Frequency and Impact of Intensive Care Unit Complications on Moderate-Severe Traumatic Brain Injury: Early Results of the Outcome Prognostication in Traumatic Brain Injury (OPTIMISM) Study

Background

Known predictors of adverse outcomes in patients with moderate-severe TBI (msTBI) explain only a relatively small proportion of patient-related outcomes. The frequency and impact of intensive care unit complications (ICU-COMPL) on msTBI-associated outcomes are poorly understood.

Methods

In 213 consecutive msTBI patients admitted to a Level I Trauma Center neuro trauma ICU, twenty-eight ICU-COMPL (21 medical and 7 neurological) were prospectively collected and adjudicated by group consensus, using pre-defined criteria. We determined frequencies, and explored associations of ICU-COMPL and hospital discharge outcomes using multivariable logistic regression.

Results

The average age of the study sample was 53 years, and the median presenting Glasgow Coma Scale and Injury Severity Scores were 5 and 27, respectively. Hyperglycemia (79 %), fever (62 %), systemic inflammatory response syndrome (60 %), and hypotension requiring vasopressors (42 %) were the four most common medical ICU-COMPL. Herniation (39 %), intracranial rebleed (39 %), and brain edema requiring osmotherapy (37 %) were the three most common neurological ICU-COMPL. After adjusting for admission variables, duration of ventilation, and ICU length-of-stay, patients with brain edema (OR 5.8; 95 % CI 2, 16.7) had a significantly increased odds for dying during hospitalization whereas patients with hospital-acquired urinary tract infection (UTI) had a decreased odds (OR 0.05; 95 % CI 0.005, 0.6). Sensitivity analysis revealed that UTI occurred later, suggesting a non-causal association with survival. Brain herniation (OR 15.7; 95 % CI 2.6, 95.4) was associated with an unfavorable functional status (GOS 1–3).

Conclusion

ICU-COMPL are very common after msTBI, have a considerable impact on short-term outcomes, and should be considered in the prognostication of these high risk patients. Survival associations of time-dependent complications warrant cautious interpretation.     

Admission Braden Skin Score Independently Predicts Mortality in Cardiac Intensive Care Patients

ObjectiveTo determine whether a low Braden skin score (BSS), reflecting increased risk for skin pressure injury, would predict lower survival in cardiac intensive care unit (CICU) patients after adjustment for illness severity and comorbidities.Patients and MethodsThis retrospective cohort study included consecutive unique adult patients admitted to a single tertiary care referral hospital CICU from January 1, 2007, through December 31, 2015, who had a BSS documented on CICU admission. The primary outcome was all-cause hospital mortality, using elastic net penalized logistic regression to determine predictors of hospital mortality. The secondary outcome was all-cause post-discharge mortality, using Cox proportional hazards models to determine predictors of post-discharge mortality.ResultsThe study included 9552 patients with a mean age of 67.4±15.2 years (3589 [37.6%] were females) and a hospital mortality rate of 8.3%. Admission BSS was inversely associated with hospital mortality (unadjusted odds ratio, 0.70; 95% CI, 0.68-0.72; P<.001; area under the receiver operator curve, 0.80; 95% CI, 0.78-0.82), with increased short-term mortality as a function of decreasing admission BSS. After adjustment for illness severity and comorbidities using multivariable analysis, admission BSS remained inversely associated with hospital mortality (adjusted odds ratio, 0.88; 95% CI, 0.85-0.92; P<.001). Among hospital survivors, admission BSS was inversely associated with post-discharge mortality after adjustment for illness severity and comorbidities (adjusted hazard ratio, 0.89; 95% CI, 0.88-0. 90; P<.001).ConclusionThe admission BSS, a simple inexpensive bedside nursing assessment potentially reflecting frailty and overall illness acuity, was independently associated with hospital and post-discharge mortality when added to established multiparametric illness severity scores among contemporary CICU patients.     

Comparison of a Low,Fixed Dose and a High,Weight-Based Dose of Recombinant Factor VIIa in the Treatment of Warfarin-Associated Intracranial Hemorrhage

Background

Recombinant activated Factor VII (rFVIIa) can be used for rapid INR normalization in patients with warfarin-associated intracranial hemorrhage (WA-ICH); however, the optimal dose to normalize INR has not been established.

Methods

This is a retrospective review comparing two rFVIIa hospital protocols for WA-ICH [weight-based dose (80 mcg/kg) or fixed dose (2 mg)]. Primary endpoint was the percentage of patients with INR reversal (INR <1.3) at the next INR draw and the need for further doses of rFVIIa. Secondary endpoints included time to documented INR reversal and sustained INR normalization, morbidity, mortality, change in hematoma size, cost, and adverse drug reactions.

Results

Twenty-nine patients were included in each group. The weight-based group received a mean dose of 78.9 ± 21 mcg/kg versus 26.6 ± 8 mcg/kg in the fixed dose group. More patients in the fixed dose protocol achieved documented INR reversal than those in the weight-based group (92.6 vs 72.4 %, p = 0.19). The weight-based group achieved INR normalization in 229.5 [102, 331] minutes versus 165 [83, 447] minutes in the fixed dose group (p=0.02). Time to sustained INR normalization was similar in both groups. Four patients in the fixed dose group received an additional dose of 1 mg per hospital protocol. With the exception of medication acquisition cost savings of about $4,300 per patient who received fixed dose protocol, all other endpoints were similar between groups.

Conclusions

A low, fixed dose of rFVIIa appears to be as effective as a high, weight-based dose in achieving INR normalization in patients with WA-ICH.