Methods
Eligibility Criteria
Randomised clinical trials involving adults with TBI, with evidence of raised ICP were examined. Eligible trials compared the effect on ICP of hypertonic sodium solutions and mannitol. Non-blinded and crossover design trials were included.
Information Sources and Search
See search terms used (see online supplementary appendix A http://emj.bmj.com/content/31/8/679/suppl/DC1) and flow diagram (Figure 1).
(Enlarge Image)
Figure 1.
Literature search strategy and results. (1) HDAS, NHS Healthcare Databases Advanced Search. (2) Sixty titles were reduced to 28 papers by abstract review. (3) SIGLE, System for Information on Grey Literature in Europe. ICP, intracranial pressure.
Study Selection
Fifty-five titles or abstracts were reviewed by ACR and eight original scientific pieces of work were included. The search and study selection were checked and confirmed by JS and AK independently.
Included
Randomised clinical trials involving adults with TBI, with evidence of raised ICP were included. Eligible trials compared the effect on ICP of hypertonic sodium solutions and mannitol. Unblinded and crossover design trials were included.
Excluded
Studies involving children, patients with hypovolaemic shock, non-TBI, studies where data were available only from an abstract, animal studies or incomplete data were excluded.
Data Collection Process
For each trial the following were extracted: number of patients; intervention performed (strength and volume of agents used and whether or not they were equiosmolar); ICP at baseline and after the study period (varied between 60 and 240 min), and the difference in mean ICP reduction. Where discrepancy in reporting existed, lead authors were contacted by ACR in an attempt to clarify the findings.
The primary outcome measure was the difference in mean ICP reduction between the hypertonic sodium solution and mannitol. The relative risk of success of treatment (as defined in each study) was also calculated for those studies that quoted a ‘threshold for successful ICP treatment’.
Risk of Bias in Individual Studies
The Cochrane collaboration's tool for assessing the risk of bias in randomised studies was used. Specifically this examined random sequence generation; allocation concealment; blinding of participants and personnel; blinding of outcome assessment; incomplete outcome data; selective reporting and other sources of bias (Figure 2).
(Enlarge Image)
Figure 2.
Risk of bias assessment table.
Summary Measures
The principal summary measure was the pooled difference in mean ICP reduction with hypertonic sodium substance compared with mannitol.
Meta-analysis
Heterogeneity among trials, arising from differences in hypertonic sodium formulations and varying osmolarities of formulations, as well as differences in subject populations, led to the use of a random-effects model. In the secondary analysis, the same strategy was used to calculate the pooled relative risk of ICP control. This analysis used DerSimonian and Laird's random-effects model. Heterogeneity among studies was assessed using Cochran's Q statistic and Higgins' I statistic. Significant heterogeneity was prespecified using p<0.05.
Risk of Bias Across Studies
Publication bias was assessed using a funnel plot.
Additional Analyses
Sensitivity analysis was performed on the primary outcome assessment by excluding the two papers that were assessed as using the weakest methodology, judged by a combination of Jadad score and risk of bias assessment.