Accuracy of Neurovascular Fellows' Prognostication of Outcome After Subarachnoid Hemorrhage

Machine learning validation of a simple prediction model for the correlation between advanced age and clinical outcomes in patients with aneurysmal subarachnoid hemorrhage

Adverse effects of advanced age and poor initial neurological status on outcomes of patients with aneurysmal subarachnoid hemorrhage (SAH) have been documented. While a predictive model of the non-linear correlation between advanced age and clinical outcome has been reported, no previous model has been validated. Therefore, we created a prediction model of the non-linear correlation between advanced age and clinical outcome by machine learning and validated it using a separate cohort. Data of aneurysmal SAH patients treated by surgical clipping or endovascular coiling between 2003 and 2019 were obtained from the Japanese Stroke Databank (derivation cohort, n = 9,657) and "Predict for Outcome Study of Aneurysmal Subarachnoid Hemorrhage" (validation cohort, n = 5,085). Generalized additive models (GAMs) for poor outcome (modified Rankin Scale score ≥ 3 at discharge) were fitted with age transformation using spline curves for each World Federation of Neurological Societies grade. The discrimination property and calibration plot of unadjusted and adjusted models were assessed using the validation cohort. The derivation and validation cohorts included 3,610 and 3,251 patients, respectively. Regarding discrimination, areas under the receiver operating characteristic curve for the derivation and validation cohorts were 0.835 and 0.827, respectively, in the unadjusted model and 0.844 and 0.836, respectively, in the adjusted model. An unbiased correlation was confirmed between predicted and observed probabilities of poor outcomes. GAM could help visualize the correlation between age and clinical outcomes. Our prediction model can quantitatively aid in treatment decision-making and can be effective for most diseases and treatment settings. Trial Registration: UMIN Clinical Trials Registry (Date 2/22/2022/ ID, UMIN000046282 number, R000052809 URL, https//www.umin.ac.jp/ctr/index.htm) and the Japan Registry of Clinical Trials (Date 3/28/2022 /No. jRCT1060210092 URL, https//jrct.niph.go.jp/).

Accuracy of Early Neuroprognostication in Pediatric Severe Traumatic Brain Injury

BACKGROUND

Children with severe traumatic brain injury (sTBI) are at risk for neurological sequelae impacting function. Clinicians are tasked with neuroprognostication to assist in decision-making. We describe a single-center study assessing clinicians' neuroprognostication accuracy.

METHODS

Clinicians of various specialties caring for children with sTBI were asked to predict their patients' functioning three to six months postinjury. Clinicians were asked to participate in the study if their patient had survived but not returned to baseline between day 4 and 7 postinjury. The outcome tool utilized was the functional status scale (FSS), ranging from 6 to 30 (best-worst function). Predicted scores were compared with actual scores three to six months postinjury. Lin concordance correlation coefficients were used to estimate agreement between predicted and actual FSS. Outcome was dichotomized as good (FSS 6 to 8) or poor (FSS ≥9). Positive and negative predictive values for poor outcome were calculated. Pessimistic prognostic prediction was defined as predicted worse outcome by ≥3 FSS points. Demographic and clinical variables were collected.

RESULTS

A total of 107 surveys were collected on 24 patients. Two children died. Fifteen children had complete (FSS = 6) or near-complete (FSS = 7) recovery. Mean predicted and actual FSS scores were 10.8 (S.D. 5.6) and 8.6 (S.D. 4.1), respectively. Predicted FSS scores were higher than actual scores (P < 0.001). Eight children had collective pessimistic prognostic prediction.

CONCLUSIONS

Clinicians predicted worse functional outcomes, despite high percentage of patients with near-normal function at follow-up clinic. Certain patient and provider factors were noted to impact accuracy and need to be studied in larger cohorts.

Machine Intelligence in Cerebrovascular and Endovascular Neurosurgery

The advent of different realms of computational neurosurgery-including not only machine intelligence but also visualization techniques such as mixed reality and robotic applications-is beginning to impact both open vascular as well as endovascular neurosurgery. Especially in this relatively common patient population of often very fragile patients, with potential for devastating complications and clinical outcomes and sometimes highly complex pathologies, computer assistance could prove particularly useful. In this chapter, state-of-the-art applications of machine learning toward vascular patients are elucidated: Beginning from simple clinical diagnostic, prognostic, and predictive modeling, to the interpretation of medical imaging (radiomics, segmentation, and diagnostic assistance) and synthetic imaging (image modality conversion, super-resolution, and 2D-to-3D-synthesis), up to intraoperative applications of computer vision (robotic steering, rapid intraoperative histopathology, and anatomical and surgical phase recognition), and natural language processing (enabling model training and big data, documentation, and large language models)-this chapter provides a "tour de force" of machine intelligence in the realm of neurovascular medicine.

Recent Trends in the Withdrawal of Life-Sustaining Treatment in Patients with Acute Cerebrovascular Disease : 2017-2021

OBJECTIVE

The Act on Life-Sustaining Treatment (LST) decisions for end-of-life patients has been effective since February 2018. An increasing number of patients and their families want to withhold or withdraw from LST when medical futility is expected. This study aimed to investigate the status of the Act on LST decisions for patients with acute cerebrovascular disease at a single hospital.

METHODS

Between January 2017 and December 2021, 227 patients with acute cerebrovascular diseases, including hemorrhagic stroke (n=184) and ischemic stroke (n=43), died at the hospital. The study period was divided into the periods before and after the Act.

RESULTS

The duration of hospitalization decreased after the Act was implemented compared to before (15.9±16.1 vs. 11.2±18.6 days, p=0.127). The rate of obtaining consent for the LST plan tended to increase after the Act (139/183 [76.0%] vs. 27/44 [61.4%], p=0.077). Notably, none of the patients made an LST decision independently. Ventilator withdrawal was more frequently performed after the Act than before (52/183 [28.4%] vs. 0/44 [0%], p<0.001). Conversely, the rate of organ donation decreased after the Act was implemented (5/183 [2.7%] vs. 6/44 [13.6%], p=0.008). Refusal to undergo surgery was more common after the Act was implemented than before (87/149 [58.4%] vs. 15/41 [36.6%], p=0.021) among the 190 patients who required surgery.

CONCLUSION

After the Act on LST decisions was implemented, the rate of LST withdrawal increased in patients with acute cerebrovascular disease. However, the decision to withdraw LST was made by the patient's family rather than the patient themselves. After the execution of the Act, we also observed an increased rate of refusal to undergo surgery and a decreased rate of organ donation. The Act on LST decisions may reduce unnecessary treatments that prolong end-of-life processes without a curative effect. However, the widespread application of this law may also reduce beneficial treatments and contribute to a decline in organ donation.

Physician experience improves ability to predict 6-month functional outcome of severe traumatic brain injury

BACKGROUND

The functional prognosis of severe traumatic brain injury (TBI) during the acute phase is often poor and uncertain. We aimed to quantify the elements that shade the degree of uncertainty in prognostic determination of TBI and to better understand the role of clinical experience in prognostic quality.

METHODS

This was an observational, prospective, multicenter study. The medical records of 16 patients with moderate or severe TBI in 2020 were randomly drawn from a previous study and submitted to two groups of physicians: senior and junior. The senior physician group had graduated from a critical care fellowship, and the junior physician group had at least 3 years of anesthesia and critical care residency. They were asked for each patient, based on the reading of clinical data and CT images of the first 24 h, to determine the probability of an unfavorable outcome (Glasgow Outcome Scale < 4) at 6 months between 0 and 100, and their level of confidence. These estimations were compared with the actual evolution.

RESULTS

Eighteen senior physicians and 18 junior physicians in 4 neuro-intensive care units were included in 2021. We observed that senior physicians performed better than junior physicians, with 73% (95% confidence interval (CI) 65-79) and 62% (95% CI 56-67) correct predictions, respectively, in the senior and junior groups (p = 0.006). The risk factors for incorrect prediction were junior group (OR 1.71, 95% CI 1.15-2.55), low confidence in the estimation (OR 1.76, 95% CI 1.18-2.63), and low level of agreement on prediction between senior physicians (OR 6.78, 95% CI 3.45-13.35).

CONCLUSIONS

Determining functional prognosis in the acute phase of severe TBI involves uncertainty. This uncertainty should be modulated by the experience and confidence of the physician, and especially on the degree of agreement between physicians.

Evolution of Making Clinical Predictions in Neurosurgery

Prediction of clinical outcomes is an essential task for every physician. Physicians may base their clinical prediction of an individual patient on their intuition and on scientific material such as studies presenting population risks and studies reporting on risk factors (prognostic factors). A relatively new and more informative approach for making clinical predictions relies on the use of statistical models that simultaneously consider multiple predictors that provide an estimate of the patient's absolute risk of an outcome. There is a growing body of literature in the neurosurgical field reporting on clinical prediction models. These tools have high potential in supporting (not replacing) neurosurgeons with their prediction of a patient's outcome. If used sensibly, these tools pave the way for more informed decision-making with or for individual patients. Patients and their significant others want to know their risk of the anticipated outcome, how it is derived, and the uncertainty associated with it. Learning from these prediction models and communicating the output to others has become an increasingly important skill neurosurgeons have to master. This article describes the evolution of making clinical predictions in neurosurgery, synopsizes key phases for the generation of a useful clinical prediction model, and addresses some considerations when deploying and communicating the results of a prediction model. The paper is illustrated with multiple examples from the neurosurgical literature, including predicting arachnoid cyst rupture, predicting rebleeding in patients suffering from aneurysmal subarachnoid hemorrhage, and predicting survival in glioblastoma patients.

External Validation of a Neural Network Model in Aneurysmal Subarachnoid Hemorrhage: A Comparison With Conventional Logistic Regression Models

BACKGROUND

Interest in machine learning (ML)-based predictive modeling has led to the development of models predicting outcomes after aneurysmal subarachnoid hemorrhage (aSAH), including the Nijmegen acute subarachnoid hemorrhage calculator (Nutshell). Generalizability of such models to external data remains unclear.

OBJECTIVE

To externally validate the performance of the Nutshell tool while comparing it with the conventional Subarachnoid Hemorrhage International Trialists (SAHIT) models and to review the ML literature on outcome prediction after aSAH and aneurysm treatment.

METHODS

A prospectively maintained database of patients with aSAH presenting consecutively to our institution in the 2013 to 2018 period was used. The web-based Nutshell and SAHIT calculators were used to derive the risks of poor long-term (12-18 months) outcomes and 30-day mortality. Discrimination was evaluated using the area under the curve (AUC), and calibration was investigated using calibration plots. The literature on relevant ML models was surveyed for a synopsis.

RESULTS

In 269 patients with aSAH, the SAHIT models outperformed the Nutshell tool (AUC: 0.786 vs 0.689, P = .025) in predicting long-term functional outcomes. A logistic regression model of the Nutshell variables derived from our data achieved adequate discrimination (AUC = 0.759) of poor outcomes. The SAHIT models outperformed the Nutshell tool in predicting 30-day mortality (AUC: 0.810 vs 0.636, P < .001). Calibration properties were more favorable for the SAHIT models. Most published aneurysm-related ML-based outcome models lack external validation and usable testing platforms.

CONCLUSION

The Nutshell tool demonstrated limited performance on external validation in comparison with the SAHIT models. External validation and the dissemination of testing platforms for ML models must be emphasized.

The intersection of prognostication and code status in patients with severe brain injury

Accurately estimating the prognosis of brain injury patients can be difficult, especially early in their course. Prognostication is important because it largely determines the care level we provide, from aggressive treatment for patients we predict could have a good outcome to withdrawal of treatment for those we expect will have a poor outcome. Accurate prognostication is required for ethical decision-making. However, several studies have shown that prognostication is frequently inaccurate and variable. Overly optimistic prognostication can lead to false hope and futile care. Overly pessimistic prognostication can lead to therapeutic nihilism. Overlapping is the powerful effect that cognitive biases, in particular code status, can play in shaping our perceptions and the care level we provide. The presence of Do Not Resuscitate orders has been shown to be associated with increased mortality. Based on a comprehensive search of peer-reviewed journals using a wide range of key terms, including prognostication, critical illness, brain injury, cognitive bias, and code status, the following is a review of prognostic accuracy and the effect of code status on outcome. Because withdrawal of treatment is the most common cause of death in the ICU, a clearer understanding of this intersection of prognostication and code status is needed.

Early-stage serum Stanniocalcin 1 as a predictor of outcome in patients with aneurysmal subarachnoid hemorrhage

Stanniocalcin-1 (STC1) takes part in anti-inflammatory and anti-oxidative processes, thus demonstrating neuroprotective properties. Early brain injuries associated with initial subarachnoid hemorrhage typically led to secondary cerebral infarction and poor outcomes. This retrospective study aimed to clarify the clinical significance of serum STC1 level in patients with subarachnoid hemorrhage.We collected demographic information, comorbidities, neurological status in detail. All blood samples were collected on admission. Enzyme-linked immunosorbent assay kits were used to detect the serum level of STC1. Spearman analysis was used to explore the relationship between STC1 and clinical severity. Multivariate logistic regression was used to investigate the prognostic role of STC1 in patients with aneurysmal subarachnoid hemorrhage (aSAH). Receiver operating characteristic curve was performed to investigate the power of STC1 in predicting outcome in aSAH patients.Serum STC1 concentration was significantly higher in aSAH patients than in healthy individuals. Serum concentration of STC1 positively correlated with Hunt-Hess grade (r = 0.62, P < .01) and Fisher grade (r = 0.48, P < .01), and negatively correlated with Glasgow Coma Scale on admission (r = -0.45, P < .01). Patients with delayed cerebral ischemia (DCI) had higher level of serum STC1 than those without DCI (13.12 ± 1.44 vs 8.56 ± 0.31, P < .01). Moreover, patients with poor outcome had higher concentration of STC1 than patients with good outcome (11.82 ± 0.62 vs 8.21 ± 0.35,P < 0.01). Results of univariate and multivariate logistic analysis revealed that Hunt-hess III-IV, DCI, and high STC1 level were independent risk factors associated with poor outcome of patients with aSAH. Further analysis revealed that combination of STC1 with Hunt-hess grade was more superior to 2 indicators alone in predicting clinical outcome of aSAH patients.STC1 can be used as a novel biomarker in predicting outcome of patients with aSAH, especially when combined with Hunt-hess grade.

Visualisation of the non-linear correlation between age and poor outcome in patients with aneurysmal subarachnoid haemorrhage

OBJECTIVE

To visualise the non-linear correlation between age and poor outcome at discharge in patients with aneurysmal subarachnoid haemorrhage (SAH) while adjusting for covariates, and to address the heterogeneity of this correlation depending on disease severity by a registry-based design.

METHODS

We extracted data from the Japanese Stroke Databank registry for patients with SAH treated via surgical clipping or endovascular coiling within 3 days of SAH onset between 2000 and 2017. Poor outcome was defined as a modified Rankin Scale Score ≥3 at discharge. Variable importance was calculated using machine learning (random forest) model. Correlations between age and poor outcome while adjusting for covariates were determined using generalised additive models in which spline-transformed age was fit to each neurological grade of World Federation of Neurological Societies (WFNS) and treatment.

RESULTS

In total, 4149 patients were included in the analysis. WFNS grade and age had the largest and second largest variable importance in predicting the outcome. The non-linear correlation between age and poor outcome was visualised after adjusting for other covariates. For grades I-III, the risk slope for unit age was relatively smaller at younger ages and larger at older ages; for grade IV, the slope was steep even in younger ages; while for grade V, it was relatively smooth, but with high risk even at younger ages.

CONCLUSIONS

The clear visualisation of the non-linear correlation between age and poor outcome in this study can aid clinical decision making and help inform patients with aneurysmal SAH and their families better.

Psychological distress after subarachnoid haemorrhage: A systematic review and meta-analysis

OBJECTIVE

Psychological distress is a common complication in patients after Subarachnoid haemorrhage (SAH) which often has significant impact on the prognosis. The objective of this study was to determine the pooled prevalence of anxiety symptoms and depressive symptoms in patients after SAH and identify relevant risk factors.

METHODS

The study adopted a systematic review and meta-analysis protocol. Multiple databases including EMBASE, Medline, PsychInfo, and Web of Science were searched for publications before 1st January 2020. Screening, data extraction, and quality assessment were undertaken following the PRISMA guidelines for preferred reporting of systematic reviews and meta-analysis. The random-effects model was used to calculate pooled prevalence rates. Meta-analysis was conducted using Comprehensive Meta-analysis software. The review protocol was registered on PROSPERO (CRD42020182594).

RESULTS

42 studies reporting anxiety symptoms and 64 studies reporting depressive symptoms were included. The pooled short term(<3 years) and long term(≥3 years) prevalence rates of anxiety symptoms were 31.4%(95% CI: 23.6%, 40.4%) and 40.4%(95% CI: 31.6%, 49.8%), respectively, whereas the pooled short term and long term prevalence rates of depressive symptoms were 25.2%(95%CI: 17.8%, 34.5%) and 35.8%(95%CI: 28.6%, 43.6%), respectively. Gender and pre-existing psychiatric conditions were identified as potential risk factors.

CONCLUSIONS

The high prevalence of anxiety symptoms and depressive symptoms after SAH highlights the need for appropriate assessment and management of psychological stress in patients after SAH. Further research is warranted to explore potential underlying mechanisms and to develop holistic interventions that incorporate understanding of both the biological and psychological impact of SAH.

Application of Near-Infrared Spectroscopy for the Detection of Delayed Cerebral Ischemia in Poor-Grade Subarachnoid Hemorrhage

BACKGROUND

The objective of this study was to investigate the clinical feasibility of near-infrared spectroscopy (NIRS) for the detection of delayed cerebral ischemia (DCI) in patients with poor-grade subarachnoid hemorrhage (SAH) treated with coil embolization.

METHODS

Cerebral regional oxygen saturation (rSO₂) was continuously monitored via two-channel NIRS for 14 days following SAH. The rSO₂ levels according to DCI were analyzed by using the Mann-Whitney U-test. A receiver operating characteristic curve was generated on the basis of changes in rSO₂ by using the rSO₂ level on day 1 as a reference value to determine the optimal cutoff value for identifying DCI.

RESULTS

Twenty-four patients with poor-grade SAH were included (DCI, n = 8 [33.3%]; non-DCI, n = 16 [66.7%]). The rSO₂ levels of patients with DCI were significantly lowered from 6 to 9 days compared with those in without DCI. The rSO₂ level was 62.55% (58.30-63.40%) on day 6 in patients with DCI versus 65.40% (60.90-68.70%) in those without DCI. By day 7, it was 60.40% (58.10-61.90%) in patients with DCI versus 64.25% (62.50-67.10%) those without DCI. By day 8, it was 58.90% (56.50-63.10%) in patients with DCI versus 66.05% (59.90-69.20%) in those without DCI, and by day 9, it was 60.85% (58.40-65.20%) in patients with DCI versus 65.80% (62.70-68.30%) in those without DCI. A decline of greater than 14.5% in the rSO₂ rate yielded a sensitivity of 92.86% (95% confidence interval: 66.1-99.8%) and a specificity of 88.24% (95% confidence interval: 72.5-96.7%) for identifying DCI. A decrease by more than 14.7% of the rSO₂ level indicates a sensitivity of 85.7% and a specificity of 85.7% for identifying DCI.

CONCLUSIONS

Near-infrared spectroscopy shows some promising results for the detection of DCI in patients with poor-grade SAH. Further studies involving a large cohort of the SAH population are required to confirm our results.

Development and Validation of Scoring Indication of Surgical Clipping and Endovascular Coiling for Aneurysmal Subarachnoid Hemorrhage from the Post Hoc Analysis of Japan Stroke Data Bank

There are no scoring methods for optimal treatment of patients with aneurysmal subarachnoid hemorrhage (aSAH). We developed a scoring model to predict clinical outcomes according to aSAH risk factors using data from the Japan Stroke Data Bank (JSDB). Of 5344 patients initially registered in the JSDB, 3547 met the inclusion criteria. Patients had been diagnosed with aSAH and treated with surgical clipping or endovascular coiling between 1998 and 2013. We performed multivariate logistic regression for poor outcomes at discharge, indicated by a modified Rankin Scale (mRS) score >2, and in-hospital mortality for both treatment methods. Based on each risk factor, we developed a scoring model assessing its validity using another dataset of our institution. In the surgical clipping group, scoring criteria for aSAH were age >72 years, history of more than once stroke, World Federation of Neurological Societies (WFNS) grades II–V, aneurysmal size >15 mm, and vertebrobasilar artery (VBA) aneurysm location. In the endovascular coiling group, scoring criteria were age >80 years, history of stroke, WFNS grades III–V, computed tomography (CT) Fisher group 4, and aneurysmal location in the middle cerebral artery (MCA) and anterior cerebral artery (ACA). The rates of poor outcome of mRS score >2 in an isolated dataset using these scoring criteria were significantly correlated with our model’s scores, so this scoring model was validated. This scoring model can help in the more objective treatment selection in patients with aSAH.

Incremental Accuracy of Blood Biomarkers for Predicting Clinical Outcomes After Intracerebral Hemorrhage

BACKGROUND

Intracerebral hemorrhage (ICH) is associated with high mortality, morbidity, and recurrence. Studies have reported the accuracy of several blood biomarkers in predicting clinical outcomes; however, their independent contribution in prediction remains to be established.

AIM

To investigate the incremental accuracy in predicting clinical outcomes in patients with ICH in a north Indian population using blood-based biomarkers.

METHODS

In this study, a total of 250 ICH cases were recruited within 72 hours of onset. Baseline clinical and CT scan measurement were recorded. Homocysteine (HCY), C-reactive protein (CRP), matrix metalloproteinase-9 (MMP9), E-selectin (SELE), and P-selectin (SELP) levels were measured through ELISA. Telephonic follow-up was done by using mRS scale at three months.

RESULTS

The mean age of cohort was 54.9 (SD±12.8) years with 64.8% patients being male. A total of 109 (43.6%) deaths were observed over three months follow-up. Area under the receiver operating characteristics curve-(AUROC) for 90-day mortality were 0.55 (HCY), 0.62 (CRP), 0.57 (MMP9), 0.60 (SELE) and 0.53 (SELP) and for poor outcome at 90-day (mRS: 3-6) were 0.60 (HCY), 0.62 (CRP), 0.54 (MMP9), 0.67 (SELE) and 0.54 (SELP). In multivariable model including age, ICH volume, IVH and GCS at admission, serum SELE (p=0.004) significant for poor outcome with improved AUROC (0.86) and HCY (p=0.04), CRP (p=0.003) & MMP9 (p=0.02) for mortality with least Akaike's Information Criterion-(AIC) (1060.5).

CONCLUSIONS

Our findings suggest that the serum SELE is a significant predictor of poor outcome and HCY, CRP & MMP9 for Mortality in patients with ICH in the north Indian population.

Predicting the Poor Recovery Risk of Aneurysmal Subarachnoid Hemorrhage: Clinical Evaluation and Management Based on a New Predictive Nomogram

PURPOSE

To develop and validate a model for identifying the risk factors of poor recovery in patients with aneurysmal subarachnoid hemorrhage (aSAH).

METHODS

A prediction model was developed using training data obtained from 1577 aSAH patients from multiple centers. The patients were followed for 6 months on average and assessed using the modified Rankin Scale; patient information was collected with a prospective case report form. The least absolute shrinkage and selection operator regression were applied to optimize factor selection for the poor recovery risk model. Multivariable logistic regression, incorporating the factors selected in the previous step, was used for model predictions. Predictive ability and clinical effectiveness of the model were evaluated using C-index, receiver operating characteristic curve, and decision curve analysis. Internal validation was performed using the C-index, taking advantage of bootstrapping validation.

RESULTS

The predictors included household income per capita, hypertension, smoking, migraine within a week before onset, Glasgow Coma Scale at admission, average blood pressure at admission, modified Fisher score at admission, treatment method, and complications. Our newly developed model made satisfactory predictions; it had a C-index of 0.796 and an area under the receiver operating characteristic curve of 0.784. The decision curve analysis showed that the poor recovery nomogram was of clinical benefit when an intervention was decided at a poor recovery threshold between 2% and 50%. Internal validation revealed a C-index of 0.760.

CONCLUSION

Our findings indicate that the novel poor recovery nomogram may be conveniently used for risk prediction in aSAH patients. For patients with intracranial aneurysms, migraine needs to be vigilant. Quitting smoking and blood pressure management are also beneficial.

Wisdom of the expert crowd prediction of response for 3 neurology randomized trials

OBJECTIVE

To explore the accuracy of combined neurology expert forecasts in predicting primary endpoints for trials.

METHODS

We identified one major randomized trial each in stroke, multiple sclerosis (MS), and amyotrophic lateral sclerosis (ALS) that was closing within 6 months. After recruiting a sample of neurology experts for each disease, we elicited forecasts for the primary endpoint outcomes in the trial placebo and treatment arms. Our main outcome was the accuracy of averaged predictions, measured using ordered Brier scores. Scores were compared against an algorithm that offered noncommittal predictions.

RESULTS

Seventy-one neurology experts participated. Combined forecasts of experts were less accurate than a noncommittal prediction algorithm for the stroke trial (pooled Brier score = 0.340, 95% subjective probability interval [sPI] 0.340 to 0.340 vs 0.185 for the uninformed prediction), and approximately as accurate for the MS study (pooled Brier score = 0.107, 95% confidence interval [CI] 0.081 to 0.133 vs 0.098 for the noncommittal prediction) and the ALS study (pooled Brier score = 0.090, 95% CI 0.081 to 0.185 vs 0.090). The 95% sPIs of individual predictions contained actual trial outcomes among 44% of experts. Only 18% showed prediction skill exceeding the noncommittal prediction. Independent experts and coinvestigators achieved similar levels of accuracy.

CONCLUSION

In this first-of-kind exploratory study, averaged expert judgments rarely outperformed noncommittal forecasts. However, experts at least anticipated the possibility of effects observed in trials. Our findings, if replicated in different trial samples, caution against the reliance on simple approaches for combining expert opinion in making research and policy decisions.

Fundamentals of Clinical Prediction Modeling for the Neurosurgeon

Clinical prediction models in neurosurgery are increasingly reported. These models aim to provide an evidence-based approach to the estimation of the probability of a neurosurgical outcome by combining 2 or more prognostic variables. Model development and model reporting are often suboptimal. A basic understanding of the methodology of clinical prediction modeling is needed when interpreting these models. We address basic statistical background, 7 modeling steps, and requirements of these models such that they may fulfill their potential for major impact for our daily clinical practice and for future scientific work.

Predictive accuracy of physicians' estimates of outcome after severe stroke

Introduction

End-of-life decisions after stroke should be guided by accurate estimates of the patient’s prognosis. We assessed the accuracy of physicians’ estimates regarding mortality, functional outcome, and quality of life in patients with severe stroke.

Methods

Treating physicians predicted mortality, functional outcome (modified Rankin scale (mRS)), and quality of life (visual analogue scale (VAS)) at six months in patients with major disabling stroke who had a Barthel Index ≤6 (of 20) at day four. Unfavorable functional outcome was defined as mRS >3, non-satisfactory quality of life as VAS <60. Patients were followed-up at six months after stroke. We compared physicians’ estimates with actual outcomes.

Results

Sixty patients were included, with a mean age of 72 years. Of fifteen patients who were predicted to die, one actually survived at six months (positive predictive value (PPV), 0.93; 95% CI, 0.66–0.99). Of thirty patients who survived, one was predicted to die (false positive rate (FPR), 0.03; 95%CI 0.00–0.20). Of forty-six patients who were predicted to have an unfavorable outcome, four had a favorable outcome (PPV, 0.93; 95% CI, 0.81–0.98; FPR, 0.30; 95% CI; 0.08–0.65). Prediction of non-satisfactory quality of life was less accurate (PPV, 0.63; 95% CI, 0.26–0.90; FPR, 0.18; 95% CI 0.05–0.44).

Conclusions

In patients with severe stroke, treating physicians’ estimation of the risk of mortality or unfavorable functional outcome at six months is relatively inaccurate. Prediction of quality of life is even more imprecise.

Family discussions on life-sustaining interventions in neurocritical care

Approximately 20% of all deaths in the USA occur in the intensive care unit (ICU) and the majority of ICU deaths involves decision of de-escalation of life-sustaining interventions. Life-sustaining interventions may include intubation and mechanical ventilation, artificial nutrition and hydration, antibiotic treatment, brain surgery, or vasoactive support. Decision making about goals of care can be defined as an end-of-life communication and the decision-making process between a clinician and a patient (or a surrogate decision maker if the patient is incapable) in an institutional setting to establish a plan of care. This process includes deciding whether to use life-sustaining treatments. Therefore, family discussion is a critical element in the decision-making process throughout the patient's stay in the neurocritical care unit. A large part of care in the neurosciences intensive care unit is discussion of proportionality of care. This chapter provides a stepwise approach to hold these conferences and discusses ways to do it effectively.

Clinician judgment vs formal scales for predicting intracerebral hemorrhage outcomes

OBJECTIVE

To compare the performance of formal prognostic instruments vs subjective clinical judgment with regards to predicting functional outcome in patients with spontaneous intracerebral hemorrhage (ICH).

METHODS

This prospective observational study enrolled 121 ICH patients hospitalized at 5 US tertiary care centers. Within 24 hours of each patient's admission to the hospital, one physician and one nurse on each patient's clinical team were each asked to predict the patient's modified Rankin Scale (mRS) score at 3 months and to indicate whether he or she would recommend comfort measures. The admission ICH score and FUNC score, 2 prognostic scales selected for their common use in neurologic practice, were calculated for each patient. Spearman rank correlation coefficients (r) with respect to patients' actual 3-month mRS for the physician and nursing predictions were compared against the same correlation coefficients for the ICH score and FUNC score.

RESULTS

The absolute value of the correlation coefficient for physician predictions with respect to actual outcome (0.75) was higher than that of either the ICH score (0.62, p = 0.057) or the FUNC score (0.56, p = 0.01). The nursing predictions of outcome (r = 0.72) also trended towards an accuracy advantage over the ICH score (p = 0.09) and FUNC score (p = 0.03). In an analysis that excluded patients for whom comfort care was recommended, the 65 available attending physician predictions retained greater accuracy (r = 0.73) than either the ICH score (r = 0.50, p = 0.02) or the FUNC score (r = 0.42, p = 0.004).

CONCLUSIONS

Early subjective clinical judgment of physicians correlates more closely with 3-month outcome after ICH than prognostic scales.

Time from onset of SIRS to antibiotic administration and outcomes after subarachnoid hemorrhage

INTRODUCTION

The interval from presentation with systemic inflammatory response syndrome (SIRS) to the start of antibiotic administration affects mortality in patients with sepsis. However, patients with subarachnoid hemorrhage (SAH) often develop SIRS directly from their brain injury, making it a less useful indicator of infection. We therefore hypothesized that SIRS would not be a suitable trigger for antibiotics in this population.

METHODS

We examined the time from the development of SIRS until antibiotic initiation and its relationship to long-term neurological outcomes in patients with nontraumatic SAH. Patients' baseline characteristics, time of antibiotic administration, and hospital course were collected from retrospective chart review. The primary outcome, 6-month functional status, was prospectively determined using blinded, structured interviews incorporating the modified Rankin Scale (mRS).

RESULTS

Sixty-six of 70 patients with SAH during the study period had 6-month follow-up and were included in this analysis. SIRS developed in 57 patients (86%, 95% CI 78-95%). In ordinal logistic regression models controlling for age and illness severity, the time from SIRS onset until antibiotic initiation was not associated with 6-month mRS scores (OR per hour, 0.994; 95% CI 0.987-1.001).

CONCLUSIONS

In this cohort of patients with SAH, time from SIRS onset until antibiotic administration was not related to functional outcomes. Our results indicate that SIRS is nonspecific in patients with SAH, and support the safety of withholding antibiotics in those who lack additional evidence of infection or hemodynamic deterioration.

Metalloporphyrins as therapeutic catalytic oxidoreductants in central nervous system disorders

SIGNIFICANCE

Metalloporphyrins, characterized by a redox-active transitional metal (Mn or Fe) coordinated to a cyclic porphyrin core ligand, mitigate oxidative/nitrosative stress in biological systems. Side-chain substitutions tune redox properties of metalloporphyrins to act as potent superoxide dismutase mimics, peroxynitrite decomposition catalysts, and redox regulators of transcription factor function. With oxidative/nitrosative stress central to pathogenesis of CNS injury, metalloporphyrins offer unique pharmacologic activity to improve the course of disease.

RECENT ADVANCES

Metalloporphyrins are efficacious in models of amyotrophic lateral sclerosis, Alzheimer's disease, epilepsy, neuropathic pain, opioid tolerance, Parkinson's disease, spinal cord injury, and stroke and have proved to be useful tools in defining roles of superoxide, nitric oxide, and peroxynitrite in disease progression. The most substantive recent advance has been the synthesis of lipophilic metalloporphyrins offering improved blood-brain barrier penetration to allow intravenous, subcutaneous, or oral treatment.

CRITICAL ISSUES

Insufficient preclinical data have accumulated to enable clinical development of metalloporphyrins for any single indication. An improved definition of mechanisms of action will facilitate preclinical modeling to define and validate optimal dosing strategies to enable appropriate clinical trial design. Due to previous failures of "antioxidants" in clinical trials, with most having markedly less biologic activity and bioavailability than current-generation metalloporphyrins, a stigma against antioxidants has discouraged the development of metalloporphyrins as CNS therapeutics, despite the consistent definition of efficacy in a wide array of CNS disorders.

FUTURE DIRECTIONS

Further definition of the metalloporphyrin mechanism of action, side-by-side comparison with "failed" antioxidants, and intense effort to optimize therapeutic dosing strategies are required to inform and encourage clinical trial design.

End-of-life decisions in patients with severe acute brain injury

Most in-hospital deaths of patients with stroke, traumatic brain injury, or postanoxic encephalopathy after cardiac arrest occur after a decision to withhold or withdraw life-sustaining treatments. Decisions on treatment restrictions in these patients are generally complex and are based only in part on evidence from published work. Prognostic models to be used in this decision-making process should have a strong discriminative power. However, for most causes of acute brain injury, prognostic models are not sufficiently accurate to serve as the sole basis of decisions to limit treatment. These decisions are also complicated because patients often do not have the capacity to communicate their preferences. Additionally, surrogate decision makers might not accurately represent the patient's preferences. Finally, in the acute stage, prediction of how a patient would adapt to a life with major disability is difficult.

Predicting future brain tissue loss from white matter connectivity disruption in ischemic stroke

BACKGROUND AND PURPOSE

The Network Modification (NeMo) Tool uses a library of brain connectivity maps from normal subjects to quantify the amount of structural connectivity loss caused by focal brain lesions. We hypothesized that the Network Modification Tool could predict remote brain tissue loss caused by poststroke loss of connectivity.

METHODS

Baseline and follow-up MRIs (10.7±7.5 months apart) from 26 patients with acute ischemic stroke (age, 74.6±14.1 years, initial National Institutes of Health Stroke Scale, 3.1±3.1) were collected. Lesion masks derived from diffusion-weighted images were superimposed on the Network Modification Tool's connectivity maps, and regional structural connectivity losses were estimated via the Change in Connectivity (ChaCo) score (ie, the percentage of tracks connecting to a given region that pass through the lesion mask). ChaCo scores were correlated with subsequent atrophy.

RESULTS

Stroke lesions' size and location varied, but they were more frequent in the left hemisphere. ChaCo scores, generally higher in regions near stroke lesions, reflected this lateralization and heterogeneity. ChaCo scores were highest in the postcentral and precentral gyri, insula, middle cingulate, thalami, putamen, caudate nuclei, and pallidum. Moderate, significant partial correlations were found between baseline ChaCo scores and measures of subsequent tissue loss (r=0.43, P=4.6×10(-9); r=0.61, P=1.4×10(-18)), correcting for the time between scans.

CONCLUSIONS

ChaCo scores varied, but the most affected regions included those with sensorimotor, perception, learning, and memory functions. Correlations between baseline ChaCo and subsequent tissue loss suggest that the Network Modification Tool could be used to identify regions most susceptible to remote degeneration from acute infarcts.

Plasma taurine as a predictor of poor outcome in patients with mild neurological deficits after aneurysmal subarachnoid hemorrhage

Object

The object of this study was to determine the relationship between plasma taurine and subarachnoid hemorrhage (SAH) outcome.

Methods

Forty patients with SAH and mild neurological deficits were included in this prospective, blinded cohort study. Plasma taurine levels were measured using high-performance liquid chromatography on admission and were correlated with patient outcomes at discharge.

Results

Twenty-five percent of the patients ultimately had a poor outcome. Plasma taurine concentrations at admission were increased (2-fold) in SAH patients with a favorable outcome and were further increased (6-fold) in those who had a poor outcome. Increased taurine levels identified patients who would be discharged with a poor outcome, with sensitivity and specificity values of approximately 80% and 100%, respectively, and positive and negative predictive values of approximately 90%. Delayed cerebral vasospasm showed an OR of 27.9 (95% CI 1.090–714.9) for a poor outcome, whereas an increased taurine concentration had an OR of 105 for a poor outcome (95% CI 8.3–1328.0, p < 0.001).

Conclusions

Increased plasma taurine concentrations on admission predict a poor outcome in SAH.

JURaSSiC: accuracy of clinician vs risk score prediction of ischemic stroke outcomes

OBJECTIVE

We compared the accuracy of clinicians and a risk score (iScore) to predict observed outcomes following an acute ischemic stroke.

METHODS

The JURaSSiC (Clinician JUdgment vs Risk Score to predict Stroke outComes) study assigned 111 clinicians with expertise in acute stroke care to predict the probability of outcomes of 5 ischemic stroke case scenarios. Cases (n = 1,415) were selected as being representative of the 10 most common clinical presentations from a pool of more than 12,000 stroke patients admitted to 12 stroke centers. The primary outcome was prediction of death or disability (modified Rankin Scale [mRS] ≥3) at discharge within the 95% confidence interval (CI) of observed outcomes. Secondary outcomes included 30-day mortality and death or institutionalization at discharge.

RESULTS

Clinicians made 1,661 predictions with overall accuracy of 16.9% for death or disability at discharge, 46.9% for 30-day mortality, and 33.1% for death or institutionalization at discharge. In contrast, 90% of the iScore-based estimates were within the 95% CI of observed outcomes. Nearly half (n = 53 of 111; 48%) of participants were unable to accurately predict the probability of the primary outcome in any of the 5 rated cases. Less than 1% (n = 1) provided accurate predictions in 4 of the 5 cases and none accurately predicted all 5 case outcomes. In multivariable analyses, the presence of patient characteristics associated with poor outcomes (mRS ≥3 or death) in previous studies (older age, high NIH Stroke Scale score, and nonlacunar subtype) were associated with more accurate clinician predictions of death at 30 days (odds ratio [OR] 2.40, 95% CI 1.57-3.67) and with a trend for more accurate predictions of death or disability at discharge (OR 1.85, 95% CI 0.99-3.46).

CONCLUSIONS

Clinicians with expertise in stroke performed poorly compared to a validated tool in predicting the outcomes of patients with an acute ischemic stroke. Use of the risk stroke outcome tool may be superior for decision-making following an acute ischemic stroke.

Trajectory of functional recovery after hospital discharge for subarachnoid hemorrhage

BACKGROUND

Although there are extensive data on long-term disability after subarachnoid hemorrhage (SAH), there are few data on the trajectory of functional recovery after hospital discharge.

METHODS

From October 2009 to April 2010, we prospectively followed consecutive patients with non-traumatic SAH discharged from a university hospital. Modified Rankin Scale (mRS) scores were calculated at discharge from chart review and at 6 months by standardized telephone interview. Good functional status was defined as a mRS score of 0-2, and poor status as an mRS score of 3-6. Descriptive statistics were used to assess the trajectory of functional recovery and determine the proportion of patients whose functional status improved from poor to good.

RESULTS

Among 52 patients with non-traumatic SAH (79 % aneurysmal) who were discharged alive, most (71 %) were discharged home. Median (IQR) mRS score was 3 (2-4) at discharge and 2 (1-2) at 6 months. Some functional recovery (any improvement in mRS score) was seen in most patients (83 %; 95 % CI, 72-93 %). Of the 28 patients with poor functional status at discharge, 16 (57 %) improved to good functional status at 6 months. All patients with Hunt-Hess grade 4 or 5 hemorrhages (n = 14) had poor functional status at discharge, but half (95 % CI, 20-80 %) recovered to a good functional status at 6 months.

CONCLUSIONS

Although our sample size is small, our findings suggest that a substantial proportion of patients with SAH who are disabled at discharge go on to regain functional independence within 6 months.