Summary of findings for the main comparison. Hypertonic saline compared with other intracranial pressure‐lowering agents for acute traumatic brain injury.
| Hypertonic saline compared with other intracranial pressure‐lowering agents for acute traumatic brain injury | ||||||
|
Patient or population: people with acute traumatic brain injury Settings: intensive care units Intervention: hypertonic saline (between 3% and 7.5% solution) delivered by infusions Comparison: all other intracranial pressure‐lowering agents eligible (but mannitol infusions the sole comparator within the included trials) | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect (95% CI) | No of participants (RCTs) | Quality of the evidence (GRADE) | Comments | |
| Assumed risk | Corresponding risk | |||||
| Other intracranial pressure‐lowering agents (mannitol or mannitol with glycerol) | Hypertonic saline | |||||
| Mortality (short‐term) | 2 RCTs reported mortality in the short term, prior to discharge from hospital. Meta‐analysis was not possible. 1 RCT (n = 38) reported that 3/18 participants in the HTS group and 1/20 in the mannitol group died in the first 6 days following treatment, after which no further deaths occurred in the HTS group but a further 9 deaths occurred in the mannitol group prior to discharge. At this time point, there is a slight trend favouring HTS compared to mannitol (RR 0.33, 95% 0.11 to 1.02). The other RCT (n = 32, in which only a third of participants had TBI) reported that 7/17 (41.2%) participants in the HTS group and 9/15 (60%) in the mannitol group died by the end of stay in the ICU. Here, HTS did not reduce all‐cause mortality in people with acute TBI (RR 0.69, 95% CI 0.34 to 1.39). |
70 (2 RCTs) |
⊕⊝⊝⊝ Very lowa |
Not advisable to pool data due to difference in time of assessment and in populations | ||
| Mortality at 6 months | 35.6% risk (available case) 35.6% risk (worst case for intervention, HTS) 44.4% risk (best case for intervention, HTS) |
30% risk (available case) 40% risk (worst case for intervention, HTS) 30% risk (best case for intervention, HTS) |
RR 0.84 (0.46 to 1.55) (available case) RR 1.12 (0.66 to 1.89) (worst case for intervention, HTS) RR 0.67 (0.38 to 1.18) (best case for intervention, HTS) |
85 (2 RCTs) | ⊕⊝⊝⊝ Very lowb |
Pooling done with available, best‐case and worst‐case scenarios. In no case do results show a clear difference between groups. |
|
Glasgow Outcome Scale: poor outcome at 6 months |
66.7% risk (available case) 66.7% risk (worst case for intervention, HTS) 75.6% risk (best case for intervention, HTS) |
72.5% risk (available case) 82.5% risk (worst case for intervention, HTS) 72.5% risk (best case for intervention, HTS) |
RR 1.09 (0.82 to 1.44) (available case) RR 1.24 (0.97 to 1.58) (worst case for intervention, HTS) RR 0.96 (0.74 to 1.24) (best case for intervention, HTS) |
85 (2 RCTs) | ⊕⊝⊝⊝ Very lowb |
Pooling done with available, best‐case and worst‐case scenarios from two studies which supplied sufficient data. One study on children reported GOS non‐traditionally (defining 'poor outcome' as survival with any level of disability excluding persistent vegetative state and found no clear difference between groups |
| Uncontrolled ICP during treatment |
HTS vs mannitol 1 RCT reported a difference in the magnitude of ICP reduction and found both HTS and mannitol effectively and equally reduced ICP levels with subsequent elevation of CPP and CBF, although this effect was significantly stronger and of longer duration after HTS. Another 2 RCTs reported a difference in the ratio of uncontrolled ICP between the 2 groups, and the definition of uncontrolled ICP in these 2 RCTs differed, as did the time frames for data collection. 1 of these RCTs found both interventions to be effective, but added pretreatment factors need to be considered (e.g. serum sodium and haemodynamics). The 5th RCT found HTS to be more effective for increased ICP than mannitol but cautioned the benefit in this trial might be explained by "local osmotic effects". 2 RCTs in which intervention was only given after CSF drainage had failed, reported the mean fall in ICP following a dose averaged over hundreds of episodes across 4‐6 days of treatment. Of these 2 RCTs, 1 found HTS to be superior to mannitol; the other found no clear difference between groups. HTS vs mannitol vs mannitol plus glycerol A 6th RCT comparing HTS with mannitol and also with mannitol plus glycerol reported means and ranges of change within an hour following a single dose. They reported all 3 hyperosmolar agents (HTS, mannitol and mannitol plus glycerol) as effective, but HTS was slightly superior, effecting a greater change in reducing ICP, and more quickly, than other agents, while at a lower dose. |
287 (6 RCTs) |
⊕⊝⊝⊝ Very lowb |
It is not possible to pool these data because of variations in timings and ways of reporting ICP | ||
| A rebound phenomenon during treatment | None of the RCTs reported on this outcome systematically, although it is mentioned in passing as potentially affecting those treated with mannitol in 1 RCT. | 32 (1 RCT) |
⊕⊝⊝⊝ Very lowc |
This outcome was reported in just 1/4 RCTs available. | ||
| Pulmonary oedema during treatment | No data available | This outcome was not reported. | ||||
| Acute renal failure during treatment | No data available | This outcome was not reported. | ||||
| *The basis for the assumed risk (e.g. the median control group risk across trials) is provided in footnotes. The corresponding risk (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; CPP: cerebral perfusion pressure; CSF: cerebrospinal fluid; HTS: hypertonic saline; ICP: Intracranial pressure: ICU: intensive care unit; RCT: randomised controlled trial; RR: risk ratio; TBI: traumatic brain injury | ||||||
| GRADE Working Group grades of evidence High quality: further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: we are very uncertain about the estimate. | ||||||
aWe downgraded for imprecision (low number of participants), risk of bias and indirectness (1 trial included a mixed population of whom only a third had TBI). bWe downgraded for imprecision, inconsistency and risk of bias. cWe downgraded for imprecision, indirectness and risk of bias.