Hospital Characteristics and Mortality in Aneurysmal Subarachnoid Hemorrhage

Serum neurofilament light as a predictor of outcome in subarachnoid haemorrhage

BACKGROUND

Prognostication of clinical outcome in patients suffering from aneurysmal subarachnoid haemorrhage (SAH) is a challenge. There are no biochemical markers in routine use that can aid in prognostication. Neurofilament light (NFL) measured in cerebrospinal fluid (CSF) has been associated with clinical outcome in previous studies.

OBJECTIVE

To investigate if serum levels of NFL correlate with CSF levels and long-term clinical outcome in patients suffering from SAH.

METHODS

We conducted an observational cohort study of 88 patients treated for SAH at Umeå University Hospital in 2014-2018. Serum and CSF samples were analysed using an enzyme-linked immunosorbent assay to quantify NFL levels. Outcome was assessed using Glasgow Outcome Scale Extended and dichotomised as favourable or unfavourable. Differences in NFL levels between outcome groups were analysed using repeated measurements ANOVA. Relationship between CSF and serum NFL levels was analysed using Pearson's correlation. A multivariate binary logistic regression model and a receiver operation characteristic curve were used to assess the predictive value of serum NFL.

RESULTS

A significant correlation between serum and CSF-NFL levels could be seen (Pearson's correlation coefficient = 0.7, p < .0001). Mean level of serum NFL was higher in the unfavourable outcome group than the favourable outcome group (p < .0001), in all epochs of SAH, and correlated with initial disease severity on the World Federation of Neurosurgical Societies scale. Serum NFL in the late phase displayed the best predictive potential in a receiver operation characteristic curve analysis (AUC=0.845, p < .0001).

CONCLUSION

Levels of NFL in serum and CSF are correlated. Early serum NFL levels seem to reflect initial tissue damage and serum NFL levels in the late phase may reflect secondary events such as vasospasm or delayed cerebral ischemia. Serum NFL may be used as a prognostic marker of clinical outcome in SAH.

Predictors of Time to Aneurysm Repair and Mortality in Aneurysmal Subarachnoid Hemorrhage

ABSTRACT

BACKGROUND : Prompt aneurysm repair is essential to prevent rebleeding after aneurysmal subarachnoid hemorrhage. To date, most studies on this topic have focused on 1 set of predictors (eg, hospital or patient characteristics) and on 1 outcome (either time to aneurysm repair or mortality). The purpose of this study was to test a model that includes hospital and patient characteristics as predictors of time to aneurysm repair and mortality, controlling for disease severity and comorbidity, and considering time to aneurysm repair as a potential influence in these relationships. METHODS : A sample of aneurysmal subarachnoid hemorrhage patients with a principal procedure of clipping or coiling was selected (n = 387) from a statewide administrative database for cross-sectional retrospective analysis. The primary study outcome was in-hospital mortality. Independent variables were level of stroke center, age, race, sex, and type of aneurysm repair. Hierarchical logistic regression was used to estimate the probability of in-hospital death. RESULTS : Patients who underwent a coiling procedure were more likely to be treated within the first 24 hours of admission compared with those undergoing clipping (odds ratio, 0.54; 95% CI, 0.35-0.84; P = .01). Patients treated at a certified comprehensive stroke center (CSC) had a 72% reduction in odds of death compared with those treated at primary stroke centers (odds ratio, 0.28; 95% CI, 0.10-0.77; P = .01), after controlling for disease severity and comorbid conditions. Time to aneurysm repair was not significantly associated with mortality and did not influence the relationship between hospital and patient characteristics and mortality. CONCLUSION : Our results indicate that treatment at a CSC was associated with a lower risk of in-hospital mortality. Time to aneurysm repair did not influence mortality and did not explain the mortality benefit observed in CSCs. Research is needed to explore interdisciplinary hospital factors including nursing and nurse-sensitive interventions that may explain the relationship between CSCs and mortality outcomes.