Very early versus delayed mobilisation after stroke
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
To conduct a systematic review of randomised controlled trials (RCTs) to establish if very early mobilisation (started as soon as possible and no later than 48 hours after onset of symptoms) compared with conventional care in patients with stroke can increase the proportion of independent survivors.
Background
Stroke presents a major global public health challenge, with 5.5 million people dying from stroke each year (WHO 2003) and many more living with chronic disability (Wolfe 2000). We know that treatment in a stroke unit (compared with treatment in a general medical ward) reduces the odds of being dead or disabled at 12 months post stroke (SUTC 2001). However, relatively little is known about which components of acute stroke unit care may be responsible for better outcomes (Langhorne 1998; Langhorne 2002). Early rehabilitation interventions, including very early mobilisation, are characteristic features of stroke unit care and have been recommended in a number of acute stroke clinical guidelines (Adams 2003; NSF 2003). It is not known whether very early mobilisation improves outcome after stroke.
Objectives
To conduct a systematic review of randomised controlled trials (RCTs) to establish if very early mobilisation (started as soon as possible and no later than 48 hours after onset of symptoms) compared with conventional care in patients with stroke can increase the proportion of independent survivors.
Methods
Criteria for considering studies for this review
Types of studies
We aim to consider all unconfounded randomised trials, with or without blinding, of very early mobilisation within 48 hours of symptom onset compared with conventional care (that is, normal practice or no routine intervention). We will include only the first period of any trial employing a randomised crossover design.
Types of participants
We will include patients of any age with a definite clinical diagnosis of stroke (focal neurological deficit caused by cerebrovascular disease) for whom mobilisation could be started within 48 hours of the onset of symptoms.
Types of interventions
For the purpose of this review, very early mobilisation is defined as any intervention aimed to reduce the time to first mobilisation (first out‐of‐bed episode) and to improve the frequency or amount, or both, of out‐of‐bed physical activity (for example, participation in activities of daily living involving mobilising, transfers, or both, such as walking to the toilet, transferring on and off the toilet with the assistance of one or more staff or equipment, sitting out of bed, standing, and walking). Conventional care is defined as usual mobilisation practice.
Any form of very early mobilisation will be considered irrespective of the member of staff assisting, duration of the mobilising intervention, or amount of intervention.
Types of outcome measures
Primary outcome of interest
(1) Death or a poor outcome: the number of patients who died or remained dependent (a Barthel score of less than 15 or equivalent, or admission to institutional care, or both, or a modified Rankin Score of 3 or more) at the end of scheduled follow up where follow up was performed at three months or longer after the stroke. Institutional care is defined as care within a residential home, nursing home, or hospital at the end of scheduled follow up or at discharge.
Secondary outcomes of interest
(1) Death: number of deaths from any cause at the end of the follow up.
(2) Death or dependence: the number of patients dead or physically dependent at the end of scheduled follow up where follow up was performed at three months or longer after the stroke.
(3) The number of patients requiring institutional care at the end of scheduled follow up where follow up was performed at three months or longer after the stroke.
(4) Performance in activities of daily living at the end of scheduled follow up where follow up was performed at three months or longer after the stroke.
(5) Performance in extended activities of daily living (community and domestic activities) at the end of scheduled follow up where follow up was performed at three months or longer after the stroke.
(6) Patient subjective health status or quality of life at the end of scheduled follow up where follow up was performed at three months or longer after the stroke.
(7) Time to walking unassisted (without help from another person) reported alone or as part of a functional mobility scale, at the end of scheduled follow up where follow up was performed at three months or longer after the stroke.
(8) Adverse events: number or severity (or both) of adverse effects including deep vein thrombosis (DVT), non‐fatal pulmonary embolism (PE), incidence and grade of new pressure sores (using standardised grading scale), number of incontinent episodes over 24 hours, severity of incontinence, chest infection, falls or injuries, and physiological variables (blood pressure, oxygen, temperature) recorded at the end of scheduled follow up.
(9) Patient mood at the end of scheduled follow up.
Search methods for identification of studies
See: 'Specialized register' section in Cochrane Stroke Group
We will search the Cochrane Stroke Group Trials Register. In addition, we will carry out searches of the following electronic bibliographic databases: the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library, latest issue); MEDLINE (1966 to 2006); EMBASE (1980 to 2006); CINAHL (1983 to 2006); PsycLIT (1974 to 2006), the Occupational Therapy Research Index and Dissertation Abstracts, AMED (1985 to 2006), PEDro (to 2006), REHABDATA database (to 2006), ongoing trials and research registers, as well as contact with researchers in the field. We will also search the three citation index databases: Science Citation Index (SCI); Social Sciences Citation Index (SSCI); and Arts and Humanities Citation Index (A&HCI) (Appendix 1).
We will also handsearch all available years of the following journals.
-
Advances in Occupational Medicine and Rehabilitation (1996 to 1999)
-
Advances in Clinical Neurosciences and Rehabilitation (2001 to 2006)
-
Advances in Clinical Rehabilitation (1987 to 1990)
-
Archives of Occupational Therapy (renamed Occupational Therapy and Rehabilitation) (1922 to 2006)
-
Canadian Journal of Rehabilitation (1987 to 1999)
-
Chinese Journal of Physical Medicine and Rehabilitation (1980 to 2006)
-
European Journal of Physical Medicine and Rehabilitation (1991 to 1999)
-
International Journal of Rehabilitation and Health (1995 to 2000)
-
Journal of Rehabilitation Administration (1987 to 2006)
-
Rehabilitation (1948 to 49, 1951 to 1977)
-
Rehabilitation in Canada (1963 to 1972)
-
Rehabilitation Nursing Research (1992 to 1996)
-
Topics in Stroke Rehabilitation (1995 to 2006)
We will search reference lists of all relevant papers.
Data collection and analysis
We will select trials for inclusion in the review from all studies found by the methods outlined in the previous section. One review author will read the titles of all the references identified and eliminate any obviously irrelevant studies. The abstracts for the remaining studies will be obtained and, based on the inclusion criteria (types of studies, types of participants, aims of interventions, outcome measures), two review authors will independently classify these as 'relevant', 'irrelevant' or 'unsure'. Any trials classified as 'irrelevant' by the review authors will be excluded. Differences in opinion regarding trial eligibility will be resolved by discussion. All remaining trials will be included at this stage.
Assessment of methodological quality
We will extract information, for each included trial, about the method of randomisation and allocation concealment, blinding of outcome assessment and whether all the randomised patients were accounted for in the analysis. The authors of the trials will be contacted if the above information is not available in the published reports. Sensitivity analyses will be based on these variables.
Data extraction
Our primary aim will be to obtain standardised data through collaboration with the original trialists. Data taken from published sources will be independently extracted by two review authors using a standard data recording form. Concealment of randomisation, blinding in outcome evaluation, and intention‐to‐treat analysis will be evaluated and graded as present or unclear. Differences occurring between the two review authors will be resolved through discussion.
Data analysis
For binary outcome variables, a weighted estimate of the intervention effects across trials (odds ratio) will be calculated using a fixed‐effect model. For continuous outcome data, we will use mean difference. Standardised mean difference will be used where different outcome measures are used to record the same outcome. Ninety‐five per cent confidence intervals will be used for interpreting the results. Inconsistency across studies will be quantified using I‐squared (that is, by describing the percentage of the variability in the effect estimates that is due to heterogeneity rather than sampling error). A value greater than 50% will be considered substantial heterogeneity. A random‐effects model will be used if there is substantial heterogeneity.
Sensitivity analysis
Sensitivity analyses will be carried out based upon the method of randomisation and adequacy of allocation concealment, presence of an intention‐to‐treat analysis, blinding of final assessment, time to first mobilisation, and amount of mobilisation.
Missing data for patients excluded or lost to follow up will be analysed using a worst case scenario for the composite outcome of 'death or a poor outcome' to ensure significance of the results. In the worst case analysis, it will be assumed that those patients who were lost to follow up in the intervention group were dead, dependent, had deteriorated in ability to perform activities of daily living, or required nursing home care at the end of scheduled follow up. If the effects of primary and worst‐case analyses are in the same direction and size of effect, we can draw a definite conclusion about the effectiveness of early mobilisation.
