Best Evidence Topics
  • Send this BET as an Email
  • Make a Comment on this BET

Hypertonic sodium solutions vs mannitol in reducing ICP in traumatic brain injury

Three Part Question

In [patients with traumatic brain injury (TBI) and signs of raised intracranial pressure (ICP)] are [hypertonic sodium solutions better than mannitol] at [reducing morbidity and mortality]?

Clinical Scenario

A 54 year old female pedestrian has been hit by a bus. She is brought into the ED by ambulance. Her GCS is 13 on arrival and examination reveals an isolated head injury with a haematoma over the left occiput. CT confirms a right frontal contusion with subdural and subarachnoid haemorrhage and a fracture of the left temporal and occipital bones. There is midline shift to the left. On return to the ED, her right pupil appears dilated and her GCS is now 10 (E2M4V4). The neurosurgical registrar is in theatre for the next 20 minutes. You intubate and ventilate the patient and wonder whether hypertonic saline would be better than mannitol at controlling the patient's ICP acutely.

Search Strategy

OVID Medline (R) 1950—Oct week 3, 2010 and OVID Medline (R) In-process (29 Oct 2010).

Cochrane Library: hypertonic and saline and head and injury Medline ((brain and injury).mp. OR exp brain injuries/OR ‘head injury’.mp) AND (exp intracranial pressure/OR ‘raised intracranial pressure’.mp) AND ((exp saline solution, hypertonic/OR ‘hypertonic saline’.mp. OR exp sodium chloride/OR ‘sodium chloride’.mp.) AND (exp mannitol/OR OR ‘osmotic diuretic’.mp.)) LIMIT to human and English language.

Search Outcome

Medline: 22 papers were found of which four were relevant. Cochrane: one review yielded one further paper.

Relevant Paper(s)

Author, date and country Patient group Study type (level of evidence) Outcomes Key results Study Weaknesses
Oddo et al,
12 consecutive patients with severe traumatic brain injury (TBI) admitted to a level 1 trauma centre who underwent brain tissue oxygen tension (PbtO2) monitoring on neuroITU. All treated with mannitol (25% 0.75g/kg) if ICP>20mmHg or Hypertonic saline (HTS)(7.5% 250ml) if ICP not controlled with mannitolRetrospective cohort, level 4 evidenceIncrease in PbtO2 from baseline at 30, 60 & 120 mins42 episodes of raised ICP treated with mannitol (28 boluses) or HTS (second line)(14 boluses). Increase in PbtO2 from baseline (28.3mmHG - 41.4 p<0.01) in HTS group. No increase in PbtO2 in mannitol group (30.4-27.5)HTS used as second line therapy after mannitol (not direct comparison with equal baseline) so observed physiological improvements may be cumulative. No power study undertaken, small numbers. Mannitol and HTS not given at equimolar doses. Retrospective data collection so possible selection bias and no control group. Outcome measures did not include death and 1/3 of patients died.
Decrease in ICPHTS associated with lower ICP (27-15) compared with mannitol ( 29-24) p<0.001)
Increase in cerebral perfusion pressure (CPP) and cardiac outputHTS associated with higher CPP (p=0.02) and higher cardiac output (p=0.002) than mannitol
Ichai et al,
34 adult patients with isolated severe TBI and raised ICP randomly allocated to mannitol or hypertonic sodium lactate (HSL) infusions for 15 minutes. Rescue therapy to alternate treatment if ICP still not controlledProspective open randomised study, level 2b evidenceICP at 4 hours post treatment34 patients recruited, 17 in each group.9 received mannitol only, 12 received lactate only, 13 received both treatments. ICP lowered by 7mmHg (HSL) compared with 4mmHg (mannitol) p=0.016 and HSL had more prolonged effect of lowering ICP (-5.9 +/- 1 vs -3.2 +/- 0.9) p=0.009Study not blinded in outcome assessment hence potential for bias. Intention to treat and actual treatemts received both analysed then statistics presented were confusing mixture of both analyses. Neurological outcomes were not adequately described. Cross-over trial with side effects poorly reported. Sample size small although authors attempt to justify this and did attempt to power the study.
Percentage of successfully treated episodesHSL successful treatment in 90.4% compared with mannitol 70.4% p=0.053
1 year neurological status (Glasgow Outcome score)Better in HSL group than in mannitol group
Ware et al,
13 adults identified retrospectively with severe TBI and raised ICP admitted to San Francisco General Hospital ITU given initially mannitol then 23.4% saline (HTS) if ICP still highRetrospective cohort study. Level 4 evidenceReduction in ICP and duration of effectMean reductions in ICP both significant, mannitol (38-18 p<0.001) and HTS (36-20 p<0.001) but no significant difference in absolute numbers. Duration of effect: HTS longer than mannitol (96 mins vs 59 mins p=0.016)Retrospective notes/ chart review with poor reliability, unsure which patients were eligible, hence likely selection bias. Small numbers. Patients not uniformly managed, some had more complex schemes, hence effect may be cumulative. HTS was rescue therapy after mannitol had not worked thus mannitol was not a "control". Effects, once again, may have been cumulative (mannitol + HTS)
Glasgow Outcome Score at dischargeNo difference in outcome (all had GOS: moderate disability - death)
Complications of treatment (electrolyte abnormalities, convulsions, CCF, coagulopathy)No complications associated with HTS treatment
Battison et al,
9 patients with brain injury admitted to an Edinburgh ITU. 6 had TBI, 3 had SAH. All required ICP monitoring. Eache received two treatments each of 7.5% HTS/6% dextran (HSD) and 20% mannitol in random orderProspective randomised, controlled cross-over trial, level 2bICP reduction after each treatment and duration of effectHSD produced greater reduction in ICP than mannitol (13mmHG 95% CI 11.5-17.3 cf. 7.5mmHg 95% CI 5.8-11.8) p=0.014. HSD produced a longer duration of effect than mannitol (46 minutes longer 95% CI 6-182 p=0.044)Ver small numbers (only 6/9 patients relevant to this BestBET question). Randomisation of treatments not adequately described. Outcomes not assessed blind (but computer generated numbers so bias unlikely). Longer term outcomes not assessed (death/ disability). Unsure whether treatment groups were comparable at baseline (no demographic data presented). No intention to treat analysis performed. As both treatments were ultimately given effect may have been cumulative.
Vialet et al,
20 consecutive patients with head trauma, persistent coma and episodes of raised ICP resistant to standard modes of treatment assigned randomly to 20% mannitol or 7.5% saline infusionsProspective randomised study level 2b evidenceRate of failure for each treatment (persistence of raised ICP despite x2 infusions of same treatment)HTS 1 of 10 patients failed vs mannitol 7 of 10 patients failed (p<0.01)Small numbers (20 patients) with no power study. Method of randomisation not stated. Evaluation not assessed blinded, hence bias possible
Number and duration of raised ICP episodes per dayFewer raised ICP episodes per day with HTS: 6.9+/-5.6 vs mannitol: 13.3+/-14.6 (p<0.01). Daily duration of raised ICP for HTS was shorter: 67+/-85 mins than mannitol: 131+/-123 mins (p<0.01)
Death/ 90 day Glasgow Outcome ScoreNo difference between 2 groups (mannitol: 5 dead, 5 severe GOS, HTS: 4 dead, 6 severe GOS)


Standard therapy for the management of raised ICP in severe TBI includes intubation, normocapnic ventilation, sedation, head-up positioning, ensuring a mean arterial pressure >90 mm Hg and osmotherapy. Evidence for the use of one osmotherapy agent over another is scarce. The literature reveals many case reports of hypertonic sodium (HTS) solutions being used as a rescue therapy when mannitol has failed. Further, mannitol can cause renal toxicity with multiple administrations and may accumulate in the brain parenchyma, worsening cerebral oedema. Diuresis may also exacerbate systemic hypotension in polytrauma patients. HTS solutions, on the other hand, have a positive effect on haemodynamics and would seem more appropriate for use in polytrauma patients with TBI. Two of the five papers reviewed here (Oddo and Ware) demonstrate the effectiveness of HTS as a rescue therapy in reducing ICP after mannitol has failed. The remaining papers conclude that HTS solutions are superior to mannitol in both size and duration of ICP reduction, although data regarding long-term outcome either show no difference from mannitol or have not been reported. Further study is needed to determine the optimum concentration and composition of hypertonic sodium solutions for use in this context, ideally in terms of survival and neurological recovery, rather than effect on ICP alone.

Clinical Bottom Line

The available evidence suggests that HTS solutions across a range of concentrations are more effective than mannitol at reducing ICP in patients with TBI and should be considered as a first-line therapy for patients with TBI in the resuscitation room who display signs of raised ICP.

Level of Evidence

Level 2 - Studies considered were neither 1 or 3.


  1. M Oddo, JM Levine, S Frangos, et al. Effect of mannitol and hypertonic saline on cerebral oxygenation in patients with severe traumatic brain injury and refractory intracranial hypertension. J Neurol Neurosurg Psychiatry 2009 Aug. 80 (8): 916-20
  2. C Ichai, G Armando, JC Orban et al. Sodium lactate versus mannitol in the treatment of intracranial hypertensive episodes in severe traumatic brain-injured patients. Intensive Care Medicine 2009 Mar. 35(3): 471-9
  3. ML Ware, VM Nemani, M Meeker et al. Effects of 23.4% sodium chloride solution in reducing intracranial pressure in patients with traumatic brain injury: A preliminary study Neurosurgery 2005 Oct. 57(4): 727-36
  4. C Battison, PJD Andrews, C Graham, T Petty. Randomized, controlled trial on the effect of a 20% mannitol solution and a 7.5% saline/ 6% dextran solution on increased intracranial pressure after brain injury. Critical Care Medicine 2005 Jan. 33(1): 196-202
  5. R Vialet, J Albanese, L Thomachot et al Isovolume hypertonic solutes (sodium chloride or mannitol) in the treatment of refractory posttraumatic intracranial hypertension: 2ml/kg 7.5% saline is more effective than 2ml/kg 20% mannitol Critical Care Medicine 2003 31(6): 1683-7