Very early mobilization of patients with stroke: a systematic review
Systematic review
|Published
Patients with stroke should be mobilized early, but how early? We were asked to identify and review studies that have evaluated this.
Key message
In 2014, 9600 persons were admitted to hospital with a stroke in Norway. Mortality caused by stroke have been substantially reduced the latest decades, but many patients with stroke get lasting neurological disabilities. An important element in the treatment of patients with stroke is that the patient is mobilized early. However, it is unclear if it is most efficient if this happens within 24 hours or between 24 and 48 hours after the stroke onset. We were asked to identify and review studies that have evaluated this.
- We identified three studies that evaluated this issue, one international study from several countries with 2104 patients, a study from Australia with 71 pasients, and a Norwegian study with 56 patients.
- In both the international study and the study from Australia, the patients that were mobilised within 24 hours of the stroke onset got more frequent and longer mobilisation compared to patients that were mobilised after 24-48 hours. Therefore we cannot isolate the effect of the time of mobilisation.The Norwegian study comprised few patients.
- Research indicate that there probably is more deaths and a higher proportion with reduced function if the patients are mobilised very early (within 24 hours) with high frequency and intensity compared to early (within 48 hours) with lower frequency and intensity.
We are still uncertain if the time for starting mobilisation of patients with stroke should be within 24 hours or between 24 and 48 hours of the stroke onset.
Summary
Background
In 2014, 9600 persons were admitted to hospital with a stroke in Norway. Mortality caused by stroke have been substantially reduced during the latest decades, but many patients with stroke get lasting neurological disabilities.
Method
We updated the search for literature performed in the Cochrane-review by Bernhardt et al from April 2008. Our literature searches were performed on 7th and 20th October 2016 in MEDLINE, EMBASE and Central.
Two project members have independently from each other evaluated identified titles and abstracts according to the inclusion criteria. Selected references was then ordered in full text and evaluated according to the inclusion criteria. We evaluated the risk of bias as described in the Cochrane review and our handbook.
The inclusion criteria were:
- Population: persons who had had a stroke
- Intervention: very early mobilisation, ie within 24 hours of stroke onset
- Comparison: early mobilisation, ie from 24 to 48 hours after stroke onset
- Outcome: mortality, dependency measured with modified Rankin Scale (mRS) and adverse events.
We summarised relevant data from the randomised controlled studies in text and tables. To do the meta analysis, we used Review Manager Software and random effect model. Dichotomous outcomes are presented as risk ratio (RR) with 95 % confidence interval.
The quality of the documentation for each of the outcomes was evaluated by GRADE (Grading of Recommendations Assessment, Development and Evaluation). GRADE is a tool for evaluating our confidence in the effect estimate.
Results
We identified three studies, an international study with 2104 patients, a study from Australia with 71 patients, and a Norwegian study with 56 patients.
In the group that was mobilised very early 103 of 1117 patients died, while in the group that was mobilised early 77 of 1111 patients died (RR=1.66, 95 % CI = 0.87-3.17). The number of patients dependent (mRS 3-6) were 596 of 1101 in the very early mobilisation group and 554 of 1106 in the early mobilisation group (RR=1.07, 95 % CI 0.88-1.29). The number of non-fatal serious adverse events after three months were 216 of 1092 in the very early mobilisation group and 222 of 1083 in the early mobilisation group (RR = 0.96, 95 % CI 0.81 – 1.13).
In the very early mobilisation group, the patients were mobilised more frequently and more intensely, and the mobilisation started after a mean of 18.5 hours after the stroke onset. In the early mobilisation group, the mobilisation started after a mean of 22.4 hours after the stroke onset, even though this was not according to protocol, and the patients were mobilised less frequently and less intensely.
We are still uncertain if an intervention that only addresses the time for start of mobilisation for patient with stroke should be within 24 hours of between 24 and 48 hours of the stroke onset.
Based on available documentation we assume that it is possible that a combination of very early mobilisation within 24 hours in addition to more frequent and intense mobilisation can result in a higher number of deaths and reduced function measured with mRS.
Discussion
When the time for start of mobilisation after the stroke onset is the issue, there is a reason to notice that the difference between the times in the two groups of the Norwegian study is 20 hours, while it in the larger international study is four hours. Possibly, differences as large as this, are important.
We have repeated the literature search from a Cochrane review from 2009, where only randomised controlled trials were included. Studies with other controlled designs could possibly have contributed with more information. We searched in three databases, and it is possible that there are more relevant studies in other databases.
Because the time for mobilisation will be depending on the time between the stroke onset and the admittance to hospital, the possibility to include patients quickly enough after the stroke onset will also effect who participates in the studies that are included in our review. The organisation of ambulance services and other prehospital services might influence which patients that can participate in the studies.
A strength with our systematic approach is that we have documented our work and our arguments. As two persons have worked independently of each other, we have reduced the risk of mistakes in evaluating titles, abstracts and articles in full text.
Conclusion
Research indicate that probably there is a danger for more deaths and a higher proportion with reduced function if the patients are mobilised very early (within 24 hours) with high frequency and intensity compared to early (within 48 hours) with lower frequency and intensity.
We are still uncertain if the time for start of mobilisation of patients with stroke should be within 24 hours or between 24 and 48 hours of the stroke onset.
