Reviewed by Sanjeev Sivakumar MD
Robba C, Graziano F, Rebora P et al. SYNAPSE-ICU Investigators. Intracranial pressure monitoring in patients with acute brain injury in the intensive care unit (SYNAPSE-ICU): an international, prospective observational cohort study. Lancet Neurol. 2021 Jul;20(7):548-558. doi: 10.1016/S1474-4422(21)00138-1. PMID: 34146513.
Read the Article
Background: In its latest iteration, the Brain Trauma Foundation downgraded the strength of recommendation for ICP monitoring in TBI. Consensus regarding specific thresholds for ICP monitoring, particularly for non-traumatic brain injury, is lacking. The BEST TRIP trial is the only randomized controlled trial that assessed the value of ICP monitoring in clinical decision making and showed no benefit in functional outcomes or mortality. The study was limited by sample size, the clinical setting, and a lack of information on the effect of ICP management. SYNAPSE-ICU aims to describe the variations in ICP monitoring practices: indications, treatment intensity in patients with and without monitoring, and association with patient outcomes
Methods: Prospective, observational cohort study performed in 146 ICUs from 42 countries. Inclusion criteria were adult patients with acute brain injury (TBI, ICH or SAH) admitted with altered consciousness (GCS eye response score of 1 and motor score ≤5). Patients with bilateral unreactive pupils, GCS of 3, death within 48h were excluded. Each center enrolled maximum of 90 patients over 12 weeks. Therapy intensity level [TIL; ranging from 1 (minimum) to 38 (maximum level)] was used to quantify practice variations in ICP interventions. Mann-Whitney U test was used for continuous data and χ² test for categorical data. Median odds ratio (MOR) – referring to the odds of ICP monitoring use between two randomly selected centers for patients with same covariates, was used to quantify intergroup differences. MOR greater than 1 denotes high between-center variability. Primary endpoints were 6m mortality and 6m Glasgow Outcome Scale Extended (GOSE; unfavorable outcome defined as GOSE < 5). Secondary outcomes were hospital mortality and GOSE score. A propensity score method was used to estimate the association between use of ICP monitoring and 6m outcomes. Authors also created models to account for pupillary reactivity. This was done since pupillary reactivity was found to modify association between recipients of ICP monitoring and outcome.
Results: 2395 patients including 1287 (54%) with TBI, 587 (25%) with ICH and 521 (22%) with SAH were included over a 12-month period. The median age was 55 years; 65% were male; 82% were from high income countries. 82% had altered consciousness on admission; 34% had at least 1 unreactive pupil. 1332 (56%) received ICP monitoring and were younger with fewer comorbidities . The most common reason for ICP monitoring was low GCS (71%), in all forms of acute brain injury. The most common reasons for not placing ICP monitor were severe clinical status (25%), severe neuroimaging findings (25%) and local policy (18%). Intraparenchymal device was preferentially used in TBI (73%), intraventricular device in SAH (53%) and ICH (54%). Majority received ICP monitor within first day of admission (86%), with median duration of monitoring 8 days (IQR 4-13. There was considerable variability in the use of ICP monitoring (MOR 4·5, 95% CI 3·8–4·9) between centers.
Therapy Intensity Levels: The median TIL score during first week of ICU was 7 (IQR 5-10); TIL was higher in patients with ICP monitoring (9 [IQR 7-12] vs. 5 [3-8]; p<0.0001), and at day 1 (8 [6-11] vs 5 [3-8]); p<0.0001), day 3 (6 [4-8] vs 4 [2-6]; p<0.0001), and on day 7 (5 [3-7] vs 3 [2-5]; p<0.0001). An increment of one point in TIL was associated with a reduction in mortality (HR 0·94, 95% CI 0·91–0·98; p=0·001).
Outcomes: Patients with ICP monitoring had significantly lower in-hospital mortality (28% Vs. 42%; p<0.0001) and 6m mortality (34% Vs. 49%; p<0.039). The 6m mortality in high income countries was lower among patients with ICP monitoring (34% Vs. 55%); no differences were noted in low- and middle-income countries. Unfavorable neurological outcomes were similar in both groups. Daily median ICP value was associated with an unfavorable neurological outcome; OR for each 5mm Hg increase in ICP value 1.04, 95% CI 1.02-1.05, p<0.0001.
Clinical status, TIL and ICP monitoring: Among patients with at least one unreactive pupil, ICP monitoring was associated with significantly lower 6m mortality (HR 0·35, 95% CI 0·26–0·47; p<0·0001) after adjustment for TIL, and better neurological outcome at 6 months (OR 0·38, 95% CI 0·26–0·56; p=0·0025). Odds for poor neurological outcome was similar in both groups. In patients with ICH, ICP monitoring was associated with lower mortality among patients with bilateral pupillary reactivity (HR 0.57, 95% CI 0.38-0.87), and lower odds of unfavorable neurological outcomes at 6m in patients with one unreactive pupil (OR 0.23, 95% CI 0.04-1.0).
Commentary: The key findings of this study are 1) Use of ICP monitoring is associated with a more aggressive therapeutic approach and improved outcomes in most severe cases. 2) There is significant variability between centers and countries in the use of ICP monitoring.
Preinjury characteristics (younger age, lesser comorbidities) and severity of injury influenced the decision to insert an ICP monitor. The absence of universal guidelines, local policies, being in a lower/middle income country influenced ICP monitoring. Contrary to the BEST TRIP study, the results of the SYNAPSE-ICU study show that in severe cases, ICP monitoring was associated with a more aggressive approach and better long-term clinical results. In the study by Chesnut et al., higher TIL was seen in patients managed using CT findings and clinical status as opposed to ICP monitoring. In severe cases with one unreactive pupil, ICP monitoring was associated with reduced hospital and 6m mortality rates across all subgroups (TBI, ICH and SAH). An increase in 1 point in TIL resulted in a reduction in probability for unfavorable outcomes.
The strengths of this study include the international, multicenter cohort with a large sample size and high rates of follow up. Limitations of the study include i) observational nature which precludes causal inferences ii) varying pathologies (TBI, ICH, SAH) were grouped for selection iii) neurological severity was evaluated on admission and disease trajectories and ICU complications were not modelled for iv) unmeasured confounders v) withdrawal of life sustaining measures was not included in analysis. However, to overcome some of these limitations, the authors had a statistical plan to include propensity score analysis, performed sensitivity analysis with latent confounder which confirmed the robust nature of results, and performed sensitivity analysis in groups excluding patients with severe neurological status on admission and death within 48 h. This supported their original results. #LiteratureWatch