Still not safe: a propensity score matched analysis of colloid use in traumatic brain injury

Introduction: The use of 4% albumin for resuscitation in patients with traumatic brain injury (TBI) has been associated with higher mortality rates compared to crystalloids. Despite this finding, there are clinical scenarios in which colloids containing albumin such as 5% and 25% albumin or fresh frozen plasma (FFP) may be indicated in certain TBI patients. Methods: We conducted a retrospective propensity score matched cohort study evaluating adult patients admitted to our tertiary care academic medical center with blunt TBI who received colloid therapy. Colloids were defined as either 5% albumin, 25% albumin, or FFP. Patients were excluded if they were diagnosed with penetrating trauma or expired within 24 hours of admission. The primary outcome was in-hospital mortality. Secondary outcomes included intensive care unit (ICU) length of stay and discharge disposition. Results: Of the 434 patients who met inclusion criteria, 39 (9.0%) received colloid therapy and were matched 1:1 to 39 patients in the control group. Baseline characteristics were well balanced except for a higher mean injury severity score (ISS) (23.13 vs. 17, P = 0.021) and a lower mean admission Glasgow Coma Score (GCS) (12 vs. 14, P = 0.02) in the colloid group. The majority (73.1%) of patients had a grade 3 Brain Injury Guideline TBI. Patients in the colloid group had significantly higher mortality rates compared to the control group (20.5% vs. 2.6%, P = 0.013). Multivariate regression analyses were performed evaluating albumin alone and FFP alone compared to the control group. Albumin alone was associated with increased mortality after adjustment for ISS [OR 12.46, 95% CI (1.26 – 123.36)] and after adjustment for GCS [OR 14.93, 95% CI (1.24 – 179.15)]. FFP alone was not associated with increased mortality compared to control [OR 8.00, 95% CI (0.84 – 76.62)]. ICU length of stay was longer in the colloid group (12.74 days vs. 4.21 days, P = 0.002). Significantly more patients in the colloid group had a non-home discharge disposition (71.8% vs. 38.5%, P= 0.003). Conclusions: Colloid therapy may be associated with increased mortality in patients with TBI. The risk of harm appears to be attributed to 5% and 25% albumin as opposed to FFP. Additional studies are needed to evaluate the impact of colloid therapy in TBI.