813 participants: OR 1.15, 95% CI 0.81 to 1.64; using LOCF; moderate‐certainty evidence) (Analysis 5.2). Heterogeneity was low for donepezil 5 mg (I² = 0%) and moderate for donepezil 10 mg (I² = 46%).

5.1 — Comparison 5: Serious adverse events, Outcome 1: Donepezil 5 mg (LOCF)

5.2 — Comparison 5: Serious adverse events, Outcome 2: Donepezil 10 mg (LOCF)

Behavioural disturbance

The included trials did not assess this outcome.

Deaths

For donepezil 5 mg, there was between‐trial heterogeneity (I² = 70%) as well as imprecision and risk of bias, so we were very uncertain of the result (3 trials, 1772 participants: OR 1.46, 95% CI 0.60 to 3.50; using LOCF; very low‐certainty evidence) (Analysis 7.1). There was probably little to no difference in deaths that occurred during the trial between donepezil 10 mg and placebo (2 trials, 813 participants: OR 0.94, 95% CI 0.34 to 2.58; using LOCF; moderate‐certainty evidence) (Analysis 7.2). Heterogeneity was low in this analysis (I² = 0%).

7.1 — Comparison 7: Death (LOCF), Outcome 1: Donepezil 5 mg

7.2 — Comparison 7: Death (LOCF), Outcome 2: Donepezil 10 mg

Rivastigmine 3 to 12 mg

Primary outcomes

Cognition

The ADAS‐Cog (range 0 to 70) was used to assess changes in cognition from baseline to 24 or 26 weeks. The results suggest that rivastigmine may have little or no effect on cognition (2 trials, 748 participants: MD 0.03, 95% CI −3.04 to 3.10; using LOCF; low‐certainty evidence) (Analysis 1.3). Heterogeneity was moderate for the pooled result (I² = 63%).

1.3 — Comparison 1: Cognition, Outcome 3: Rivastigmine ADAS‐Cog (LOCF)

Clinical global impression

The included trials did not assess this outcome.

Functional performance in activities of daily living

The ADCS‐ADL (range 0 to 54) and Instrumental Activities of Daily Living Scale (IADL) (range 0 to 8) were used to assess changes in functional performance in activities of daily living, from baseline to 24 or 26 weeks. The results showed that there may be little to no benefit in functional performance with rivastigmine (3 trials, 800 participants: standardised mean difference (SMD) 0.02, 95% CI −0.12 to 0.16; using LOCF; low‐certainty evidence) (Analysis 3.3). Heterogeneity was low for the pooled result (I² = 0%).

3.3 — Comparison 3: Functional performance in activities of daily living, Outcome 3: Rivastigmine ADCS‐ADL & IADL (LOCF)

Adverse events

The evidence is very uncertain regarding the effect of rivastigmine on adverse events (3 trials, 831 participants: OR 3.21, 95% CI 0.36 to 28.88; using LOCF; very low‐certainty evidence) (Analysis 4.3). Heterogeneity was high for the pooled result (I² = 95%).

4.3 — Comparison 4: Adverse events, Outcome 3: Rivastigmine (LOCF)

Secondary outcomes

Serious adverse events

There may be little to no difference in serious adverse events reported between rivastigmine and placebo (2 trials, 622 participants: OR 1.42, 95% CI 0.90 to 2.25; using LOCF; low‐certainty evidence) (Analysis 5.3), (I² = 0%).

5.3 — Comparison 5: Serious adverse events, Outcome 3: Rivastigmine (LOCF)

Behavioural disturbance

The NPI (range 0 to 144) was used to assess changes in behavioural disturbance from baseline to 24 or 26 weeks. There was probably little to no effect of rivastigmine on behavioural disturbance (3 trials, 796 participants: MD 0.21, 95% CI −1.25 to 1.66; using LOCF; moderate‐certainty evidence) (Analysis 6.1). Heterogeneity was low for the pooled result (I² = 0%).

6.1 — Comparison 6: Behavioural disturbance, Outcome 1: Rivastigmine NPI (LOCF)

Deaths

The results showed no evidence of increased deaths with rivastigmine (3 trials, 800 participants: OR 1.45, 95% CI 0.51 to 4.15; using LOCF; low‐certainty evidence) (Analysis 7.3). Heterogeneity was low for the pooled result (I² = 0%).

7.3 — Comparison 7: Death (LOCF), Outcome 3: Rivastigmine

Galantamine 16 to 24 mg

Primary outcomes

Cognition

The ADAS‐Cog (range 0 to 70) was used to assess changes in cognition from baseline to 26 weeks. The results showed that galantamine probably improves cognition slightly, although the size of the change may not reach clinical importance (2 trials, 1188 participants: MD −2.01, 95% CI −3.18 to −0.85; using OC or LOCF; moderate‐certainty evidence) (Analysis 1.4). Heterogeneity was moderate for the pooled result (I² = 57%).

1.4 — Comparison 1: Cognition, Outcome 4: Galantamine ADAS‐Cog (OC or LOCF)

Clinical global impression

The CIBIC‐Plus (7‐point Likert scale) was used to assess changes in clinical global impression from baseline to 24 or 26 weeks. The results showed that galantamine improves clinical global impression although this may not be clinically important (2 trials, 1326 participants: OR 1.32, 95% CI 1.03 to 1.70; using LOCF; high‐certainty evidence) (Analysis 2.3). Heterogeneity was low for the pooled result (I² = 0%).

2.3 — Comparison 2: Clinical global impression, Outcome 3: Galantamine CIBIC‐Plus (Improvement) (LOCF)

Functional performance in activities of daily living

The ADCS‐ADL (range 0 to 54) and Disability Assessment for Dementia Scale (DAD) (range 0 to 100) were used to assess changes in functional performance in ADL, from baseline to 24 or 26 weeks. The results showed that there may be little or no benefit in functional performance with galantamine (2 trials, 1174 participants: SMD 0.11, 95% CI −0.24 to 0.46; using OC or LOCF; low‐certainty evidence) (Analysis 3.4). Heterogeneity was high for the pooled result (I² = 88%).

3.4 — Comparison 3: Functional performance in activities of daily living, Outcome 4: Galantamine ADCS‐ADL & DAD (OC or LOCF)

Adverse events

The results showed that galantamine probably leads to a slight increase in adverse events (2 trials, 1378 participants: OR 1.57, 95% CI 1.02 to 2.43; using LOCF; moderate‐certainty evidence) (Analysis 4.4). Heterogeneity was moderate for the pooled result (I² = 67%).

4.4 — Comparison 4: Adverse events, Outcome 4: Galantamine (LOCF)

Secondary outcomes

Serious adverse events

The results showed that galantamine probably has little or no effect on the occurrence of serious adverse events (1 trial, 786 participants: OR 1.12, 95% CI 0.78 to 1.59; using LOCF; moderate‐certainty evidence) (Analysis 5.4).

5.4 — Comparison 5: Serious adverse events, Outcome 4: Galantamine (LOCF)

Behavioural disturbance

The NPI (range 0 to 144) was used to assess changes in behavioural disturbance from baseline to 24 or 26 weeks. The results showed that galantamine may not improve behavioural disturbance (2 trials, 1151 participants: MD −0.13, 95% CI −4.05 to 3.79; using OC or LOCF; low‐certainty evidence) (Analysis 6.2). Heterogeneity was high for the pooled result (I² = 84%).

6.2 — Comparison 6: Behavioural disturbance, Outcome 2: Galantamine NPI (OC or LOCF)

Deaths

Galantamine probably does not lead to a difference in number of deaths compared with placebo (2 trials, 1378 participants: OR 0.53, 95% CI 0.26 to 1.10; using LOCF; moderate‐certainty evidence) (Analysis 7.4). Heterogeneity was low for the pooled result (I² = 0%).

7.4 — Comparison 7: Death (LOCF), Outcome 4: Galantamine

Sensitivity analyses

Two sets of analyses were compared for the primary cognitive outcome (measured using ADAS‐Cog), one using OC data, and the other using intention‐to‐treat/LOCF data. Two trials investigating donepezil 5 mg and donepezil 10 mg, Black 2003; Wilkinson 2003, and one trial investigating galantamine, Erkinjuntti 2002, reported both OC and LOCF data for cognitive function. Within‐trial comparisons are reported in Analysis 8.1; Analysis 8.2; Analysis 8.3. There is generally little difference between the results for OC and LOCF data, but the sensitivity analysis result for cognition with donepezil 5 mg is more uncertain when LOCF data are used, suggesting the result should be interpreted with caution.

8.1 — Comparison 8: Sensitivity analyses: comparison of OC and LOCF analyses, Outcome 1: Donepezil 5 mg. Outcome: Cognition (ADAS‐Cog)

8.2 — Comparison 8: Sensitivity analyses: comparison of OC and LOCF analyses, Outcome 2: Donepezil 10 mg. Outcome: Cognition (ADAS‐Cog)

8.3 — Comparison 8: Sensitivity analyses: comparison of OC and LOCF analyses, Outcome 3: Galantamine. Outcome: Cognition (ADAS‐Cog)

Results of NMA for treatment effects

We conducted an NMA for the main metrics of benefit and harm (cognition and adverse events), for which results were available in the standard pairwise meta‐analysis, for two or more of the following; donepezil 5 mg, donepezil 10 mg, rivastigmine, and galantamine. For each NMA, Bayesian models employing LOCF data were used (as LOCF data were available for all drugs), for each of the included outcome measures. Cognition (ADAS‐Cog) was used to demonstrate benefit, and rate of adverse events to demonstrate harm. See Table 5 for Summary of findings: NMA results.

1. Network meta‐analysis results: 'Summary of findings' table.

Bayesian network meta‐analysis: 'Summary of findings' table
BENEFITS
Patients: vascular dementia and other vascular cognitive impairments Interventions: cholinesterase inhibitors Comparator: placebo Outcome: cognition (using ADAS‐Cog,* continuous) (follow‐up at 24 or 26 weeks) Setting: outpatients
Total RCTs: 7 Total participants: 3537	Relative effect (95% CI)	Anticipated absolute effect				Certainty of evidence (GRADE)	Ranking	Interpretation of findings
		With treatment	Without treatment		Difference
Donepezil 10 mg 2 RCTs; 608 participants	‐	MD −2.18 (−3.87 to −0.47) Network estimates	Mean change from baseline ranged across control groups from −0.58 to 0.34.		‐	⊕⊕⊕⊕ HIGH	1	Definitely superior to placebo
Galantamine 2 RCTs; 1188 participants (Dose ranges: 16 to 24 mg/day)	‐	MD −1.84 (−3.63 to −0.14) Network estimates	Mean change from baseline ranged across control groups from 0.3 to 1.		‐	⊕⊕⊕⊕ HIGH	2	Definitely superior to placebo
Donepezil 5 mg 3 RCTs; 1601 participants	‐	MD −0.74 (−2.14 to 0.71) Network estimates	Mean change from baseline ranged across control groups from −0.58 to 0.34.		‐	⊕⊕⊕⊝ MODERATE¹	3	Probably not superior to placebo
Rivastigmine 2 RCTs; 748 participants (Dose ranges: 3 to 12 mg/day)	‐	MD −0.53 (−2.35 to 1.94) Network estimates	Mean change from baseline ranged across control groups from −2.8 to 0.4.		‐	⊕⊕⊕⊝ MODERATE²	4	Probably not superior to placebo
HARMS
Patients: vascular dementia and other vascular cognitive impairments Interventions: cholinesterase inhibitors Comparator: placebo Outcome: adverse events (follow‐up at 24 or 26 weeks) Setting: outpatients
Total RCTs: 8 Total participants: 3981	Relative effect (95% CI)	Anticipated absolute effect				Certainty of evidence (GRADE)	Ranking	Interpretation of findings
		With treatment	Without treatment	Difference
Donepezil 10 mg 2 RCTs; 813 participants	OR 2.00 (0.36 to 11.50) Network estimates	933 per 1000	875 per 1000	58 per 1000 more (from 159 fewer to 113 more)		⊕⊕⊕⊝ MODERATE¹	3	Probably inferior to placebo
Galantamine 2 RCTs; 1378 participants (Dose ranges: 16 to 24 mg/day)	OR 1.60 (0.28 to 9.25) Network estimates	790 per 1000	701 per 1000	89 per 1000 more (from 309 fewer to 255 more)		⊕⊕⊕⊝ MODERATE¹	2	Probably inferior to placebo
Donepezil 5 mg 3 RCTs; 1772 participants	OR 1.23 (0.29 to 5.28) Network estimates	857 per 1000	830 per 1000	27 per 1000 more (from 244 fewer to 133 more)		⊕⊕⊝⊝ LOW^1,2	1	May be inferior to placebo
Rivastigmine 3 RCTs; 831 participants (Dose ranges: 3 to 12 mg/day)	OR 3.75 (0.74 to 15.40) Network estimates	607 per 1000	287 per 1000	314 per 1000 more (from 58 fewer to 574 more)		⊕⊝⊝⊝ VERY LOW^1,3	4	Uncertain evidence suggesting inferior to placebo
CI: confidence interval; MD: mean difference; OR: odds ratio; RCT: randomised controlled trial *Alzheimer's Disease Assessment Scale‐Cognitive Subscale (ADAS‐Cog) ranges from 0 to 70, with a higher score indicating greater impairment. A 4‐point ADAS‐Cog change at 6 months is considered clinically meaningful.
GRADE Working Group grades of evidence: High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Open in a new tab

¹Some imprecision (wide 95% CI) (downgraded once). ²Downgraded for unclear risk of bias for more than 2 domains. ³Some inconsistency in point estimates (downgraded once).

Benefit: effect on cognition

Overall cognition data (see network plot in Figure 4), using ADAS‐Cog results, were available from seven RCTs (3537 participants). Seven indirect comparisons were possible. A Bayesian model was used. The between‐study standard deviation was 0.95. The results of the NMA in Table 5 and the forest plot in Figure 5 show the superiority of donepezil 10 mg (MD −2.18, 95% CI −3.87 to −0.47) over the other drugs and placebo, followed in order of superiority by galantamine (MD −1.84, 95% CI −3.63 to −0.14), donepezil 5 mg (MD −0.74, 95% CI −2.14 to 0.71), and rivastigmine (MD −0.53, 95% CI −2.35 to 1.94) (Figure 6). Estimates of benefits as calculated in the NMA are shown in Table 6.

Network plot: Cognition.

The nodes represent an intervention. The solid lines connecting each pair of interventions represent a direct comparison, and the dotted lines an indirect comparison. The numbers on the lines represent the number of trials available for direct comparison.

Forest plot (Bayesian model) network meta‐analysis results: Cognition.

Ranking probabilities for (a) Outcome: cognition, (b) Outcome: adverse events. The horizontal axis shows the possible ranks, and the vertical axis the ranking probabilities. Each line connects the estimated probabilities of being at a particular rank for every intervention.

2. Network meta‐analysis results: estimates of benefits*.

	Donepezil 10 mg	Donepezil 5 mg	Galantamine	Placebo	Rivastigmine
Donepezil 10 mg	Donepezil 10 mg	1.44 (−0.26, 3.14)	0.32 (−2.17, 2.72)	2.18 (0.47, 3.87)	1.63 (−0.76, 4.72)
Donepezil 5 mg	−1.44 (−3.14, 0.26)	Donepezil 5 mg	−1.11 (−3.42, 1.1)	0.74 (−0.71, 2.14)	0.19 (−2.05, 3.09)
Galantamine	−0.32 (−2.72, 2.17)	1.11 (−1.1, 3.42)	Galantamine	1.84 (0.14, 3.63)	1.31 (−1.08, 4.46)
Placebo	−2.18 (−3.87, −0.47)	−0.74 (−2.14, 0.71)	−1.84 (−3.63, −0.14)	Placebo	−0.53 (−2.35, 1.94)
Rivastigmine	−1.63 (−4.72, 0.76)	−0.19 (−3.09, 2.05)	−1.31 (−4.46, 1.08)	0.53 (−1.94, 2.35)	Rivastigmine

Open in a new tab

*Estimates of benefit (cognition, as assessed using Alzheimer's Disease Assessment Scale‐Cognitive Subscale (ADAS‐Cog)) from the Bayesian network meta‐analysis. The estimates are mean differences (95% confidence intervals) of cognition, for treatment columns compared to treatment rows.

Harm: effect on adverse events

Overall adverse events (see network plot in Figure 7) were available from eight RCTs (3981 participants). Seven indirect comparisons were possible. A Bayesian model was used. The between‐study standard deviation was 1.12. The results of the NMA in Table 5 and the forest plot in Figure 8 show the superiority of donepezil 5 mg (OR 1.23, 95% CI 0.29 to 5.28, but low‐certainty evidence) over the other drugs (placebo showing best effect), followed in order of superiority by galantamine (OR 1.60, 95% CI 0.28 to 9.25), donepezil 10 mg (OR 2.00, 95% CI 0.36 to 11.50), and rivastigmine (OR 3.75, 95% CI 0.74 to 15.40, with very low‐certainty evidence) (Figure 6). Estimates of harms as calculated in the NMA are shown in Table 7. The results apply to oral rivastigmine only, as the transdermal route of administration was not investigated in the included trials.

Network plot: Adverse events.

The nodes represent an intervention. The solid lines connecting each pair of interventions represent a direct comparison, and the dotted lines an indirect comparison. The numbers on the lines represent the number of trials available for direct comparison.

Forest plot (Bayesian model) network meta‐analysis results: Adverse events.

3. Network meta‐analysis results: estimates of harms*.

	Donepezil 10 mg	Donepezil 5 mg	Galantamine	Placebo	Rivastigmine
Donepezil 10 mg	Donepezil 10 mg	0.61 (0.11, 3.51)	0.8 (0.07, 9.32)	0.5 (0.09, 2.78)	1.88 (0.17, 16.82)
Donepezil 5 mg	1.63 (0.28, 9.18)	Donepezil 5 mg	1.3 (0.13, 12.77)	0.81 (0.19, 3.48)	3.04 (0.34, 22.08)
Galantamine	1.25 (0.11, 14.72)	0.77 (0.08, 7.48)	Galantamine	0.63 (0.11, 3.63)	2.35 (0.21, 21.19)
Placebo	2.00 (0.36, 11.50)	1.23 (0.29, 5.28)	1.60 (0.28, 9.25)	Placebo	3.75 (0.74, 15.40)
Rivastigmine	0.53 (0.06, 5.94)	0.33 (0.05, 2.98)	0.43 (0.05, 4.86)	0.27 (0.06, 1.35)	Rivastigmine

Open in a new tab

*Estimates of harm (adverse events) from the Bayesian network meta‐analysis. The estimates are odd ratios (95% confidence intervals) of adverse events, for treatment columns compared to treatment rows.

Discussion

By combining and updating three previous Cochrane Reviews, Malouf 2004; Birks 2006; Birks 2013, and adding a network meta‐analysis, we aimed to assess the benefits and harms of cholinesterase inhibitors for vascular dementia and other VCI using contemporary systematic review methods. Through a comprehensive search, we identified eight eligible RCTs with a total of 4373 participants, of which three studied donepezil, three oral rivastigmine, and two galantamine.

It is important to note that the mean clinical important difference (MCID) is not known for any of the outcome measures in the populations studied in this review. Interpretation of effect estimates is therefore challenging. An MCID for cognition of a change of three ADAS‐Cog points over six months in mild Alzheimer's disease has been suggested (Schrag 2012). Manero 2013 proposed an MCID for functional performance of two points on the ADFACS, based on a two‐point difference between cognitively normal people and those with mild cognitive impairment in Alzheimer's disease.

Summary of main results

In the pairwise comparisons, we reported small potential beneficial effects over placebo for: cognition (donepezil 5 mg or 10 mg daily and galantamine 16 to 24 mg daily); clinical global impression (donepezil 5 mg and galantamine); and functional performance (donepezil 10 mg), although the certainty of the evidence varied, and the size of the cognitive and functional effects probably did not reach clinical importance. We found no evidence of effects on behavioural disturbance.

With regard to harms, we found evidence of an increase in the rate of adverse events with donepezil 10 mg and galantamine, but not donepezil 5 mg. We found no evidence of increases in the numbers of serious adverse events or deaths with any of the cholinesterase inhibitors.

For oral rivastigmine, we found no evidence of benefits and were unable to draw any conclusions regarding harmful effects. However, inadequate dosage may have been an issue, as all the trials included participants who achieved a maximum daily dosage of less than 6 mg.

In the NMA of cholinesterase inhibitors, cognition represented benefit and adverse events harm. All drugs ranked above placebo for cognition and below placebo for adverse events. Donepezil 10 mg ranked first in terms of benefit, but third in terms of harms, when considering the network estimates and certainty of the evidence. Galantamine ranked second in terms of both benefit and harm. Rivastigmine had the lowest ranking of the cholinesterase inhibitors in both benefit and harm NMA estimates, but this may reflect possible inadequate doses received by some participants and small sample sizes. No trial investigated transdermal rivastigmine, which may have led to different conclusions.

Overall completeness and applicability of evidence

Despite conducting a comprehensive search for new trials, we found only one additional trial that met our review inclusion criteria since the completion of the previous reviews. The trials appeared to have included a representative population of patients with vascular dementia and VCI seen in clinical practice, where the trials were conducted. We did not include participants with CADASIL, so our results do not apply to this patient group. We selected seven commonly investigated outcome measures used in dementia research; however, not all of the included trials investigated all of the selected outcomes.

No trial investigated transdermal rivastigmine, which is reportedly better tolerated than any of the oral cholinesterase inhibitors when considering gastrointestinal adverse events. Possible underdosing of rivastigmine was also evident, as all trials included participants who received a maximum dose of less than 6 mg/day.

Quality of the evidence

Overall, the certainty of the evidence across the included trials in the pairwise comparisons varied from low to high. A major limitation of this review was the paucity of trials and data. Reliability and quality of data were poor for rivastigmine. Many of the pooled analyses were underpowered. Few trials were included for each cholinesterase inhibitor, and for some comparisons only one trial was included. This makes it difficult to accurately assess whether the effect estimates are reproducible. The certainty of evidence for many of the outcomes was downgraded for imprecision in the GRADE assessment.

A lack of consensus on what is a minimal clinically important difference for the included outcomes for this population renders interpretation difficult and reduces confidence in the conclusions.

In the NMA, we judged the overall certainty of the evidence to be moderate to high for the benefits estimates, but very low to moderate for the harms estimates.

Potential biases in the review process

We did not find any methodological issues in the preparation of this review that could put it at risk of potential biases. We selected a range of databases to search, without any language restrictions. We conducted the NMA according to applicable Cochrane guidance and sought advice from the NIHR‐CRSU when required.

In the NMA, one of the underpinning assumptions is that the participants in the different trials are similar. We found no evidence of systematic differences across the included trials from either clinical or methodological points of view. However, we cannot rule out violation of the transitivity assumption because of the paucity of data.

Agreements and disagreements with other studies or reviews

Despite the inclusion of an additional trial (Roman 2010), we agree with the findings of the previous Cochrane Review (Malouf 2004), that donepezil 5 mg has a potential small benefit that probably does not reach clinical importance in cognition and clinical global impression at 24 weeks' follow‐up. Donepezil 10 mg has a small potential benefit that probably does not reach clinical importance in cognition and functional performance on activities of daily living at 24 weeks' follow‐up. We also agree that donepezil 10 mg leads to a higher rate of adverse events than placebo.

The original rivastigmine review did not pool the results due to differences in study populations (Birks 2013), therefore we cannot directly compare it with our review results. In our review we pooled the results and acknowledged potential heterogeneity in the interpretation of results. The authors of Birks 2013 suggested some evidence of benefit with rivastigmine, which was not supported by the results of our review. They also reported a significant number of withdrawals due to side effects, which is supported by the results of our NMA, which demonstrated oral rivastigmine to be the most inferior of all cholinesterase inhibitors investigated with regard to adverse events.

We agree with the findings of the original galantamine review (Birks 2006), that the data from two trials suggest some advantage over placebo in cognition and clinical global impression measures (but probably not reaching clinical importance), but with a concomitant significant increase in the number of adverse events reported compared with placebo. We agree that no differences were reported between galantamine and placebo for any of the other outcome measures investigated.

Our NMA results also found potential beneficial effects on cognition (but not reaching clinical importance) with donepezil 5 mg, donepezil 10 mg, and galantamine, when compared with placebo. The results also agree that donepezil 10 mg, galantamine, and oral rivastigmine lead to a significantly higher rate of adverse events. These results were to be expected, as only one new trial was included in this update. Our update consolidates the results of the previous reviews through the inclusion of robust 'Risk of bias' assessments and GRADE ratings (Puhan 2014).

Donepezil 5 mg and 10 mg, galantamine, and rivastigmine were also reported to have a potential beneficial effect on cognition in another previous meta‐analysis investigating cholinesterase inhibitors for vascular dementia (Kavirajan 2007). There was also a higher number of reported adverse events with cholinesterase inhibitors, compared with placebo. In comparison with Kavirajan 2007, our update included three additional trials: Roman 2010, Ballard 2008, and Narasimhalu 2009. These results are also supported in a more recent meta‐analysis in vascular dementia patients that demonstrated an improvement in cognition with donepezil 5 mg and 10 mg and galantamine (Chen 2016), but as found in our study, no significant beneficial effect with rivastigmine. They also reported an increase in adverse events with cholinesterase inhibitor use. The review by Chen 2016 included two trials not included in this update, as we excluded trials investigating a primary population of patients with CADASIL and Alzheimer's disease. In a Cochrane Review investigating the efficacy of cholinesterase inhibitors in rarer dementias associated with neurological conditions, no clear benefit on cognition was reported in participants with CADASIL (Li 2015). As in our review, cholinesterase inhibitors were associated with more gastrointestinal side effects compared with placebo (Li 2015).

Authors' conclusions

Implications for practice.

Based on moderate‐ to high‐certainty evidence, we found donepezil 5 mg, donepezil 10 mg, and galantamine 16 to 24 mg to have small beneficial effects on cognition (probably not reaching clinical importance) in participants with vascular dementia and other forms of vascular cognitive impairment, at 24 or 26 weeks' follow‐up. Rates of adverse events were probably higher when compared with placebo for donepezil 10 mg and galantamine, but not for donepezil 5 mg. We could not draw any conclusions regarding adverse events for rivastigmine. We also found that donepezil 5 mg and galantamine had a small potentially beneficial effect that probably did not reach clinical importance on clinical global impression. Donepezil 10 mg had a small beneficial effect that probably did not reach clinical importance on performance of activities of daily living.

Our network meta‐analysis results suggest that donepezil 10 mg ranks highest amongst the cholinesterase inhibitors in terms of cognition, followed by galantamine. In terms of adverse events, the network meta‐analysis results and certainty of the evidence suggest that donepezil 5 mg ranks highest, followed by galantamine. Rivastigmine was not shown to have a beneficial effect on any of the outcomes investigated, but this may reflect underdosing, the poor quality of the evidence, and the high level of heterogeneity between the trials. Conclusions regarding rivastigmine should therefore be interpreted with caution. The results for rivastigmine only apply to the oral preparation, as the transdermal mode of delivery was not investigated in any of the included trials.

Implications for research.

The signal of small benefits found in this review is sufficient to justify further research into donepezil 5 mg, donepezil 10 mg, and galantamine 16 to 24 mg for vascular dementia and other vascular cognitive impairments. For rivastigmine, further trials into the benefits and harms are needed for this patient population, as the results of this review are inconclusive due to poor reliability of data in the included trials. Furthermore, large, good‐quality rivastigmine trials would be beneficial due to the potential underdosing reported in previous trials. The transdermal route of delivery of rivastigmine is a potential area for future research, as there is a need for evidence as to whether this could provide beneficial effects without the known deleterious side effects seen with the oral route. This is particularly important to clinicians, as rivastigmine patches are increasingly being used in some countries as a first‐line treatment option for patients with vascular dementia and other vascular cognitive impairments.

Mixed dementias with a vascular component are common, especially in dementia syndromes seen in older age. It is uncertain whether the results described for the populations in the trials included in this review (who were felt to have predominant vascular pathology) would apply to those with more mixed neurodegeneration. Based on the results of this review, we also consider further research into the benefits of cholinesterase inhibitors for mixed vascular dementia and Alzheimer's disease to be justified.

The included studies all pre‐date 2011, perhaps supporting the suggestion that further trials are required. A longer follow‐up period should be investigated in future trials, as there may be longer‐term effects of cholinesterase inhibitors that are not demonstrated in the trials. Our review has also highlighted that a core outcome set may be needed for research into outcomes in vascular dementia and other vascular cognitive impairments, due to the fact that some important outcomes were infrequently assessed. This would ensure that direct comparisons are possible between trials.

History

Protocol first published: Issue 4, 2019 Review first published: Issue 2, 2021

Acknowledgements

We are grateful to peer reviewers Sebastian Köhler and Peter Langhorne and consumer reviewer Cath Hofstetter for their comments and feedback.

Appendices

Appendix 1. Sources searched and search strategies

Source		Search strategy	Hits retrieved
1. CENTRAL (The Cochrane Library) (issue 8 of 12, 2020) http://crso.cochrane.org/SearchSimple.php [Date of most recent search 19 August 2020]		#1 MESH DESCRIPTOR CADASIL EXPLODE ALL TREES 7 #2 MESH DESCRIPTOR Cerebrovascular Disorders EXPLODE ALL TREES 12799 #3 MESH DESCRIPTOR Dementia, Multi‐Infarct EXPLODE ALL TREES 63 #4 MESH DESCRIPTOR Dementia, Vascular EXPLODE ALL TREES 308 #5 MESH DESCRIPTOR Neurocognitive Disorders EXPLODE ALL TREES 9256 #6 (subcortical ischemic vascular disease):TI,AB,KY 3 #7 (vascular cognitive impairment):TI,AB,KY 156 #8 (vascular dement):TI,AB,KY 754 #9 dement:TI,AB,KY 11281 #10 VaD:TI,AB,KY 423 #11 VCI:TI,AB,KY 67 #12 #1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8 OR #9 OR #10 OR #11 29557 #13 MESH DESCRIPTOR Cholinesterase Inhibitors EXPLODE ALL TREES 1774 #14 MESH DESCRIPTOR TACRINE EXPLODE ALL TREES 106 #15 MESH DESCRIPTOR GALANTAMINE EXPLODE ALL TREES 201 #16 (acetylcholinesterase inhibitor):TI,AB,KY 475 #17 (anti‐alzheimer ADJ2 drug):TI,AB,KY 7 #18 anti‐cholinesteras:TI,AB,KY 18 #19 (anti‐dementia drug):TI,AB,KY 43 #20 (cholinesterase inhibitor):TI,AB,KY 1496 #21 (memory drug):TI,AB,KY 1719 #22 (SDZ ENA 713):TI,AB,KY 7 #23 Anticholinesterase*:TI,AB,KY 133 #24 anti‐cholinesterase:TI,AB,KY 15 #25 aricept:TI,AB,KY 142 #26 cognex:TI,AB,KY 11 #27 donezepil:TI,AB,KY 11 #28 E2020:TI,AB,KY 48 #29 exelon:TI,AB,KY 91 #30 galantamine:TI,AB,KY 583 #31 Nivalin:TI,AB,KY 3 #32 Razadyne:TI,AB,KY 5 #33 reminyl:TI,AB,KY 49 #34 rivastigmine:TI,AB,KY 647 #35 tacrine:TI,AB,KY 223 #36 #13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22 OR #23 OR #24 OR #25 OR #26 OR #27 OR #28 OR #29 OR #30 OR #31 OR #32 OR #33 OR #34 OR #35 5068 #37 #12 AND #36 1649	April 2019 ‐ 1652 August 2020 ‐ 251
2. MEDLINE In‐process and other non‐indexed citations and MEDLINE 1950‐present (Ovid SP) [Date of most recent search 19 August 2020]		1 exp CADASIL/ 2 exp Cerebrovascular Disorders/ 3 exp Dementia, Multi‐Infarct/ 4 exp Dementia, Vascular/ 5 exp Neurocognitive Disorders/ 6 "subcortical ischemic vascular disease".ti,ab. 7 "vascular cognitive impairment".ti,ab. 8 "vascular dement".ti,ab. 9 dement.ti,ab. 10 VaD.ti,ab. 11 VCI.ti,ab. 12 or/1‐11 13 exp Cholinesterase Inhibitors/ 14 exp Tacrine/ 15 exp Galantamine/ 16 exp Donepezil/ 17 "acetylcholinesterase inhibitor".ti,ab. 18 "anti‐alzheimer ADJ2 drug".ti,ab. 19 "anti‐cholinesteras".ti,ab. 20 "anti‐dementia drug".ti,ab. 21 "cholinesterase inhibitor".ti,ab. 22 "memory drug".ti,ab. 23 "SDZ ENA 713".ti,ab. 24 Anticholinesterase*.ti,ab. 25 anti‐cholinesterase.ti,ab. 26 aricept.ti,ab. 27 cognex.ti,ab. 28 donezepil.ti,ab. 29 E2020.ti,ab. 30 exelon.ti,ab. 31 galantamine.ti,ab. 32 galanthamine.ti,ab. 33 Nivalin.ti,ab. 34 Razadyne.ti,ab. 35 reminyl.ti,ab. 36 rivastigmine.ti,ab. 37 tacrine.ti,ab. 38 or/13‐37 39 12 and 38 40 randomized controlled trial.pt. 41 controlled clinical trial.pt. 42 randomized.ab. 43 placebo.ab. 44 drug therapy.fs. 45 randomly.ab. 46 trial.ab. 47 groups.ab. 48 or/40‐47 49 exp animals/ not humans.sh. 50 48 not 49 51 39 and 50	April 2019 ‐ 5899 August 2020 ‐ 584
3. EMBASE 1974 to 18 August 2020 [Date of most recent search 19 August 2020]		1 exp CADASIL/ 2 exp cerebrovascular disease/ 3 exp multiinfarct dementia/ 4 "subcortical ischemic vascular disease".ti,ab. 5 "vascular cognitive impairment".ti,ab. 6 "vascular dement".ti,ab. 7 dement.ti,ab. 8 VaD.ti,ab. 9 VCI.ti,ab. 10 or/1‐9 11 exp cholinesterase inhibitor/ 12 exp tacrine/ 13 exp galantamine/ 14 exp donepezil/ 15 "acetylcholinesterase inhibitor".ti,ab. 16 "anti‐alzheimer ADJ2 drug".ti,ab. 17 "anti‐cholinesteras".ti,ab. 18 "anti‐dementia drug".ti,ab. 19 "cholinesterase inhibitor".ti,ab. 20 "memory drug".ti,ab. 21 "SDZ ENA 713".ti,ab. 22 Anticholinesterase*.ti,ab. 23 anti‐cholinesterase.ti,ab. 24 aricept.ti,ab. 25 cognex.ti,ab. 26 donezepil.ti,ab. 27 E2020.ti,ab. 28 exelon.ti,ab. 29 galantamine.ti,ab. 30 galanthamine.ti,ab. 31 Nivalin.ti,ab. 32 Razadyne.ti,ab. 33 reminyl.ti,ab. 34 rivastigmine.ti,ab. 35 tacrine.ti,ab. 36 or/11‐35 37 10 and 36 38 randomized controlled trial/ 39 controlled clinical trial/ 40 random$.ti,ab. 41 randomization/ 42 intermethod comparison/ 43 placebo.ti,ab. 44 (compare or compared or comparison).ti. 45 ((evaluated or evaluate or evaluating or assessed or assess) and (compare or compared or comparing or comparison)).ab. 46 (open adj label).ti,ab. 47 ((double or single or doubly or singly) adj (blind or blinded or blindly)).ti,ab. 48 double blind procedure/ 49 parallel group$1.ti,ab. 50 (crossover or cross over).ti,ab. 51 ((assign$ or match or matched or allocation) adj5 (alternate or group$1 or intervention$1 or patient$1 or subject$1 or participant$1)).ti,ab. 52 (assigned or allocated).ti,ab. 53 (controlled adj7 (study or design or trial)).ti,ab. 54 (volunteer or volunteers).ti,ab. 55 trial.ti. 56 or/38‐55 57 37 and 56	April 2019 ‐ 2406 August 2020 ‐ 262
4. PsycINFO (Ovid SP) (1806 to 18 August 2020) [Date of most recent search 19 August 2020]		1 exp Cerebrovascular Disorders/ 2 exp Vascular Dementia/ 3 CADASIL.ti,ab. 4 "subcortical ischemic vascular disease".ti,ab. 5 "vascular cognitive impairment".ti,ab. 6 "vascular dement".ti,ab. 7 dement.ti,ab. 8 VaD.ti,ab. 9 VCI.ti,ab. 10 or/1‐9 11 exp Cholinesterase Inhibitors/ 12 exp Galanthamine/ 13 "acetylcholinesterase inhibitor".ti,ab. 14 "anti‐alzheimer ADJ2 drug".ti,ab. 15 "anti‐cholinesteras".ti,ab. 16 "anti‐dementia drug".ti,ab. 17 "cholinesterase inhibitor".ti,ab. 18 "memory drug".ti,ab. 19 "SDZ ENA 713".ti,ab. 20 Anticholinesterase.ti,ab. 21 anti‐cholinesterase.ti,ab. 22 aricept.ti,ab. 23 cognex.ti,ab. 24 donezepil.ti,ab. 25 E2020.ti,ab. 26 exelon.ti,ab. 27 galantamine.ti,ab. 28 galanthamine.ti,ab. 29 Nivalin.ti,ab. 30 Razadyne.ti,ab. 31 reminyl.ti,ab. 32 rivastigmine.ti,ab. 33 tacrine.ti,ab. 34 or/11‐33 35 10 and 34 36 exp Clinical Trials/ 37 randomly.ab. 38 randomi?ed.ti,ab. 39 placebo.ti,ab. 40 groups.ab. 41 "double‐blind".ti,ab. 42 "single‐blind*".ti,ab. 43 RCT.ti,ab. 44 or/36‐43 45 35 and 44 46 from 45 keep 1‐496	April 2019 ‐ 496 August 2020 ‐ 25
5. CINAHL (EBSCOhost) [Date of most recent search 19 August 2020]		S52 S38 AND S51 S51 S39 OR S40 OR S41 OR S42 OR S43 OR S44 OR S45 OR S46 OR S47 OR S48 OR S49 OR S50 S50 MH "Random Assignment" S49 MH "Single‐Blind Studies" or MH "Double‐Blind Studies" or MH "Triple‐Blind Studies" S48 MH "Crossover Design" S47 MH "Factorial Design" S46 MH "Placebos" S45 MH "Clinical Trials" S44 TX "multi‐centre study" OR "multi‐center study" OR "multicentre study" OR "multicenter study" OR "multi‐site study" S43 TX crossover OR "cross‐over" S42 AB placebo* S41 TX random* S40 TX trial* S39 TX "latin square" S38 S11 AND S37 S37 S12 OR S13 OR S14 OR S15 OR S16 OR S17 OR S18 OR S19 OR S20 OR S21 OR S22 OR S23 OR S24 OR S25 OR S26 OR S27 OR S28 OR S29 OR S30 OR S31 OR S32 OR S33 OR S34 OR S35 OR S36 S36 TX tacrine S35 TX rivastigmine S34 TX reminyl S33 TX Razadyne S32 TX Nivalin S31 TX galanthamine S30 TX galantamine S29 TX exelon S28 TX E2020 S27 TX donezepil S26 TX cognex S25 TX aricept S24 TX anti‐cholinesterase S23 TX Anticholinesterase* S22 TX SDZ ENA 713 S21 TX memory drug S20 TX cholinesterase inhibitor* S19 TX anti‐dementia drug* S18 TX anti‐cholinesteras* S17 TX anti‐alzheimer* N2 drug* S16 TX acetylcholinesterase inhibitor* S15 (MH "Donepezil") S14 (MH "Galanthamine") S13 (MH "Tacrine") S12 (MH "Cholinesterase Inhibitors+") S11 S1 OR S2 OR S3 OR S4 OR S5 OR S6 OR S7 OR S8 OR S9 OR S10 S10 TX VCI S9 TX VaD S8 TX dement* S7 TX vascular dement* S6 TX vascular cognitive impairment* S5 TX subcortical ischemic vascular disease S4 (MH "Dementia, Vascular+") S3 (MH "Dementia, Multi‐Infarct") S2 (MH "Cerebrovascular Disorders+") S1 (MH "CADASIL")	April 2019 ‐ 670 August 2020 ‐ 47
6. ISI Web of Science – core collection (ISI Wed of Science) (from 1900) [Date of most recent search 19 August 2020]		TOPIC: (CADASIL OR Vascular Dementia OR vascular cognitive impairment OR VaD OR VCI OR vascular cognitive impairment)AND TOPIC: (Cholinesterase Inhibitors OR TACRINE OR GALANTAMINE OR Donepezil OR acetylcholinesterase inhibitors OR Anticholinesterase OR aricept OR cognex OR exelon) ANDTOPIC: (randomly OR randomised OR randomized OR "random allocat" OR RCT OR CCT OR "double blind" OR "single blind" OR "double blind" OR "single blind*" OR trial)	April 2019 ‐ 450 August 2020 ‐ 21
7. LILACS (BIREME) [Date of most recent search 19 August 2020]		dementia OR demencia OR demência OR CADASIL OR Vascular Dementia OR vascular cognitive impairment OR VaD OR VCI OR vascular cognitive impairment [Words] and Cholinesterase Inhibitors OR TACRINE OR GALANTAMINE OR Donepezil OR acetylcholinesterase inhibitors OR Anticholinesterase OR aricept OR cognex OR exelon [Words] and randomly OR randomised OR randomized OR trial OR ensaio clínico OR control OR controlled [Words]	April 2019 ‐ 14 August 2020 ‐ 1
8. ClinicalTrials.gov (www.clinicaltrials.gov) [Date of most recent search 17 April 2019. NB ICTRP databse not available 19 August 2020]		INTERVENTIONAL AND dementia OR CADASIL OR Vascular Dementia OR vascular cognitive impairment OR VaD OR VCI OR vascular cognitive impairment AND Cholinesterase Inhibitors OR TACRINE OR GALANTAMINE OR Donepezil OR acetylcholinesterase inhibitors OR Anticholinesterase OR aricept OR cognex OR exelon	April 2019 ‐ 268 August 2020 ‐ 10
9. ICTRP (http://apps.who.int/trialsearch) [Date of most recent search 19 August 2020]		dementia OR CADASIL OR Vascular Dementia OR vascular cognitive impairment OR VaD OR VCI OR vascular cognitive impairment AND Cholinesterase Inhibitors OR TACRINE OR GALANTAMINE OR Donepezil OR acetylcholinesterase inhibitors OR Anticholinesterase OR aricept OR cognex OR exelon	April 2019 ‐ 269 August 2020 ‐
10. ALOIS (CRS web) [Date of most recent search 19 August 2020]	1 Cholinesterase Inhibitors OR TACRINE OR GALANTAMINE OR Donepezil OR acetylcholinesterase inhibitors OR Anticholinesterase OR aricept OR cognex OR exelon AND INREGISTER 2 (Vascular OR VAD OR CADASIL OR VCI ):AB,TI AND INREGISTER 3 #1 AND #2			April 2019 ‐ 90 August 2020 ‐ 7
Total hits retrieved			April 2019 ‐ 12214 August 2020 ‐1208
Total after deduplication			April 2019 ‐ 7402 August 2020 ‐ 1037

Open in a new tab

Data and analyses

Comparison 1. Cognition.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 Donepezil 5 mg ADAS‐Cog (OC or LOCF)	3	1601	Mean Difference (IV, Random, 95% CI)	‐0.92 [‐1.44, ‐0.40]
1.2 Donepezil 10 mg ADAS‐Cog (OC)	2	608	Mean Difference (IV, Random, 95% CI)	‐2.21 [‐3.07, ‐1.35]
1.3 Rivastigmine ADAS‐Cog (LOCF)	2	748	Mean Difference (IV, Random, 95% CI)	0.03 [‐3.04, 3.10]
1.4 Galantamine ADAS‐Cog (OC or LOCF)	2	1188	Mean Difference (IV, Random, 95% CI)	‐2.01 [‐3.18, ‐0.85]
1.5 Donepezil 5 mg ADAS‐Cog (LOCF)	3	1772	Mean Difference (IV, Random, 95% CI)	‐0.73 [‐1.67, 0.22]
1.6 Donepezil 10 mg ADAS‐Cog (LOCF)	2	813	Mean Difference (IV, Random, 95% CI)	‐2.17 [‐2.97, ‐1.37]
1.7 Galantamine ADAS‐Cog (LOCF)	2	1279	Mean Difference (IV, Random, 95% CI)	‐1.80 [‐2.55, ‐1.05]

Open in a new tab

1.5 — Comparison 1: Cognition, Outcome 5: Donepezil 5 mg ADAS‐Cog (LOCF)

1.6 — Comparison 1: Cognition, Outcome 6: Donepezil 10 mg ADAS‐Cog (LOCF)

1.7 — Comparison 1: Cognition, Outcome 7: Galantamine ADAS‐Cog (LOCF)

Comparison 2. Clinical global impression.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
2.1 Donepezil 5 mg CIBIC‐Plus (Improvement) (OC or LOCF)	2	712	Odds Ratio (M‐H, Random, 95% CI)	1.58 [1.10, 2.27]
2.2 Donepezil 10 mg CIBIC‐Plus (Improvement) (OC or LOCF)	2	699	Odds Ratio (M‐H, Random, 95% CI)	1.15 [0.78, 1.70]
2.3 Galantamine CIBIC‐Plus (Improvement) (LOCF)	2	1326	Odds Ratio (M‐H, Random, 95% CI)	1.32 [1.03, 1.70]
2.4 Donepezil 5 mg CIBIC‐Plus (Improvement) (LOCF)	2	792	Odds Ratio (M‐H, Random, 95% CI)	1.55 [1.14, 2.10]
2.5 Donepezil 10 mg CIBIC‐Plus (Improvement) (LOCF)	2	795	Odds Ratio (M‐H, Random, 95% CI)	1.11 [0.71, 1.72]

Open in a new tab

2.4 — Comparison 2: Clinical global impression, Outcome 4: Donepezil 5 mg CIBIC‐Plus (Improvement) (LOCF)

2.5 — Comparison 2: Clinical global impression, Outcome 5: Donepezil 10 mg CIBIC‐Plus (Improvement) (LOCF)

Comparison 3. Functional performance in activities of daily living.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
3.1 Donepezil 5 mg ADFACS (LOCF)	2	798	Mean Difference (IV, Random, 95% CI)	‐0.73 [‐1.52, 0.06]
3.2 Donepezil 10 mg ADFACS (LOCF)	2	813	Mean Difference (IV, Random, 95% CI)	‐0.95 [‐1.73, ‐0.17]
3.3 Rivastigmine ADCS‐ADL & IADL (LOCF)	3	800	Std. Mean Difference (IV, Random, 95% CI)	0.02 [‐0.12, 0.16]
3.4 Galantamine ADCS‐ADL & DAD (OC or LOCF)	2	1174	Std. Mean Difference (IV, Random, 95% CI)	0.11 [‐0.24, 0.46]
3.5 Galantamine ADCS‐ADL & DAD (LOCF)	2	1228	Std. Mean Difference (IV, Random, 95% CI)	0.10 [‐0.23, 0.42]

Open in a new tab

3.5 — Comparison 3: Functional performance in activities of daily living, Outcome 5: Galantamine ADCS‐ADL & DAD (LOCF)

Comparison 4. Adverse events.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
4.1 Donepezil 5 mg (LOCF)	3	1772	Odds Ratio (M‐H, Random, 95% CI)	1.22 [0.94, 1.58]
4.2 Donepezil 10 mg (LOCF)	2	813	Odds Ratio (M‐H, Fixed, 95% CI)	1.95 [1.20, 3.15]
4.3 Rivastigmine (LOCF)	3	831	Odds Ratio (M‐H, Random, 95% CI)	3.21 [0.36, 28.88]
4.4 Galantamine (LOCF)	2	1378	Odds Ratio (M‐H, Random, 95% CI)	1.57 [1.02, 2.43]

Open in a new tab

Comparison 5. Serious adverse events.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
5.1 Donepezil 5 mg (LOCF)	3	1772	Odds Ratio (M‐H, Random, 95% CI)	0.94 [0.72, 1.22]
5.2 Donepezil 10 mg (LOCF)	2	813	Odds Ratio (M‐H, Fixed, 95% CI)	1.15 [0.81, 1.64]
5.3 Rivastigmine (LOCF)	2	622	Odds Ratio (M‐H, Random, 95% CI)	1.42 [0.90, 2.25]
5.4 Galantamine (LOCF)	1	786	Odds Ratio (M‐H, Random, 95% CI)	1.12 [0.78, 1.59]

Open in a new tab

Comparison 6. Behavioural disturbance.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
6.1 Rivastigmine NPI (LOCF)	3	796	Mean Difference (IV, Random, 95% CI)	0.21 [‐1.25, 1.66]
6.2 Galantamine NPI (OC or LOCF)	2	1151	Mean Difference (IV, Random, 95% CI)	‐0.13 [‐4.05, 3.79]
6.3 Galantamine NPI (LOCF)	2	1205	Mean Difference (IV, Random, 95% CI)	0.26 [‐2.90, 3.42]

Open in a new tab

6.3 — Comparison 6: Behavioural disturbance, Outcome 3: Galantamine NPI (LOCF)

Comparison 7. Death (LOCF).

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
7.1 Donepezil 5 mg	3	1772	Odds Ratio (M‐H, Fixed, 95% CI)	1.46 [0.60, 3.50]
7.2 Donepezil 10 mg	2	813	Odds Ratio (M‐H, Random, 95% CI)	0.94 [0.34, 2.58]
7.3 Rivastigmine	3	800	Odds Ratio (M‐H, Random, 95% CI)	1.45 [0.51, 4.15]
7.4 Galantamine	2	1378	Odds Ratio (M‐H, Random, 95% CI)	0.53 [0.26, 1.10]

Open in a new tab

Comparison 8. Sensitivity analyses: comparison of OC and LOCF analyses.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
8.1 Donepezil 5 mg. Outcome: Cognition (ADAS‐Cog)	2		Mean Difference (IV, Random, 95% CI)	Subtotals only
8.1.1 OC then LOCF	1	710	Mean Difference (IV, Random, 95% CI)	‐0.32 [‐1.43, 0.80]
8.1.2 OC then LOCF	1	715	Mean Difference (IV, Random, 95% CI)	‐1.55 [‐2.29, ‐0.82]
8.2 Donepezil 10 mg. Outcome: Cognition (ADAS‐Cog)	2		Mean Difference (IV, Random, 95% CI)	Subtotals only
8.2.1 OC then LOCF	1	708	Mean Difference (IV, Random, 95% CI)	‐2.28 [‐3.09, ‐1.47]
8.2.2 OC then LOCF	1	723	Mean Difference (IV, Random, 95% CI)	‐2.52 [‐3.40, ‐1.63]
8.3 Galantamine. Outcome: Cognition (ADAS‐Cog)	1		Mean Difference (IV, Random, 95% CI)	Subtotals only
8.3.1 OC then LOCF	1	995	Mean Difference (IV, Random, 95% CI)	‐2.48 [‐3.34, ‐1.62]

Open in a new tab

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Auchus 2007.

*Study characteristics*
Methods	Randomised, placebo‐controlled, multicentre, double‐blind trial of 6 months duration
Participants	788 participants with probable vascular dementia (NINDS‐AIREN criteria, clinical confirmation on MRI scan); mean age 72.3 years, 64% male participants
Interventions	Galantamine (8 or 12 mg twice a day) versus placebo for 6 months
Outcomes	ADAS‐Cog/11, ADCS‐ADL, CIBIC‐Plus, ADAS‐Cog/13, ADAS‐Cog/10, NPI, EXIT25, adverse events
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: "...subjects were randomly assigned (1:1) to either increasing doses of galantamine or placebo." Comment: unclear random sequence generation
Allocation concealment (selection bias)	Unclear risk	Quote: "After placebo run‐in, all patients were randomly assigned to either increasing doses of galantamine, or placebo." Comment: methods used to randomise unclear
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "To ensure blinding, placebo tablets were administered using the same escalation schedule. Quote: "To ensure blinding, the examiner who measured the drug’s efficacy in a particular subject was not the same person who treated that subject and recorded adverse events (AEs)." Comment: adequate blinding of participants and personnel
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "To ensure blinding, the examiner who measured the drug’s efficacy in a particular subject was not the same person who treated that subject and recorded adverse events (AEs)." Comment: adequate
Incomplete outcome data (attrition bias) All outcomes	Low risk	n = 303 (77%) participants in the treatment arm completed the trial, compared to n = 331 (85%) in the placebo arm. Comment: adequate comparable outcome data available
Selective reporting (reporting bias)	Low risk	All prespecified participant outcome data presented.
Other bias	Low risk	No other sources of bias

Open in a new tab

Ballard 2008.

*Study characteristics*
Methods	Randomised, placebo‐controlled, multicentre, double‐blind trial of 6 months duration
Participants	710 participants with vascular dementia (DSM‐IV) and probable vascular dementia (NINDS‐AIREN criteria); mean age 72.8 years, 62.3% male participants
Interventions	Rivastigmine (3 to 12 mg/day) versus placebo for 6 months
Outcomes	VaDAS, ADCS‐CGIC, ADAS‐Cog, ADCS‐ADL, NPI, MMSE, GDS, adverse events
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "All patients were assigned an identification number. Randomization of drug treatment was performed using a validated system that automated the random assignment of treatment groups." Comment: adequate
Allocation concealment (selection bias)	Low risk	Quote: "Randomization of drug treatment was performed using a validated system that automated the random assignment of treatment groups. Patients, caregivers, study site personnel, or any other personnel involved in the conduct of the study remained unaware of treatment groups until all patients had completed the study..." Comment: well‐described concealment
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "Patients, caregivers, study site personnel, or any other personnel involved in the conduct of the study remained unaware of the treatment groups until all patients had completed the study and all data had been retrieved and finalised for analysis." Comment: well‐described blinding of participants and personnel
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "Patients, caregivers, study site personnel, or any other personnel involved in the conduct of the study remained unaware of the treatment groups until all patients had completed the study and all data had been retrieved and finalised for analysis." Comment: clear blinding of outcome assessment
Incomplete outcome data (attrition bias) All outcomes	Low risk	n = 275 (75.3%) participants in the treatment arm completed the trial, compared to n = 297 (81.6%) in the placebo arm. Comment: adequate comparable outcome data available
Selective reporting (reporting bias)	Low risk	All prespecified participant outcome data presented.
Other bias	Low risk	No other sources of bias

Open in a new tab

Black 2003.

*Study characteristics*
Methods	Randomised, placebo‐controlled, multicentre, double‐blind trial of 6 months duration
Participants	603 participants with probable or possible vascular dementia (NINDS‐AIREN criteria); mean age 73.9 years, 55.2% male participants
Interventions	Donepezil (5 and 10 mg/day) versus placebo for 6 months
Outcomes	ADAS‐Cog, CIBIC‐Plus, MMSE, CDR‐SB, ADFACS, adverse events
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Computer generated randomisation protocol." Comment: adequate
Allocation concealment (selection bias)	Unclear risk	Quote "Patients were assigned to 1 of 3 treatment groups by a computer generated randomisation protocol" Comment: unclear concealment
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Comment: all the key groups of people were successfully blinded
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Quote: "The clinician rating the CIBIC‐plus was blind to the patient’s psychometric test scores and adverse events." Comment: blinding of other outcome measures was not described
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	n = 161 (81.3%) participants in the donepezil 5 mg arm and n = 148 (71.8%) in the donepezil 10 mg arm completed the trial, compared to n = 169 (84.9%) in the placebo arm. Comment: greater attrition in the donepezil 10 mg treatment arm
Selective reporting (reporting bias)	Low risk	All prespecified participant outcome data presented.
Other bias	Low risk	No other sources of bias

Open in a new tab

Erkinjuntti 2002.

*Study characteristics*
Methods	Randomised, placebo‐controlled, multicentre, double‐blind trial of 6 months duration
Participants	592 participants with probable vascular dementia or Alzheimer’s disease combined with CVD (NINDS‐AIREN or NINCDS‐ADRDA criteria plus significant radiological evidence of CVD on CT or MRI); mean age 75.1 years, 53% male participants
Interventions	Galantamine (24 mg/day) versus placebo for 6 months
Outcomes	ADAS‐Cog/11, CIBIC‐Plus, ADAS‐Cog/13, DAD, NPI, adverse events
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "...patients were randomly assigned placebo or galantamine 24 mg/day for 6 months according to a randomisation code generated by the Janssen Research Foundation." Quote: "The randomisation ratio was two to one for galantamine versus placebo." Comment: adequate
Allocation concealment (selection bias)	Unclear risk	Quote: "...patients were randomly assigned... according to a randomisation code generated by Janssen Research Foundation" Comment: unclear concealment
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Quote: "Galantamine and placebo were administered as identical single tablets taken orally twice daily." Comment: adequate blinding of the participants, but limited information on blinding of the study team
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Quote: "To avoid any potential unmasking of treatment allocation, an independent (masked to other components of the study), experienced, trained clinician undertook the CIBIC‐plus assessments to provide an overall impression over the course of the trial." Comment: blinding of other outcome measures was not described
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Quote: "More Galantamine patients than placebo patients discontinued the trial, mostly as a result of adverse events (19.7% vs 8.2%)" Comment: greater attrition in the galantamine treatment arm
Selective reporting (reporting bias)	Low risk	All prespecified participant outcome data reported.
Other bias	Low risk	No other sources of bias

Open in a new tab

Mok 2007.

*Study characteristics*
Methods	Randomised, double‐blind, placebo‐controlled trial of 26 weeks duration
Participants	40 participants with subcortical vascular dementia (diagnosis included CT or MRI evidence); mean age 74.9 years, MMSE 3 to 24, mean 13.2
Interventions	Rivastigmine (mean 6 mg/day) versus placebo
Outcomes	MMSE (Chinese version), FAB, NPI (Chinese version), IADL, CDR, adverse events
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Forty eligible patients were assigned randomly to either placebo (n = 20) or rivastigmine (n = 20) via a computer program generated code." Comment: adequate
Allocation concealment (selection bias)	Unclear risk	Quote: "Eligible patients were assigned randomly... via a computer generated code" Comment: unclear concealment
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Quote: "The study was a 26‐week, double‐blind, placebo‐controlled, single‐centre study in which treatment with 6 mg daily of rivastigmine or placebo was evaluated." Comment: no further description on blinding was reported
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Quote: "The study was a 26‐week, double‐blind, placebo‐controlled, single‐centre study in which treatment with 6 mg daily of rivastigmine or placebo was evaluated." Comment: outcome vulnerable to bias as no blinding of the assessors is described
Incomplete outcome data (attrition bias) All outcomes	Low risk	This is a small study with 20 participants in each arm and only 1 dropout (due to death).
Selective reporting (reporting bias)	Low risk	All prespecified participant outcome data reported.
Other bias	Low risk	No other sources of bias

Open in a new tab

Narasimhalu 2009.

*Study characteristics*
Methods	Randomised, double‐blind, placebo‐controlled trial of 6 months duration
Participants	50 participants with poststroke cognitive impairment (DSM‐IV and MRI evidence), MMSE 16 to 29, mean 23.8; mean age 68.7 years, 44% male participants, 70% Chinese
Interventions	Rivastigmine (up to 9 mg/day) versus placebo
Outcomes	MMSE, Clock Test, Color Trails Test, ADAS‐Cog, cognitive battery, FAB, ADCS‐ADL, MCI, NPI, GDS, adverse events
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Randomization was performed in blocks of 4 using a randomisation list that was produced by the pharmaceutical company's supply management using a validated system that automates the random assignment of treatment groups to randomisation numbers in the specified ratio. The randomisation scheme was reviewed by a Biostatistics Quality Assurance Group and locked by them after approval." Comment: adequate
Allocation concealment (selection bias)	Low risk	Quote: "A trial coordinator blind to the treatment allocation randomised each patient. The patient, psychologist, trial coordinator, clinicians and investigators were blinded to the treatment allocation." Comment: well‐described concealment
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "A trial coordinator blind to the treatment allocation randomised each patient. The patient, psychologist, trial coordinator, clinicians and investigators were blinded to the treatment allocation." Comment: well‐described blinding of participants and personnel
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "A trial coordinator blind to the treatment allocation randomised each patient. The patient, psychologist, trial coordinator, clinicians and investigators were blinded to the treatment allocation." Comment: well‐described blinding of outcome assessment
Incomplete outcome data (attrition bias) All outcomes	Low risk	Quote: "Of the 50 randomised, information on study outcomes was available for 49 patients (98%) as one patient in the placebo group died before a follow‐up cognitive evaluation carried out at week 12." Comment: proportionally similar patients in treatment and placebo groups. Comparable numbers of death and discontinuation of study medication.
Selective reporting (reporting bias)	Low risk	All prespecified participant outcome data reported.
Other bias	Low risk	No other sources of bias

Open in a new tab

Roman 2010.

*Study characteristics*
Methods	Randomised, placebo‐controlled, multicentre, double‐blind trial of 6 months duration
Participants	974 participants with possible or probable vascular dementia (NINDS‐AIREN); mean age 73.0 years, 59% male participants
Interventions	Donepezil (5 mg/day) versus placebo for 6 months
Outcomes	ADAS‐Cog, CIBIC‐Plus, MMSE, CLOX 1/2, EXIT25, DAD, CDR‐SB
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	No method of sequence generation described.
Allocation concealment (selection bias)	Unclear risk	Quote: "Participants were randomly assigned 2:1 to donepezil 5mg or placebo" Comment: method used to randomly assign participants not described. Unclear concealment
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Blinding of participants and personnel not described.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Blinding of outcome assessment not described.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Proportionally similar completion rates: placebo (86.8%) vs treatment (82.6%)
Selective reporting (reporting bias)	Unclear risk	All expected outcomes included; however, CIBIC‐Plus was reported on a graph only.
Other bias	Low risk	No other sources of bias

Open in a new tab

Wilkinson 2003.

*Study characteristics*
Methods	Randomised, placebo‐controlled, multicentre, double‐blind trial of 6 months duration
Participants	616 participants with possible or probable vascular dementia (NINDS‐AIREN) with clinical and radiologic evidence of CVD; mean age 75.0, 60% male participants
Interventions	Donepezil (5 or 10 mg/day) versus placebo
Outcomes	ADAS‐Cog, CIBIC‐Plus, MMSE, CDR‐SB, ADFACS, adverse events
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Patients were assigned to treatment groups by a computer‐generated randomisation protocol." Comment: adequate
Allocation concealment (selection bias)	Unclear risk	Quote: "Patients were assigned to treatment groups by a computer generated randomisation protocol" Comment: unclear concealment
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "Blinding was ensured by the use of identical‐appearing placebo and donepezil tablets." Comment: adequate
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Quote: "ADAS‐cog assessment was performed by a trained clinician. An independent clinician, who was blinded to the patient’s psycho‐ metric test scores and AE information, performed the CIBIC‐plus." Comment: blinding of other outcome measures was not described
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	n = 168 (80.8%) participants in the donepezil 5 mg arm and n = 162 (75.3%) in the donepezil 10 mg arm completed the trial, compared to n = 161 (83.4%) in the placebo arm. Comment: greater attrition in the donepezil 10 mg treatment arm
Selective reporting (reporting bias)	Low risk	All prespecified participant outcome data presented.
Other bias	Low risk	No other sources of bias

Open in a new tab

ADAS‐Cog/10: Alzheimer's Disease Assessment Scale ‐ Cognitive subscale / 10 questions; ADAS‐Cog/11: Alzheimer's Disease Assessment Scale ‐ Cognitive subscale / 11 questions; ADAS‐Cog/13: Alzheimer's Disease Assessment Scale ‐ Cognitive subscale / 13 questions; ADCS‐ADL: Alzheimer's Disease Cooperative Study ‐ Activities of Daily Living; ADCS‐CGIC: Alzheimer's Disease Cooperative Study ‐ Clinical Global Impression of Change; ADFACS: Alzheimer's Disease Functional Assessment of Change Scale; CAB: Community Advisory Board; CDR: Clinical Dementia Rating; CDR‐SB: Clinical Dementia Rating ‐ Sum of Boxes; CIBIC‐Plus: Clinician Interview Based Impression of Change ‐ Plus Carergiver Interview; CLOX1/2: Executive Clock Drawing Task; CVD; Cardiovascular Disease; CT: Computerised Tomography; DAD: Disability Assessment for Dementia; DSM‐IV: Diagnostic and Statistical Manual of Mental Disorders, 4th Edition; EXIT25: Executive Interview 25 item; FAB: Frontal Assessment Battery; GDS: Global Deterioration Scale; IADL; Instrumental Activities of Daily Living; MCI: Mild Cognitive Impairment; MMSE: Mini Mental State Examination; MRI: Magnetic Resonance Imaging; NINCDS‐ADRA: National Institute of Neurological and Communicative Disorders and Stroke ‐ Alzheimer's Disease Related Disorders Association; NINDS‐AIREN: National Institute of Neurological Disorders and Stroke Association ‐ Internationale Pour La Recherche et l'Enseignement en Neurosciences; NPI: Neuropsychiatric Inventory; VaDAS; Vascular Dementia Assessment.

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Dichgans 2008	Study population cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL)

Open in a new tab

Differences between protocol and review

Due to clinical diversity between the studies, we opted to report all direct treatment comparisons using random‐effects models. This was to ensure consistency in reporting (rather than reporting some fixed‐effect and some random‐effects models) and so that more conservative estimates were reported.

For rivastigmine, we were unable to separate participants who received the manufacturer's recommended final dose from participants receiving lower doses, as planned in our protocol. We pooled all doses of rivastigmine.

As fewer than 10 studies were included in the network meta‐analysis, funnel plots were not used.

We were unable to complete a sensitivity analysis including only trials rated as at low risk of bias across all domains, and those trials reporting observed case (OC) and last observation carried forward (LOCF) results for all outcomes, due to an insufficient number of trials. However, we have completed a sensitivity analysis that compared the primary cognitive outcome (measured using ADAS‐Cog scale), one using OC data, and the other using LOCF data.

We were unable to complete a sensitivity analysis including only participants with vascular dementia (excluding those with mild vascular cognitive impairment), due to an insufficient number of trials.

We were unable to complete a network meta‐regression or subgroup analysis (or both) using any effect modifiers, due to an insufficient number of trials.

Contributions of authors

All authors have made a substantial contribution to the conception and design of the protocol; have drafted the work or revised it critically for intellectual content; have approved the final version to be published; and agree to be accountable for all aspects of the work.

Sources of support

Internal sources

No sources of support supplied

External sources

NIHR, UK

This review was supported by the National Institute for Health Research (NIHR), via Cochrane Infrastructure funding to the Cochrane Dementia and Cognitive Improvement Group. The views and opinions expressed therein are those of the authors and do not necessarily reflect those of the Systematic Reviews Programme, NIHR, National Health Service, or the Department of Health.

Declarations of interest

Ceri E Battle: none known

Azmil H Abdul‐Rahim: none known

Susan D Shenkin: none known

Jonathan Hewitt: none known

Terry J Quinn: none known

These authors contributed equally to this work

New

References

References to studies included in this review

Auchus 2007 {published data only}

Auchus A. A randomized, 26-week, double-blind, placebo-controlled trial to evaluate the safety and efficacy of galantamine in the treatment of dementia secondary to cerebrovascular disease. IFPMA register 2004;1:1-3.
Auchus AP, Brashear HR, Salloway S, Korczyn AD, De Deyn PP, Gassmann-Mayer C, for the GAL-INT-26 Study Group. Galantamine treatment of vascular dementia: a randomised trial. Neurology 2007;69:448-58. [DOI] [PubMed] [Google Scholar]
NCT00035191. A placebo controlled trial to evaluate the safety and efficacy of galantamine in the treatment of vascular dementia [A randomized 26 week double blind placebo controlled trial to evaluate the safety and efficacy of galantamine in the treatment of vascular dementia]. https://clinicaltrials.gov/ct2/show/NCT00035191 First received: 03 May 2002.

Ballard 2008 {published data only}

Ballard C, Sauter M, Scheltens P, He Y, Barkhof F, Van Straaten ECW, et al. Efficacy, safety and tolerability of rivastigmine capsules in patients with probable vascular dementia: the Vantage study. Current Medical Research and Opinion 2008;24(9):2561-74. [DOI: 10.1185/03007990802328142] [DOI] [PubMed] [Google Scholar]

Black 2003 {published data only}

Black S, Roman GC, Geldmacker DS, Salloway S, Hecker J, Burns A, et al, the Donepezil 307 Vascular Dementia Study Group. Efficacy and tolerability of Donepezil in vascular dementia: positive results of a 24-week, multi-centre, international, randomised, placebo-controlled trial. Stroke 2003;34:2323-32. [DOI: 10.1161/01.STR.0000091396.95360.E1] [DOI] [PubMed] [Google Scholar]
Doody R, Pratt RD, Posner H, Kumar D. Vascular dementia patients who receive Donepezil treatment for 12 to 18 months maintain cognitive benefits. Neurobiology of Aging 2004;25:469. [Google Scholar]
Geldmacher DS, Pratt PD, Perdomo CA. Donepezil slows functional deterioration in patients with vascular dementia. European Journal of Neurology 2002;9:35. [Google Scholar]
Pratt D. Patient populations in clinical trials of the efficacy and tolerability of donepezil in patients with vascular dementia. Journal of the Neurological Sciences 2002;203:57-65. [DOI] [PubMed] [Google Scholar]
Pratt R, Perdomo C, The Donepezil 307 and 308 Study Groups. Donepezil is well tolerated in patients with vascular dementia: a comparison of safety and tolerability results from randomized, placebo-controlled clinical trials in vascular dementia patients and Alzheimer's disease patients. Neurobiology of Aging 2002;23(1 Suppl S57):S57. [Google Scholar]
Pratt RD, Perdomo CA, The Donepezil 307 VaD Study Group. Population characteristics and pattern of cognitive decline in patients with vascular dementia enrolled in two 24-week, randomized, double-blind, placebo-controlled trials. In: 7th International Geneva/Springfield Symposium on Advances in Alzheimer therapy; 2002 Apr 03-06; Geneva. 2002.
Pratt RD, Perdomo CA, The Donepezil VaD 307 and 308 Study Groups. Donepezil-treated patients with probable vascular dementia demonstrate cognitive benefits. Annals of the New York Academy of Sciences 2002;977:513-22. [DOI] [PubMed] [Google Scholar]
Roman G, Black S, Royall DR, Surick I, Kumar D, Schindler R, et al. Efficacy and safety of donepezil in vascular dementia: results from the largest double-blind trial in vascular dementia patients. European Neuropsychopharmacology 2006;16:S481-2. [Google Scholar]
Roman GC, Pratt RD, Perdomo CA, The Donepezil Study Group. Donepezil improves cognition in patients with vascular dementia: results from Study 307, a 24-week randomized, double-blind, placebo-controlled trial. In: 8th International Conference on Alzheimer's Disease and Related Disorders; 2002 Jul 20-25, Stockholm. 2002.
Schindler R, Perdomo CA, Pratt RD. Donepezil provides significant benefits to patients with vascular dementia in their ability to perform everyday activities. In: 8th Congress of the European Federation of the Neurological Sciences; 2004 Sep 04-07; Paris. 2004.
Seltzer B, Perdomo CA, Pratt RD, Schindler R. Donepezil provides significant benefits to patients with vascular dementia in their ability to perform everyday activities. Neurobiology of Aging 2004;25:470. [Google Scholar]

Erkinjuntti 2002 {published data only}

Bullock R, Erkinjuntti T, Lilienfeld S, GAL-INT-6 Study Group. Management of patients with Alzheimer's disease plus cerebrovascular disease: 12-month treatment with galantamine. Dementia and Geriatric Cognitive Disorders 2004;17:29-34. [DOI] [PubMed] [Google Scholar]
Burke W, Lilienfeld S. Galantamine improves behaviour and relieves caregiver distress in Alzheimer's disease (AD), vascular dementia and AD with cerebrovascular disease. In: 8th international conference on Alzheimer's disease and related disorders; 2002 Jul 20-25; Stockholm. 2002.
Erkinjuntti T, Gauthier S, Bullock R, Kurz A, Hammond G, Schwalen S, et al. Galantamine treatment in Alzheimer's disease with cerebrovascular disease: responder analyses from a randomized, controlled trial (GAL-INT-6). Journal of Psychopharmacology 2008;22(7):761-8. [DOI] [PubMed] [Google Scholar]
Erkinjuntti T, Kurz A, Gauthier A, Bullock R, Lillienfeld S, Damaraju CV. Efficacy of Galantamine in probable vascular dementia and Alzheimer's disease combined with cardiovascular disease: a randomised trial. Lancet 2002;359:1283-90. [DOI] [PubMed] [Google Scholar]
Erkinjuntti T, Kurz A, Small GW, Bullock R, Lilienfeld S, Damaraju CV, GAL-INT-6 Study Group. An open-label extension trial of galantamine in patients with probable vascular dementia and mixed dementia. Clinical Therapeutics 2003;25:1765-82. [DOI] [PubMed] [Google Scholar]
Erkinjuntti T, Lilienfeld S, Damaraju CV. Long-term treatment with galantamine is effective in slowing cognitive decline in patients with probable vascular dementia. In: 7th International Geneva/Springfield Symposium on Advances in Alzheimer Therapy; 2002 Apr 03-06; Geneva. 2002.
Erkinjuntti T. Broad therapeutic benefits in patients with probable vascular dementia or Alzheimer's disease with cerebrovascular disease after treatment with galantamine. European Journal of Neurology 2002;9:545. [DOI] [PubMed] [Google Scholar]
Kertesz A. Efficacy of galantamine in probable vascular dementia and Alzheimer's disease combined with cerebrovascular disease: a randomized trial. Current Neurology and Neuroscience Reports 2002;2(6):503-4. [DOI] [PubMed] [Google Scholar]
Kertesz A. Galantamine in vascular dementia and Alzheimer's disease combined with cerebrovascular disease. Current Neurology and Neuroscience Reports 2002;2(6):503-4. [DOI] [PubMed] [Google Scholar]
Kurz A, Erkinjuntti T, Bullock R, Lilienfeld S, Brashear HR. Long-term safety, tolerability, and efficacy of galantamine in the treatment of probable vascular dementia or Alzheimer's disease with cerebrovascular disease: an interim analysis. International Psychogeriatrics 2003;11:273. [Google Scholar]
Kurz A, Lilienfeld S, Damaraju CV. Galantamine is safe and effective for the long-term treatment of cognitive decline in patients with Alzheimer disease with cerebrovascular components or probable vascular dementia. In: 7th International Geneva/Springfield Symposium on Advances in Alzheimer therapy; 2002 Apr 03-06; Geneva. 2002.
Kurz A. Non-cognitive benefits of galantamine (Reminyl(R)) treatment in vascular dementia. Acta Neurologica Scandinavica 2002;106:19-24. [DOI] [PubMed] [Google Scholar]
Kurz AF, Erkinjuntti T, Small GW, Lilienfeld S, Venkata Damaraju CR. Long-term safety and cognitive effects of galantamine in the treatment of probable vascular dementia or Alzheimer's disease with cerebrovascular disease. European Journal of Neurology 2003;10:633-40. [DOI] [PubMed] [Google Scholar]
Small G, Erkinjuntti T, Kurz A, Lilienfeld S. Galantamine in the treatment of cognitive decline in patients with vascular dementia or Alzheimer's disease with cerebrovascular disease. CNS Drugs 2003;17:905-14. [DOI] [PubMed] [Google Scholar]

Mok 2007 {published data only}

Mok V, Wong A, Ho S, Leung T, Lam WWM, Wong KS. Rivastigmine in Chinese patients with subcortical vascular dementia. Neuropsychiatric Disease and Treatment 2007;3(6):943-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

Narasimhalu 2009 {published data only}

Narasimhalu K, Effendy S, Sim CH, Lee JM, Chen I, Hia SB, et al. A randomised controlled trial of Rivastigmine in patients with cognitive impairment no dementia because of cerebrovascular disease. Acta Neurologica Scandinavica 2010;121:217-24. [DOI] [PubMed] [Google Scholar]

Roman 2010 {published data only}

Pratt RD, Perdomo CA. Donepezil improves cognitive function in patients with vascular dementia: results from study 307, a 24-week, randomized, double-blind, placebo-controlled trial. Portal of Geriatrics Online Education 2002;7:A161-2.
Roman GC, Salloway S, Black S, Royall DR, DeCarli C, Weiner MW, et al. Randomised, placebo-controlled, clinical trial of Donepezil in vascular dementia. Stroke 2010;41(6):1213-21. [DOI: 10.1161/STROKEAHA.109.570077] [DOI] [PMC free article] [PubMed] [Google Scholar]

Wilkinson 2003 {published data only}

Aguilar M, Roman G, Black S, Royall D, Surick I, Kumar D, et al. Efficacy and safety of donepezil in vascular dementia results from largest double-blind trial in vascular dementia patients. In: 10th International Conference on Alzheimer's Disease and Related Disorders; 2006 Jul 15-20; Madrid. 2006.
Bayer A, Pratt RD, Kumar D. A comparison of the cognitive benefits of donepezil in patients with cortical and subcortical vascular dementia. In: 8th Congress of the European Federation of the Neurological Sciences; 2004 Sep 04-07; Paris. 2004.
Boundy K, Pratt R. Donepezil provides significant benefits in patients with vascular dementia. Internal Medicine Journal 2003;33:A46. [Google Scholar]
Farlow M. Efficacy of donepezil in vascular dementia. Neurology 2003;61(4):429. [DOI] [PubMed] [Google Scholar]
Pratt R, Perdomo C. Donepezil is well tolerated in patients with vascular dementia: a comparison of safety and tolerability results from randomized, placebo-controlled clinical trials in vascular dementia patients and Alzheimer's disease patients. Neurobiology of Aging 2002;1:S57. [Google Scholar]
Pratt RD, Perdomo CA, The Donepezil 308 VaD Study Group. Cognitive and global benefits of donepezil in vascular dementia: results from study 308, a 24-week, randomized, double-blind, placebo-controlled trial. In: 7th International Geneva/Springfield Symposium on Advances in Alzheimer therapy; 2002 Apr 03-06; Geneva. 2002.
Roman GC, Wilkinson DG, Doody RS, Black SE, Salloway SP, Schindler RJ. Donepezil in vascular dementia: combined analysis of two large-scale clinical trials. Dementia and Geriatric Cognitive Disorders 2005;20(6):338-44. [DOI] [PubMed] [Google Scholar]
Salloway S, Pratt RD, Perdomo CA. Donepezil-treated patients with vascular dementia demonstrate cognitive and global benefits: results from study 308, a 24-week, randomized, double-blind, placebo-controlled trial. Neurobiology of Aging 2002;1:57. [Google Scholar]
Salloway SP, Pratt RD, Perdomo CA. Donepezil is well tolerated in patients with vascular dementia: a comparison of tolerability in vascular dementia patients and Alzheimer's disease patients. European Journal of Neurology 2002;9:165. [Google Scholar]
Wilkinson D, Doody R, Helme R, Taubman K, Mintzer J, Kertesz A, et al, The Donepezil 308 Study Group. Donepezil in vascular dementia: a randomised, placebo-controlled study. Neurology 2003;61:479-86. [DOI] [PubMed] [Google Scholar]
Wilkinson D, Roman G, Salloway S, Hecker J, Boundy K, Kumar D, Posner H, Schindler R. The long-term efficacy and tolerability of donepezil in patients with vascular dementia. Int J Geriat Psychiatry 2010;25:305-315. [DOI] [PubMed] [Google Scholar]

References to studies excluded from this review

Dichgans 2008 {published data only}

Dichgans M, Markus HS, Salloway S, Verkkoniemi A, Moline M, Wang Q, et al. Donepezil in patients with subcortical vascular cognitive impairment: a randomised double-blind trial in CADASIL. Lancet Neurology 2008;7:310-8. [DOI] [PubMed] [Google Scholar]

Additional references

APA 2013

American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 5th edition. Washington, DC: American Psychiatric Society, 2013. [Google Scholar]

Birks 2006

Birks J. Cholinesterase inhibitors for Alzheimer's disease. Cochrane Database of Systematic Reviews 2006, Issue 1. Art. No: CD005593. [DOI: 10.1002/14651858.CD005593] [DOI] [PMC free article] [PubMed] [Google Scholar]

Birks 2013

Birks J, McGuinness B, Craig D. Rivastigmine for vascular cognitive impairment. Cochrane Database of Systematic Reviews 2013, Issue 5. Art. No: CD004744. [DOI: 10.1002/14651858.CD004744.pub3] [DOI] [PMC free article] [PubMed] [Google Scholar]

Chen 2016

Chen Y, Zhang J, Wang Y, Yuan J, Hu W. Efficacy of cholinesterase inhibitors in vascular dementia: an updated meta-analysis. European Neurology 2016;75:132-41. [DOI] [PubMed] [Google Scholar]

Colović 2013

Colović MB, Krstić DZ, Lazarević-Pašti TD, Bondžić AM, Vasić VM. Acetylcholinesterase inhibitors: pharmacology and toxicology. Current Neuropharmacy 2013;11(3):315-35. [DOI] [PMC free article] [PubMed] [Google Scholar]

Craig 2006

Craig D, Birks J. Galantamine for vascular cognitive impairment. Cochrane Database of Systematic Reviews 2006, Issue 1. Art. No: CD004746. [DOI: 10.1002/14651858.CD004746.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]

Dawbarn 2001

Dawbarn D, Shelly JA. Neurobiology of Alzheimer’s Disease. 2nd edition. Oxford (UK): Oxford University Press, 2001. [Google Scholar]

Dichgans 2017

Dichgans M, Leys D. Vascular cognitive impairment. Circulation Research 2017;120:573-91. [DOI: 10.1161/CIRCRESAHA.116.308426] [DOI] [PubMed] [Google Scholar]

Erkinjuntti 2004

Erkinjuntti T, Román G, Gauthier S, Feldman H, Rockwood K. Emerging therapies for vascular dementia and vascular cognitive impairment. Stroke 2004;35:1010-7. [DOI] [PubMed] [Google Scholar]

GRADEpro GDT [Computer program]

McMaster University (developed by Evidence Prime) GRADEpro GDT. Version accessed 12 July 2019. Hamilton (ON): McMaster University (developed by Evidence Prime), 2008. Available at gradepro.org.

Higgins 2017

Higgins JP, Altman DG, Sterne JA. Chapter 8: Assessing risk of bias in included studies. In: Higgins JP, Churchill R, Chandler J, Cumpston MS, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 5.2.0 (updated June 2017). The Cochrane Collaboration, 2017. Available from training.cochrane.org/handbook/archive/v5.2.

Hort 2010

Hort J, O’Brien JT, Gainotti G, Pirttila T, Popescu BO, Rektorova I, et al. EFNS guidelines for the diagnosis and management of Alzheimer’s disease. European Journal of Neurology 2010;17:1236-48. [DOI] [PubMed] [Google Scholar]

Jackson 2014

Jackson D, Barrett JK, Rice S, White IR, Higgins JP. A design-by-treatment interaction model for network meta-analysis with random inconsistency effects. Statistics in Medicine 2014;33:3639-54. [DOI] [PMC free article] [PubMed] [Google Scholar]

Kavirajan 2007

Kavirajan H, Schneider LS. Efficacy and adverse effects of cholinesterase inhibitors and memantine in vascular dementia: a meta-analysis of randomised controlled trials. Lancet Neurology 2007;6:782-92. [DOI] [PubMed] [Google Scholar]

Kibret 2014

Kibret T, Richer D, Beyene J. Bias in identification of the best treatment in a Bayesian network meta-analysis for binary outcome: a simulation study. Clinical Epidemiology 2014;6:451-60. [DOI: 10.2147/CLEP.S69660] [DOI] [PMC free article] [PubMed] [Google Scholar]

Li 2015

Li Y, Hai S, Zhou Y, Dong BR. Cholinesterase inhibitors for rarer dementias associated with neurological conditions. Cochrane Database of Systematic Reviews 2015, Issue 3. Art. No: CD009444. [DOI: 10.1002/14651858.CD009444.pub3] [DOI] [PMC free article] [PubMed] [Google Scholar]

Lilienfeld 2002

Lilienfeld S. Galantamine: a novel cholinergic drug with a unique dual mode of action for the treatment of patients with Alzheimer’s disease. CNS Drug Reviews 2002;8:159-76. [DOI] [PMC free article] [PubMed] [Google Scholar]

Malouf 2004

Malouf R, Birks J. Donepezil for vascular cognitive impairment. Cochrane Database of Systematic Reviews 2004, Issue 1. Art. No: CD004395. [DOI: 10.1002/14651858.CD004395.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]

Manero 2013

Manero RM, Casals-Coll M, Sánchez-Benavides G, Rodríguez-de los Reyes ON, Aguilar M, Badenes D, et al, for the NEURONORMA Study Team. Diagnostic validity of the Alzheimer’s Disease Functional Assessment and Change Scale in mild cognitive impairment and mild to moderate Alzheimer’s disease. Dementia and Geriatric Cognitive Disorders 2013;37(5):366-75. [DOI: 10.1159/000350800] [DOI] [PubMed] [Google Scholar]

Marshall 2018

Marshall IJ, Noel-Storr AH, Kuiper J, Thomas J, Wallace BC. Machine Learning for Identifying Randomized Controlled Trials: an evaluation and practitioner’s guide. Research Synthesis Methods 2018;9(4):602-14. [DOI] [PMC free article] [PubMed] [Google Scholar]

McDonald 2017

McDonald S, Noel-Storr AH, Thomas J. Harnessing the efficiencies of machine learning and Cochrane Crowd to identify randomised trials for individual Cochrane reviews. In: Global Evidence Summit; 2017 Sep 13-16; Cape Town. 2017.

Mills 2013

Mills EJ, Thorlund K, Ioannidis JPA. Demystifying trial networks and network meta-analysis. BMJ 2013;346:f2914. [DOI: 10.1136/bmj.f2914] [DOI] [PubMed] [Google Scholar]

NICE 2018

National Institute for Health and Care Excellence. NICE guidelines: donepezil, galantamine, rivastigmine and memantine for the treatment of Alzheimer's disease. Technology appraisal guidance [TA217]. www.nice.org.uk/guidance/ta217 (accessed 6 January 2019).

Noel‐Storr 2018

Noel-Storr AH, Project Transform Team. Cochrane Crowd: new ways of working together to produce health evidence. In: Evidence Live; 2018 Jun 18-20; Oxford. 2018.

O'Brien 2003

O’Brien JT, Erkinjuntti T, Reisberg B, Roman G, Sawada T, Pantoni L, et al. Vascular cognitive impairment. Lancet Neurology 2003;2:89-98. [DOI] [PubMed] [Google Scholar]

Owen 2019

Owen RK, Bradbury N, Xin Y, Copper N, Sutton A. MetaInsight: an interactive web‐based tool for analyzing, interrogating, and visualizing network meta‐analyses using R‐shiny and netmeta. Research Synthesis Methods 2019;10(4):569-81. [DOI: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

Perry 1997

Perry E, Kay DW. Some developments in brain ageing and dementia. British Journal of Biomedical Science 1997;54:201-15. [PubMed] [Google Scholar]

Puhan 2014

Puhan MA, Schünemann HJ, Murad MH, Li T, Brignardello-Petersen R, Singh JA, et al GRADE Working Group. A GRADE Working Group approach for rating the quality of treatment effect estimates from network meta-analysis. BMJ 2014;349:g5630. [DOI] [PubMed] [Google Scholar]

Qiu 2009

Qiu C, Kivipelto M, Strauss E. Epidemiology of Alzheimer's disease: occurrence, determinants, and strategies toward intervention. Dialogues in Clinical Neuroscience 2009;11:111-28. [DOI] [PMC free article] [PubMed] [Google Scholar]

Review Manager 2014 [Computer program]

Nordic Cochrane Centre, The Cochrane Collaboration Review Manager 5 (RevMan 5). Version 5.3. Copenhagen: Nordic Cochrane Centre, The Cochrane Collaboration, 2014.

Ritter 2015

Ritter A, Pillai JA. Treatment of vascular cognitive impairment. Current Treatment Options in Neurology 2015;17:367. [DOI: 10.1007/s11940-015-0367-0] [DOI] [PubMed] [Google Scholar]

Roman 1993

Roman G, Tatemichi T, Erkinjuntti T, Cummings JL, Masdeu JC, Garcia JH, et al. Vascular dementia: diagnostic criteria for research studies. Report of the NINDS-AIREN International Workshop. Neurology 1993;43:250-60. [DOI] [PubMed] [Google Scholar]

Rücker 2015

Rücker G, Schwarzer G. Ranking treatments in frequentist network meta-analysis works without resampling methods. BMC Medical Research Methodology 2015;15:58. [DOI: 10.1186/s12874-015-0060-8] [DOI] [PMC free article] [PubMed] [Google Scholar]

Salanti 2011

Salanti G, Ades AE, Ioannidis JP. Graphical methods and numerical summaries for presenting results from multiple-treatment meta-analysis: an overview and tutorial. Journal of Clinical Epidemiology 2011;64:163-71. [DOI] [PubMed] [Google Scholar]

Salanti 2012

Salanti G. Indirect and mixed-treatment comparison, network, or multiple-treatments meta-analysis: many names, many benefits, many concerns for the next generation evidence synthesis tool. Research Synthesis Methods 2012;3:80-97. [DOI] [PubMed] [Google Scholar]

Schrag 2012

Schrag A, Schott JM, Alzheimer's Disease Neuroimaging Initiative. What is the clinically relevant change on the ADAS-Cog? Journal of Neurology, Neurosurgery, and Psychiatry 2012;83(2):171-3. [DOI: 10.1136/jnnp-2011-300881] [DOI] [PubMed] [Google Scholar]

Schünemann 2011

Schünemann HJ, Oxman AD, Higgins JP, Vist GE, Glasziou P, Guyatt GH. Chapter 11: Presenting results and ‘Summary of findings’ tables. In: Higgins JP, Green S, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. Available from training.cochrane.org/handbook/archive/v5.1/.

Skrobot 2017

Skrobot OA, O’Brien J, Black S, Chen C, DeCarli C, Erkinjuntti T, et al. The Vascular Impairment of Cognition Classification Consensus Study. Alzheimer’s & Dementia 2017;13:624-33. [DOI] [PubMed] [Google Scholar]

Thomas 2017

Thomas J, Noel-Storr AH, Marshall I, Wallace B, McDonald S, Mavergames C, et al. Living Systematic Review Network. Living Systematic Reviews: 2. Combining Human and Machine Effort. Journal of Clinical Epidemiology 2017;91:31-7. [DOI] [PubMed] [Google Scholar]

Toghi 1996

Toghi H, Abe T, Kimura M, Saheki M, Takahashi S. Cerebrospinal fluid acetylcholine and choline in vascular dementia of Binswanger and multiple small infarcts types as compared with Alzheimer-type dementia. Journal of Neural Transmission 1996;103:1211-20. [DOI] [PubMed] [Google Scholar]

Turner 2012

Turner RM, Davey J, Clarke MJ, Thompson SG, Higgins JP. Predicting the extent of heterogeneity in meta-analysis, using empirical data from the Cochrane Database of Systematic Reviews. International Journal of Epidemiology 2012;41:818-27. [DOI] [PMC free article] [PubMed] [Google Scholar]

van der Flier 2018

Flier WM, Skoog I, Schneider JA, Pantoni L, Mok V, Chen CL, et al. Vascular cognitive impairment. Nature Reviews Disease Primers 2018;15:18003. [DOI] [PubMed] [Google Scholar]

Veroniki 2013

Veroniki AA, Vasiliadis HS, Higgins JP, Salanti G. Evaluation of inconsistency in networks of interventions. International Journal of Epidemiology 2013;42:332-45. [DOI] [PMC free article] [PubMed] [Google Scholar]

WHO 1992

World Health Organization. The ICD-11 Classification of Mental and Behavioural Disorders: Clinical Description and Diagnostic Guidelines. 10th edition. Geneva: World Health Organization, 1992. [Google Scholar]

Wortmann 2012

Wortmann M. Dementia: a global health priority - highlights from an ADI and World Health Organization report. Alzheimer's Research & Therapy 2012;4:40. [DOI: 10.1186/alzrt143] [DOI] [PMC free article] [PubMed] [Google Scholar]

Wu 2016

Wu YT, Fratiglioni L, Matthews FE, Lobo A, Breteler MM, Skoog I, et al. Dementia in western Europe: epidemiological evidence and implications for policy making. Lancet Neurology 2016;15:116-24. [DOI: 10.1016/S1474-4422(15)00092-7] [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0001] Auchus A. A randomized, 26-week, double-blind, placebo-controlled trial to evaluate the safety and efficacy of galantamine in the treatment of dementia secondary to cerebrovascular disease. IFPMA register 2004;1:1-3.

[CD013306-bib-0002] Auchus AP, Brashear HR, Salloway S, Korczyn AD, De Deyn PP, Gassmann-Mayer C, for the GAL-INT-26 Study Group. Galantamine treatment of vascular dementia: a randomised trial. Neurology 2007;69:448-58. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0003] NCT00035191. A placebo controlled trial to evaluate the safety and efficacy of galantamine in the treatment of vascular dementia [A randomized 26 week double blind placebo controlled trial to evaluate the safety and efficacy of galantamine in the treatment of vascular dementia]. https://clinicaltrials.gov/ct2/show/NCT00035191 First received: 03 May 2002.

[CD013306-bib-0004] Ballard C, Sauter M, Scheltens P, He Y, Barkhof F, Van Straaten ECW, et al. Efficacy, safety and tolerability of rivastigmine capsules in patients with probable vascular dementia: the Vantage study. Current Medical Research and Opinion 2008;24(9):2561-74. [DOI: 10.1185/03007990802328142] [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0005] Black S, Roman GC, Geldmacker DS, Salloway S, Hecker J, Burns A, et al, the Donepezil 307 Vascular Dementia Study Group. Efficacy and tolerability of Donepezil in vascular dementia: positive results of a 24-week, multi-centre, international, randomised, placebo-controlled trial. Stroke 2003;34:2323-32. [DOI: 10.1161/01.STR.0000091396.95360.E1] [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0006] Doody R, Pratt RD, Posner H, Kumar D. Vascular dementia patients who receive Donepezil treatment for 12 to 18 months maintain cognitive benefits. Neurobiology of Aging 2004;25:469. [Google Scholar]

[CD013306-bib-0007] Geldmacher DS, Pratt PD, Perdomo CA. Donepezil slows functional deterioration in patients with vascular dementia. European Journal of Neurology 2002;9:35. [Google Scholar]

[CD013306-bib-0008] Pratt D. Patient populations in clinical trials of the efficacy and tolerability of donepezil in patients with vascular dementia. Journal of the Neurological Sciences 2002;203:57-65. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0009] Pratt R, Perdomo C, The Donepezil 307 and 308 Study Groups. Donepezil is well tolerated in patients with vascular dementia: a comparison of safety and tolerability results from randomized, placebo-controlled clinical trials in vascular dementia patients and Alzheimer's disease patients. Neurobiology of Aging 2002;23(1 Suppl S57):S57. [Google Scholar]

[CD013306-bib-0010] Pratt RD, Perdomo CA, The Donepezil 307 VaD Study Group. Population characteristics and pattern of cognitive decline in patients with vascular dementia enrolled in two 24-week, randomized, double-blind, placebo-controlled trials. In: 7th International Geneva/Springfield Symposium on Advances in Alzheimer therapy; 2002 Apr 03-06; Geneva. 2002.

[CD013306-bib-0011] Pratt RD, Perdomo CA, The Donepezil VaD 307 and 308 Study Groups. Donepezil-treated patients with probable vascular dementia demonstrate cognitive benefits. Annals of the New York Academy of Sciences 2002;977:513-22. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0012] Roman G, Black S, Royall DR, Surick I, Kumar D, Schindler R, et al. Efficacy and safety of donepezil in vascular dementia: results from the largest double-blind trial in vascular dementia patients. European Neuropsychopharmacology 2006;16:S481-2. [Google Scholar]

[CD013306-bib-0013] Roman GC, Pratt RD, Perdomo CA, The Donepezil Study Group. Donepezil improves cognition in patients with vascular dementia: results from Study 307, a 24-week randomized, double-blind, placebo-controlled trial. In: 8th International Conference on Alzheimer's Disease and Related Disorders; 2002 Jul 20-25, Stockholm. 2002.

[CD013306-bib-0014] Schindler R, Perdomo CA, Pratt RD. Donepezil provides significant benefits to patients with vascular dementia in their ability to perform everyday activities. In: 8th Congress of the European Federation of the Neurological Sciences; 2004 Sep 04-07; Paris. 2004.

[CD013306-bib-0015] Seltzer B, Perdomo CA, Pratt RD, Schindler R. Donepezil provides significant benefits to patients with vascular dementia in their ability to perform everyday activities. Neurobiology of Aging 2004;25:470. [Google Scholar]

[CD013306-bib-0016] Bullock R, Erkinjuntti T, Lilienfeld S, GAL-INT-6 Study Group. Management of patients with Alzheimer's disease plus cerebrovascular disease: 12-month treatment with galantamine. Dementia and Geriatric Cognitive Disorders 2004;17:29-34. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0017] Burke W, Lilienfeld S. Galantamine improves behaviour and relieves caregiver distress in Alzheimer's disease (AD), vascular dementia and AD with cerebrovascular disease. In: 8th international conference on Alzheimer's disease and related disorders; 2002 Jul 20-25; Stockholm. 2002.

[CD013306-bib-0018] Erkinjuntti T, Gauthier S, Bullock R, Kurz A, Hammond G, Schwalen S, et al. Galantamine treatment in Alzheimer's disease with cerebrovascular disease: responder analyses from a randomized, controlled trial (GAL-INT-6). Journal of Psychopharmacology 2008;22(7):761-8. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0019] Erkinjuntti T, Kurz A, Gauthier A, Bullock R, Lillienfeld S, Damaraju CV. Efficacy of Galantamine in probable vascular dementia and Alzheimer's disease combined with cardiovascular disease: a randomised trial. Lancet 2002;359:1283-90. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0020] Erkinjuntti T, Kurz A, Small GW, Bullock R, Lilienfeld S, Damaraju CV, GAL-INT-6 Study Group. An open-label extension trial of galantamine in patients with probable vascular dementia and mixed dementia. Clinical Therapeutics 2003;25:1765-82. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0021] Erkinjuntti T, Lilienfeld S, Damaraju CV. Long-term treatment with galantamine is effective in slowing cognitive decline in patients with probable vascular dementia. In: 7th International Geneva/Springfield Symposium on Advances in Alzheimer Therapy; 2002 Apr 03-06; Geneva. 2002.

[CD013306-bib-0022] Erkinjuntti T. Broad therapeutic benefits in patients with probable vascular dementia or Alzheimer's disease with cerebrovascular disease after treatment with galantamine. European Journal of Neurology 2002;9:545. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0023] Kertesz A. Efficacy of galantamine in probable vascular dementia and Alzheimer's disease combined with cerebrovascular disease: a randomized trial. Current Neurology and Neuroscience Reports 2002;2(6):503-4. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0024] Kertesz A. Galantamine in vascular dementia and Alzheimer's disease combined with cerebrovascular disease. Current Neurology and Neuroscience Reports 2002;2(6):503-4. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0025] Kurz A, Erkinjuntti T, Bullock R, Lilienfeld S, Brashear HR. Long-term safety, tolerability, and efficacy of galantamine in the treatment of probable vascular dementia or Alzheimer's disease with cerebrovascular disease: an interim analysis. International Psychogeriatrics 2003;11:273. [Google Scholar]

[CD013306-bib-0026] Kurz A, Lilienfeld S, Damaraju CV. Galantamine is safe and effective for the long-term treatment of cognitive decline in patients with Alzheimer disease with cerebrovascular components or probable vascular dementia. In: 7th International Geneva/Springfield Symposium on Advances in Alzheimer therapy; 2002 Apr 03-06; Geneva. 2002.

[CD013306-bib-0027] Kurz A. Non-cognitive benefits of galantamine (Reminyl(R)) treatment in vascular dementia. Acta Neurologica Scandinavica 2002;106:19-24. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0028] Kurz AF, Erkinjuntti T, Small GW, Lilienfeld S, Venkata Damaraju CR. Long-term safety and cognitive effects of galantamine in the treatment of probable vascular dementia or Alzheimer's disease with cerebrovascular disease. European Journal of Neurology 2003;10:633-40. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0029] Small G, Erkinjuntti T, Kurz A, Lilienfeld S. Galantamine in the treatment of cognitive decline in patients with vascular dementia or Alzheimer's disease with cerebrovascular disease. CNS Drugs 2003;17:905-14. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0030] Mok V, Wong A, Ho S, Leung T, Lam WWM, Wong KS. Rivastigmine in Chinese patients with subcortical vascular dementia. Neuropsychiatric Disease and Treatment 2007;3(6):943-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013306-bib-0031] Narasimhalu K, Effendy S, Sim CH, Lee JM, Chen I, Hia SB, et al. A randomised controlled trial of Rivastigmine in patients with cognitive impairment no dementia because of cerebrovascular disease. Acta Neurologica Scandinavica 2010;121:217-24. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0032] Pratt RD, Perdomo CA. Donepezil improves cognitive function in patients with vascular dementia: results from study 307, a 24-week, randomized, double-blind, placebo-controlled trial. Portal of Geriatrics Online Education 2002;7:A161-2.

[CD013306-bib-0033] Roman GC, Salloway S, Black S, Royall DR, DeCarli C, Weiner MW, et al. Randomised, placebo-controlled, clinical trial of Donepezil in vascular dementia. Stroke 2010;41(6):1213-21. [DOI: 10.1161/STROKEAHA.109.570077] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013306-bib-0034] Aguilar M, Roman G, Black S, Royall D, Surick I, Kumar D, et al. Efficacy and safety of donepezil in vascular dementia results from largest double-blind trial in vascular dementia patients. In: 10th International Conference on Alzheimer's Disease and Related Disorders; 2006 Jul 15-20; Madrid. 2006.

[CD013306-bib-0035] Bayer A, Pratt RD, Kumar D. A comparison of the cognitive benefits of donepezil in patients with cortical and subcortical vascular dementia. In: 8th Congress of the European Federation of the Neurological Sciences; 2004 Sep 04-07; Paris. 2004.

[CD013306-bib-0036] Boundy K, Pratt R. Donepezil provides significant benefits in patients with vascular dementia. Internal Medicine Journal 2003;33:A46. [Google Scholar]

[CD013306-bib-0037] Farlow M. Efficacy of donepezil in vascular dementia. Neurology 2003;61(4):429. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0038] Pratt R, Perdomo C. Donepezil is well tolerated in patients with vascular dementia: a comparison of safety and tolerability results from randomized, placebo-controlled clinical trials in vascular dementia patients and Alzheimer's disease patients. Neurobiology of Aging 2002;1:S57. [Google Scholar]

[CD013306-bib-0039] Pratt RD, Perdomo CA, The Donepezil 308 VaD Study Group. Cognitive and global benefits of donepezil in vascular dementia: results from study 308, a 24-week, randomized, double-blind, placebo-controlled trial. In: 7th International Geneva/Springfield Symposium on Advances in Alzheimer therapy; 2002 Apr 03-06; Geneva. 2002.

[CD013306-bib-0040] Roman GC, Wilkinson DG, Doody RS, Black SE, Salloway SP, Schindler RJ. Donepezil in vascular dementia: combined analysis of two large-scale clinical trials. Dementia and Geriatric Cognitive Disorders 2005;20(6):338-44. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0041] Salloway S, Pratt RD, Perdomo CA. Donepezil-treated patients with vascular dementia demonstrate cognitive and global benefits: results from study 308, a 24-week, randomized, double-blind, placebo-controlled trial. Neurobiology of Aging 2002;1:57. [Google Scholar]

[CD013306-bib-0042] Salloway SP, Pratt RD, Perdomo CA. Donepezil is well tolerated in patients with vascular dementia: a comparison of tolerability in vascular dementia patients and Alzheimer's disease patients. European Journal of Neurology 2002;9:165. [Google Scholar]

[CD013306-bib-0043] Wilkinson D, Doody R, Helme R, Taubman K, Mintzer J, Kertesz A, et al, The Donepezil 308 Study Group. Donepezil in vascular dementia: a randomised, placebo-controlled study. Neurology 2003;61:479-86. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0044] Wilkinson D, Roman G, Salloway S, Hecker J, Boundy K, Kumar D, Posner H, Schindler R. The long-term efficacy and tolerability of donepezil in patients with vascular dementia. Int J Geriat Psychiatry 2010;25:305-315. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0045] Dichgans M, Markus HS, Salloway S, Verkkoniemi A, Moline M, Wang Q, et al. Donepezil in patients with subcortical vascular cognitive impairment: a randomised double-blind trial in CADASIL. Lancet Neurology 2008;7:310-8. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0046] American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 5th edition. Washington, DC: American Psychiatric Society, 2013. [Google Scholar]

[CD013306-bib-0047] Birks J. Cholinesterase inhibitors for Alzheimer's disease. Cochrane Database of Systematic Reviews 2006, Issue 1. Art. No: CD005593. [DOI: 10.1002/14651858.CD005593] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013306-bib-0048] Birks J, McGuinness B, Craig D. Rivastigmine for vascular cognitive impairment. Cochrane Database of Systematic Reviews 2013, Issue 5. Art. No: CD004744. [DOI: 10.1002/14651858.CD004744.pub3] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013306-bib-0049] Chen Y, Zhang J, Wang Y, Yuan J, Hu W. Efficacy of cholinesterase inhibitors in vascular dementia: an updated meta-analysis. European Neurology 2016;75:132-41. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0050] Colović MB, Krstić DZ, Lazarević-Pašti TD, Bondžić AM, Vasić VM. Acetylcholinesterase inhibitors: pharmacology and toxicology. Current Neuropharmacy 2013;11(3):315-35. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013306-bib-0051] Craig D, Birks J. Galantamine for vascular cognitive impairment. Cochrane Database of Systematic Reviews 2006, Issue 1. Art. No: CD004746. [DOI: 10.1002/14651858.CD004746.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013306-bib-0052] Dawbarn D, Shelly JA. Neurobiology of Alzheimer’s Disease. 2nd edition. Oxford (UK): Oxford University Press, 2001. [Google Scholar]

[CD013306-bib-0053] Dichgans M, Leys D. Vascular cognitive impairment. Circulation Research 2017;120:573-91. [DOI: 10.1161/CIRCRESAHA.116.308426] [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0054] Erkinjuntti T, Román G, Gauthier S, Feldman H, Rockwood K. Emerging therapies for vascular dementia and vascular cognitive impairment. Stroke 2004;35:1010-7. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0055] McMaster University (developed by Evidence Prime) GRADEpro GDT. Version accessed 12 July 2019. Hamilton (ON): McMaster University (developed by Evidence Prime), 2008. Available at gradepro.org.

[CD013306-bib-0056] Higgins JP, Altman DG, Sterne JA. Chapter 8: Assessing risk of bias in included studies. In: Higgins JP, Churchill R, Chandler J, Cumpston MS, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 5.2.0 (updated June 2017). The Cochrane Collaboration, 2017. Available from training.cochrane.org/handbook/archive/v5.2.

[CD013306-bib-0057] Hort J, O’Brien JT, Gainotti G, Pirttila T, Popescu BO, Rektorova I, et al. EFNS guidelines for the diagnosis and management of Alzheimer’s disease. European Journal of Neurology 2010;17:1236-48. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0058] Jackson D, Barrett JK, Rice S, White IR, Higgins JP. A design-by-treatment interaction model for network meta-analysis with random inconsistency effects. Statistics in Medicine 2014;33:3639-54. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013306-bib-0059] Kavirajan H, Schneider LS. Efficacy and adverse effects of cholinesterase inhibitors and memantine in vascular dementia: a meta-analysis of randomised controlled trials. Lancet Neurology 2007;6:782-92. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0060] Kibret T, Richer D, Beyene J. Bias in identification of the best treatment in a Bayesian network meta-analysis for binary outcome: a simulation study. Clinical Epidemiology 2014;6:451-60. [DOI: 10.2147/CLEP.S69660] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013306-bib-0061] Li Y, Hai S, Zhou Y, Dong BR. Cholinesterase inhibitors for rarer dementias associated with neurological conditions. Cochrane Database of Systematic Reviews 2015, Issue 3. Art. No: CD009444. [DOI: 10.1002/14651858.CD009444.pub3] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013306-bib-0062] Lilienfeld S. Galantamine: a novel cholinergic drug with a unique dual mode of action for the treatment of patients with Alzheimer’s disease. CNS Drug Reviews 2002;8:159-76. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013306-bib-0063] Malouf R, Birks J. Donepezil for vascular cognitive impairment. Cochrane Database of Systematic Reviews 2004, Issue 1. Art. No: CD004395. [DOI: 10.1002/14651858.CD004395.pub2] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013306-bib-0064] Manero RM, Casals-Coll M, Sánchez-Benavides G, Rodríguez-de los Reyes ON, Aguilar M, Badenes D, et al, for the NEURONORMA Study Team. Diagnostic validity of the Alzheimer’s Disease Functional Assessment and Change Scale in mild cognitive impairment and mild to moderate Alzheimer’s disease. Dementia and Geriatric Cognitive Disorders 2013;37(5):366-75. [DOI: 10.1159/000350800] [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0065] Marshall IJ, Noel-Storr AH, Kuiper J, Thomas J, Wallace BC. Machine Learning for Identifying Randomized Controlled Trials: an evaluation and practitioner’s guide. Research Synthesis Methods 2018;9(4):602-14. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013306-bib-0066] McDonald S, Noel-Storr AH, Thomas J. Harnessing the efficiencies of machine learning and Cochrane Crowd to identify randomised trials for individual Cochrane reviews. In: Global Evidence Summit; 2017 Sep 13-16; Cape Town. 2017.

[CD013306-bib-0067] Mills EJ, Thorlund K, Ioannidis JPA. Demystifying trial networks and network meta-analysis. BMJ 2013;346:f2914. [DOI: 10.1136/bmj.f2914] [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0068] National Institute for Health and Care Excellence. NICE guidelines: donepezil, galantamine, rivastigmine and memantine for the treatment of Alzheimer's disease. Technology appraisal guidance [TA217]. www.nice.org.uk/guidance/ta217 (accessed 6 January 2019).

[CD013306-bib-0069] Noel-Storr AH, Project Transform Team. Cochrane Crowd: new ways of working together to produce health evidence. In: Evidence Live; 2018 Jun 18-20; Oxford. 2018.

[CD013306-bib-0070] O’Brien JT, Erkinjuntti T, Reisberg B, Roman G, Sawada T, Pantoni L, et al. Vascular cognitive impairment. Lancet Neurology 2003;2:89-98. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0071] Owen RK, Bradbury N, Xin Y, Copper N, Sutton A. MetaInsight: an interactive web‐based tool for analyzing, interrogating, and visualizing network meta‐analyses using R‐shiny and netmeta. Research Synthesis Methods 2019;10(4):569-81. [DOI: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013306-bib-0072] Perry E, Kay DW. Some developments in brain ageing and dementia. British Journal of Biomedical Science 1997;54:201-15. [PubMed] [Google Scholar]

[CD013306-bib-0073] Puhan MA, Schünemann HJ, Murad MH, Li T, Brignardello-Petersen R, Singh JA, et al GRADE Working Group. A GRADE Working Group approach for rating the quality of treatment effect estimates from network meta-analysis. BMJ 2014;349:g5630. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0074] Qiu C, Kivipelto M, Strauss E. Epidemiology of Alzheimer's disease: occurrence, determinants, and strategies toward intervention. Dialogues in Clinical Neuroscience 2009;11:111-28. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013306-bib-0075] Nordic Cochrane Centre, The Cochrane Collaboration Review Manager 5 (RevMan 5). Version 5.3. Copenhagen: Nordic Cochrane Centre, The Cochrane Collaboration, 2014.

[CD013306-bib-0076] Ritter A, Pillai JA. Treatment of vascular cognitive impairment. Current Treatment Options in Neurology 2015;17:367. [DOI: 10.1007/s11940-015-0367-0] [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0077] Roman G, Tatemichi T, Erkinjuntti T, Cummings JL, Masdeu JC, Garcia JH, et al. Vascular dementia: diagnostic criteria for research studies. Report of the NINDS-AIREN International Workshop. Neurology 1993;43:250-60. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0078] Rücker G, Schwarzer G. Ranking treatments in frequentist network meta-analysis works without resampling methods. BMC Medical Research Methodology 2015;15:58. [DOI: 10.1186/s12874-015-0060-8] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013306-bib-0079] Salanti G, Ades AE, Ioannidis JP. Graphical methods and numerical summaries for presenting results from multiple-treatment meta-analysis: an overview and tutorial. Journal of Clinical Epidemiology 2011;64:163-71. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0080] Salanti G. Indirect and mixed-treatment comparison, network, or multiple-treatments meta-analysis: many names, many benefits, many concerns for the next generation evidence synthesis tool. Research Synthesis Methods 2012;3:80-97. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0081] Schrag A, Schott JM, Alzheimer's Disease Neuroimaging Initiative. What is the clinically relevant change on the ADAS-Cog? Journal of Neurology, Neurosurgery, and Psychiatry 2012;83(2):171-3. [DOI: 10.1136/jnnp-2011-300881] [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0082] Schünemann HJ, Oxman AD, Higgins JP, Vist GE, Glasziou P, Guyatt GH. Chapter 11: Presenting results and ‘Summary of findings’ tables. In: Higgins JP, Green S, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. Available from training.cochrane.org/handbook/archive/v5.1/.

[CD013306-bib-0083] Skrobot OA, O’Brien J, Black S, Chen C, DeCarli C, Erkinjuntti T, et al. The Vascular Impairment of Cognition Classification Consensus Study. Alzheimer’s & Dementia 2017;13:624-33. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0084] Thomas J, Noel-Storr AH, Marshall I, Wallace B, McDonald S, Mavergames C, et al. Living Systematic Review Network. Living Systematic Reviews: 2. Combining Human and Machine Effort. Journal of Clinical Epidemiology 2017;91:31-7. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0085] Toghi H, Abe T, Kimura M, Saheki M, Takahashi S. Cerebrospinal fluid acetylcholine and choline in vascular dementia of Binswanger and multiple small infarcts types as compared with Alzheimer-type dementia. Journal of Neural Transmission 1996;103:1211-20. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0086] Turner RM, Davey J, Clarke MJ, Thompson SG, Higgins JP. Predicting the extent of heterogeneity in meta-analysis, using empirical data from the Cochrane Database of Systematic Reviews. International Journal of Epidemiology 2012;41:818-27. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013306-bib-0087] Flier WM, Skoog I, Schneider JA, Pantoni L, Mok V, Chen CL, et al. Vascular cognitive impairment. Nature Reviews Disease Primers 2018;15:18003. [DOI] [PubMed] [Google Scholar]

[CD013306-bib-0088] Veroniki AA, Vasiliadis HS, Higgins JP, Salanti G. Evaluation of inconsistency in networks of interventions. International Journal of Epidemiology 2013;42:332-45. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013306-bib-0089] World Health Organization. The ICD-11 Classification of Mental and Behavioural Disorders: Clinical Description and Diagnostic Guidelines. 10th edition. Geneva: World Health Organization, 1992. [Google Scholar]

[CD013306-bib-0090] Wortmann M. Dementia: a global health priority - highlights from an ADI and World Health Organization report. Alzheimer's Research & Therapy 2012;4:40. [DOI: 10.1186/alzrt143] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD013306-bib-0091] Wu YT, Fratiglioni L, Matthews FE, Lobo A, Breteler MM, Skoog I, et al. Dementia in western Europe: epidemiological evidence and implications for policy making. Lancet Neurology 2016;15:116-24. [DOI: 10.1016/S1474-4422(15)00092-7] [DOI] [PubMed] [Google Scholar]

PERMALINK

Cholinesterase inhibitors for vascular dementia and other vascular cognitive impairments: a network meta‐analysis

Ceri E Battle

Azmil H Abdul-Rahim

Susan D Shenkin

Jonathan Hewitt

Terry J Quinn

Abstract

Background

Objectives

Search methods

Selection criteria

Data collection and analysis

Main results

Authors' conclusions

Plain language summary

Summary of findings

Background

Description of the condition

Description of the intervention

How the intervention might work

Why it is important to do this review

Objectives

Methods

Criteria for considering studies for this review

Types of studies

Types of participants

Types of interventions

Types of outcome measures

Primary outcomes

Secondary outcomes

NMA outcomes

Search methods for identification of studies

Electronic searches

Searching other resources

Data collection and analysis

Selection of studies

1.

Data extraction and management

Assessment of risk of bias in included studies

Measures of treatment effect

Unit of analysis issues

Dealing with missing data

Assessment of heterogeneity

Assessment of clinical and methodological heterogeneity within treatment comparisons

Assessment of transitivity across treatment comparisons

Assessment of reporting biases

Data synthesis

Methods for direct treatment comparisons

Methods for indirect and mixed comparisons

Subgroup analysis and investigation of heterogeneity

Assessment of statistical heterogeneity

Assumptions when estimating the heterogeneity

Measures and tests for heterogeneity

Assessment of statistical inconsistency

Local approaches for evaluating inconsistency

Global approaches for evaluating inconsistency

Investigation of heterogeneity and inconsistency

Sensitivity analysis

Summary of findings and assessment of the certainty of the evidence

Results

Description of studies

Results of the search

Included studies

Excluded studies

Risk of bias in included studies

2.

3.

Allocation

Donepezil

Rivastigmine

Galantamine

Blinding

Donepezil

Rivastigmine

Galantamine

Incomplete outcome data

Donepezil

Rivastigmine

Galantamine