Multiparametric monitoring in patients with traumatic brain injury: association with outcome

Objective

To investigate the association between multiple monitoring parameters and the outcome in patients with severe brain trauma.

Patients and methods

Thirty-one patients with a mean age of 43 ± 18 years were included in this prospective study. All of them had suffered serious traumatic brain damage (GCS ≤ 8 or GCS < 10 accompanied by a mass lesion in the initial CT scan). Their outcome was evaluated with the GOS in 6 months and they were divided according to that into two subgroups (favorable vs unfavorable outcome). During their ICU hospitalization, these patients were under brain function monitoring, consisting of three intra-cranial catheters: two for continuous monitoring of ICP and ptiO₂, and a third microdialysis catheter for the measurement of glucose, lactate, pyruvate and glycerol in the interstitial fluid of the brain. The microdialysis samples were collected every 2 hours and were automatically analyzed. The monitoring period lasted for up to 10 days.

Results

From the patients included, 15 (48.39%) had bad outcome (GOS = 1, 2, 3) and the remaining 16 (51.61%) good outcome (GOS = 4, 5). The group of patients with the good outcome had lower lactate/pyruvate ratio and glycerol values, compared with the group with the bad outcome (P < 0.001 and P = 0.044, respectively). This difference also existed when the maximum values of these parameters were evaluated (P = 0.03 and P = 0.011, respectively). Furthermore, mean L/P values >25 correlated with unfavorable outcome (P = 0.002). No relationship between glucose and 6-month outcome was found. From all other measurements, only mean CPP values correlated with outcome (P = 0.036). Mean ICP values were only marginally significant between the two subgroups. We could also demonstrate some correlations between the studying parameters: mean glucose values with mean pyruvate and mean L/P values (P = 0.001 and P = 0.048), maximum glycerol values with minimum lactate and maximum L/P values (P = 0.001 and P = 0.019) and maximum ICP values with minimum ptiO₂ values (0.021). Additionally, the presence of L/P ratio >40 was associated with glucose levels less than 0.5 (P = 0.001), and ptiO₂ values <5 mmHg presented a borderline correlation with glucose values <0.25 mg/dl (P = 0.058).

Conclusion

Microdialysis, ptiO₂ and ICP/CPP monitoring, when used together, may be a useful tool in the ICU setting, allowing continuous evaluation of brain damage and thus prognosis in patients with traumatic brain damage.