A simulation study evaluating approaches to the analysis of ordinal outcome data in randomized controlled trials in traumatic brain injury: results from the IMPACT Project

Bayesian transition models for ordinal longitudinal outcomes

Ordinal longitudinal outcomes are becoming common in clinical research, particularly in the context of COVID-19 clinical trials. These outcomes are information-rich and can increase the statistical efficiency of a study when analyzed in a principled manner. We present Bayesian ordinal transition models as a flexible modeling framework to analyze ordinal longitudinal outcomes. We develop the theory from first principles and provide an application using data from the Adaptive COVID-19 Treatment Trial (ACTT-1) with code examples in R. We advocate that researchers use ordinal transition models to analyze ordinal longitudinal outcomes when appropriate alongside standard methods such as time-to-event modeling.

Statistical analyses of ordinal outcomes in randomised controlled trials: a scoping review

BACKGROUND

Randomised controlled trials (RCTs) aim to estimate the causal effect of one or more interventions relative to a control. One type of outcome that can be of interest in an RCT is an ordinal outcome, which is useful to answer clinical questions regarding complex and evolving patient states. The target parameter of interest for an ordinal outcome depends on the research question and the assumptions the analyst is willing to make. This review aimed to provide an overview of how ordinal outcomes have been used and analysed in RCTs.

METHODS

The review included RCTs with an ordinal primary or secondary outcome published between 2017 and 2022 in four highly ranked medical journals (the British Medical Journal, New England Journal of Medicine, The Lancet, and the Journal of the American Medical Association) identified through PubMed. Details regarding the study setting, design, the target parameter, and statistical methods used to analyse the ordinal outcome were extracted.

RESULTS

The search identified 309 studies, of which 144 were eligible for inclusion. The most used target parameter was an odds ratio, reported in 78 (54%) studies. The ordinal outcome was dichotomised for analysis in 47 ( 33 % ) studies, and the most common statistical model used to analyse the ordinal outcome on the full ordinal scale was the proportional odds model (64 [ 44 % ] studies). Notably, 86 (60%) studies did not explicitly check or describe the robustness of the assumptions for the statistical method(s) used.

CONCLUSIONS

The results of this review indicate that in RCTs that use an ordinal outcome, there is variation in the target parameter and the analytical approaches used, with many dichotomising the ordinal outcome. Few studies provided assurance regarding the appropriateness of the assumptions and methods used to analyse the ordinal outcome. More guidance is needed to improve the transparent reporting of the analysis of ordinal outcomes in future trials.

Research considerations for prospective studies of patients with coma and disorders of consciousness

Disorders of consciousness are neurological conditions characterized by impaired arousal and awareness of self and environment. Behavioural responses are absent or are present but fluctuate. Disorders of consciousness are commonly encountered as a consequence of both acute and chronic brain injuries, yet reliable epidemiological estimates would require inclusive, operational definitions of the concept, as well as wider knowledge dissemination among involved professionals. Whereas several manifestations have been described, including coma, vegetative state/unresponsive wakefulness syndrome and minimally conscious state, a comprehensive neurobiological definition for disorders of consciousness is still lacking. The scientific literature is primarily observational, and studies-specific aetiologies lead to disorders of consciousness. Despite advances in these disease-related forms, there remains uncertainty about whether disorders of consciousness are a disease-agnostic unitary entity with a common mechanism, prognosis or treatment response paradigm. Our knowledge of disorders of consciousness has also been hampered by heterogeneity of study designs, variables, and outcomes, leading to results that are not comparable for evidence synthesis. The different backgrounds of professionals caring for patients with disorders of consciousness and the different goals at different stages of care could partly explain this variability. The Prospective Studies working group of the Neurocritical Care Society Curing Coma Campaign was established to create a platform for observational studies and future clinical trials on disorders of consciousness and coma across the continuum of care. In this narrative review, the author panel presents limitations of prior observational clinical research and outlines practical considerations for future investigations. A narrative review format was selected to ensure that the full breadth of study design considerations could be addressed and to facilitate a future consensus-based statement (e.g. via a modified Delphi) and series of recommendations. The panel convened weekly online meetings from October 2021 to December 2022. Research considerations addressed the nosographic status of disorders of consciousness, case ascertainment and verification, selection of dependent variables, choice of covariates and measurement and analysis of outcomes and covariates, aiming to promote more homogeneous designs and practices in future observational studies. The goal of this review is to inform a broad community of professionals with different backgrounds and clinical interests to address the methodological challenges imposed by the transition of care from acute to chronic stages and to streamline data gathering for patients with disorders of consciousness. A coordinated effort will be a key to allow reliable observational data synthesis and epidemiological estimates and ultimately inform condition-modifying clinical trials.

Selection of a statistical analysis method for the Glasgow Outcome Scale-Extended endpoint for estimating the probability of favorable outcome in future severe TBI clinical trials

The Glasgow outcome scale-extended (GOS-E), an ordinal scale measure, is often selected as the endpoint for clinical trials of traumatic brain injury (TBI). Traditionally, GOS-E is analyzed as a fixed dichotomy with favorable outcome defined as GOS-E ≥ 5 and unfavorable outcome as GOS-E < 5. More recent studies have defined favorable vs unfavorable outcome utilizing a sliding dichotomy of the GOS-E that defines a favorable outcome as better than a subject's predicted prognosis at baseline. Both dichotomous approaches result in loss of statistical and clinical information. To improve on power, Yeatts et al proposed a sliding scoring of the GOS-E as the distance from the cutoff for favorable/unfavorable outcomes, and therefore used more information found in the original GOS-E to estimate the probability of favorable outcome. We used data from a published TBI trial to explore the ramifications to trial operating characteristics by analyzing the sliding scoring of the GOS-E as either dichotomous, continuous, or ordinal. We illustrated a connection between the ordinal data and time-to-event (TTE) data to allow use of Bayesian software that utilizes TTE-based modeling. The simulation results showed that the continuous method with continuity correction offers higher power and lower mean squared error for estimating the probability of favorable outcome compared to the dichotomous method, and similar power but higher precision compared to the ordinal method. Therefore, we recommended that future severe TBI clinical trials consider analyzing the sliding scoring of the GOS-E endpoint as continuous with continuity correction.

Using metabolomics to predict severe traumatic brain injury outcome (GOSE) at 3 and 12 months

BACKGROUND

Prognostication is very important to clinicians and families during the early management of severe traumatic brain injury (sTBI), however, there are no gold standard biomarkers to determine prognosis in sTBI. As has been demonstrated in several diseases, early measurement of serum metabolomic profiles can be used as sensitive and specific biomarkers to predict outcomes.

METHODS

We prospectively enrolled 59 adults with sTBI (Glasgow coma scale, GCS ≤ 8) in a multicenter Canadian TBI (CanTBI) study. Serum samples were drawn for metabolomic profiling on the 1st and 4th days following injury. The Glasgow outcome scale extended (GOSE) was collected at 3- and 12-months post-injury. Targeted direct infusion liquid chromatography-tandem mass spectrometry (DI/LC-MS/MS) and untargeted proton nuclear magnetic resonance spectroscopy (¹H-NMR) were used to profile serum metabolites. Multivariate analysis was used to determine the association between serum metabolomics and GOSE, dichotomized into favorable (GOSE 5-8) and unfavorable (GOSE 1-4), outcomes.

RESULTS

Serum metabolic profiles on days 1 and 4 post-injury were highly predictive (Q² > 0.4-0.5) and highly accurate (AUC > 0.99) to predict GOSE outcome at 3- and 12-months post-injury and mortality at 3 months. The metabolic profiles on day 4 were more predictive (Q² > 0.55) than those measured on day 1 post-injury. Unfavorable outcomes were associated with considerable metabolite changes from day 1 to day 4 compared to favorable outcomes. Increased lysophosphatidylcholines, acylcarnitines, energy-related metabolites (glucose, lactate), aromatic amino acids, and glutamate were associated with poor outcomes and mortality.

DISCUSSION

Metabolomic profiles were strongly associated with the prognosis of GOSE outcome at 3 and 12 months and mortality following sTBI in adults. The metabolic phenotypes on day 4 post-injury were more predictive and significant for predicting the sTBI outcome compared to the day 1 sample. This may reflect the larger contribution of secondary brain injury (day 4) to sTBI outcome. Patients with unfavorable outcomes demonstrated more metabolite changes from day 1 to day 4 post-injury. These findings highlighted increased concentration of neurobiomarkers such as N-acetylaspartate (NAA) and tyrosine, decreased concentrations of ketone bodies, and decreased urea cycle metabolites on day 4 presenting potential metabolites to predict the outcome. The current findings strongly support the use of serum metabolomics, that are shown to be better than clinical data, in determining prognosis in adults with sTBI in the early days post-injury. Our findings, however, require validation in a larger cohort of adults with sTBI to be used for clinical practice.

AutoScore-Ordinal: an interpretable machine learning framework for generating scoring models for ordinal outcomes

Background

Risk prediction models are useful tools in clinical decision-making which help with risk stratification and resource allocations and may lead to a better health care for patients. AutoScore is a machine learning–based automatic clinical score generator for binary outcomes. This study aims to expand the AutoScore framework to provide a tool for interpretable risk prediction for ordinal outcomes.

Methods

The AutoScore-Ordinal framework is generated using the same 6 modules of the original AutoScore algorithm including variable ranking, variable transformation, score derivation (from proportional odds models), model selection, score fine-tuning, and model evaluation. To illustrate the AutoScore-Ordinal performance, the method was conducted on electronic health records data from the emergency department at Singapore General Hospital over 2008 to 2017. The model was trained on 70% of the data, validated on 10% and tested on the remaining 20%.

Results

This study included 445,989 inpatient cases, where the distribution of the ordinal outcome was 80.7% alive without 30-day readmission, 12.5% alive with 30-day readmission, and 6.8% died inpatient or by day 30 post discharge. Two point-based risk prediction models were developed using two sets of 8 predictor variables identified by the flexible variable selection procedure. The two models indicated reasonably good performance measured by mean area under the receiver operating characteristic curve (0.758 and 0.793) and generalized c-index (0.737 and 0.760), which were comparable to alternative models.

Conclusion

AutoScore-Ordinal provides an automated and easy-to-use framework for development and validation of risk prediction models for ordinal outcomes, which can systematically identify potential predictors from high-dimensional data.

Interleukin-6 as a prognostic biomarker of clinical outcomes after traumatic brain injury: a systematic review

Traumatic brain injury (TBI) is a major cause of mortality and morbidity worldwide. There are currently no early biomarkers for prognosis in routine clinical use. Interleukin-6 (IL-6) is a potential biomarker in the context of the established role of neuroinflammation in TBI recovery. Therefore, a systematic review of the literature was performed to assess and summarise the evidence for IL-6 secretion representing a useful biomarker for clinical outcomes. A multi-database literature search between January 1946 and July 2021 was performed. Studies were included if they reported adult TBI patients with IL-6 concentration in serum, cerebrospinal fluid (CSF) and/or brain parenchyma analysed with respect to functional outcome and/or mortality. A synthesis without meta-analysis is reported. Fifteen studies were included, reporting 699 patients. Most patients were male (71.7%), and the pooled mean age was 40.8 years; 78.1% sustained severe TBI. Eleven studies reported IL-6 levels in serum, six in CSF and one in the parenchyma. Five studies on serum demonstrated higher IL-6 concentrations were associated with poorer outcomes, and five showed no signification association. In CSF studies, one found higher IL-6 levels were associated with poorer outcomes, one found them to predict better outcomes and three found no association. Greater parenchymal IL-6 was associated with better outcomes. Despite some inconsistency in findings, it appears that exaggerated IL-6 secretion predicts poor outcomes after TBI. Future efforts require standardisation of IL-6 measurement practices as well as assessment of the importance of IL-6 concentration dynamics with respect to clinical outcomes, ideally within large prospective studies. Prospero registration number: CRD42021271200

Tenets of Good Practice in Regression Analysis. A Brief Tutorial

BACKGROUND

Regression analysis quantifies the relationships between one or more independent variables and a dependent variable and is one of the most frequently used types of analysis in medical research. The aim of this article is to provide a brief theoretical and practical tutorial for neurosurgeons wishing to conduct or interpret regression analyses.

METHODS AND RESULTS

Data preparation, univariable and multivariable analysis, choice of model, model requirements and assumptions are discussed, as essential prerequisites to any regression analysis. Four main types of regression techniques are presented: linear, logistic, multinomial logistic, and proportional odds logistic. To illustrate the applications of regression to real-world data and exemplify the concepts introduced, we used a previously reported data set of patients with intracranial aneurysms treated by microsurgical clip reconstruction at the Department of Neurosurgery of Erasmus MC University Medical Center Rotterdam, between January 2000 and January 2019.

CONCLUSIONS

Regression analysis is a powerful and versatile instrument in data analysis. This material is intended as a starter for those wishing to critically interpret or perform regression analysis and we recommend multidisciplinary collaborations with trained methodologists, statisticians, or epidemiologists.

The influence of cognitive reserve in the protection of the cognitive status after an acquired brain injury: A systematic review

BACKGROUND

Cognitive Reserve (CR) hypothesis was introduced to account for the variability in cognitive performance of patients with similar degrees of brain injury or pathology. The individual variability of CR is modulated by the interaction of innate capacities and exposures throughout life, which can act as protectors against neuropathology's clinical effects. Individuals with higher CR appear to have better cognitive performance after a brain injury. The present review aimed to identify and map the scientific evidence available in literature regarding CR's influence in protecting the cognitive status after an Acquired Brain Injury (ABI).

METHOD

A systematic review was performed for published studies until October 2020 in PubMed, Scopus, and CINAHL electronic databases. Studies regarding CR's influence in protecting the cognitive status after an ABI were included in this review. The Newcastle-Ottawa Scale was used to assess risk of bias in the included studies. This systematic review was recorded in the International Prospective Register of Systematic Reviews (PROSPERO) under the number CRD42021236594.

RESULTS

Twenty-one studies published between 2003 and 2020 were selected and analyzed. The literature analysis showed that CR has a positive effect on cognitive status after an ABI. Various proxies were used to estimate CR, including estimated premorbid IQ, education, occupation attainment, socioeconomic status, leisure activities, bilingualism, and social integration. CR proxies constitute a set of variables that may have a significant influence on cognitive status. Higher CR levels were associated with lower cognitive impairment after an ABI.

CONCLUSIONS

Although more research is necessary for a complete understanding of CR's impact on cognition, the synthesis of these studies confirmed that there is evidence on the beneficial impact of CR on cognitive status after an ABI. These findings support CR's cognitive status role following an ABI and may provide additional information for prognosis and rehabilitation plans.

Pretreatment neutrophil-to-lymphocyte/monocyte-to-lymphocyte ratio as prognostic biomarkers in glioma patients

OBJECTIVES

To evaluate the ability for pre-treatment NLR and MLR to predict overall survival (OS) and modified Rankin Scale (mRS) and to explore their relationship with clinicopathological parameters.

METHODS

Retrospective analysis of pretreatment NLR and MLR from 64 glioma patients.

RESULTS

Higher pretreatment NLR (>4.7) predicted higher mean admission mRS (p < 0.001) and 6-month mRS (p = 0.02). Higher pretreatment MLR (>0.35) was a risk factor for poorer OS in glioma patients (p = 0.024). Higher pretreatment NLR was significantly associated with larger tumor diameter (p = 0.02).

CONCLUSION

NLR and MLR can serve as prognostic markers to predict functional outcomes and OS in glioma patients.

Risk prediction models for discrete ordinal outcomes: Calibration and the impact of the proportional odds assumption

Calibration is a vital aspect of the performance of risk prediction models, but research in the context of ordinal outcomes is scarce. This study compared calibration measures for risk models predicting a discrete ordinal outcome, and investigated the impact of the proportional odds assumption on calibration and overfitting. We studied the multinomial, cumulative, adjacent category, continuation ratio, and stereotype logit/logistic models. To assess calibration, we investigated calibration intercepts and slopes, calibration plots, and the estimated calibration index. Using large sample simulations, we studied the performance of models for risk estimation under various conditions, assuming that the true model has either a multinomial logistic form or a cumulative logit proportional odds form. Small sample simulations were used to compare the tendency for overfitting between models. As a case study, we developed models to diagnose the degree of coronary artery disease (five categories) in symptomatic patients. When the true model was multinomial logistic, proportional odds models often yielded poor risk estimates, with calibration slopes deviating considerably from unity even on large model development datasets. The stereotype logistic model improved the calibration slope, but still provided biased risk estimates for individual patients. When the true model had a cumulative logit proportional odds form, multinomial logistic regression provided biased risk estimates, although these biases were modest. Nonproportional odds models require more parameters to be estimated from the data, and hence suffered more from overfitting. Despite larger sample size requirements, we generally recommend multinomial logistic regression for risk prediction modeling of discrete ordinal outcomes.

ICU delirium burden predicts functional neurologic outcomes

Background

We investigated the effect of delirium burden in mechanically ventilated patients, beginning in the ICU and continuing throughout hospitalization, on functional neurologic outcomes up to 2.5 years following critical illness.

Methods

Prospective cohort study of enrolling 178 consecutive mechanically ventilated adult medical and surgical ICU patients between October 2013 and May 2016. Altogether, patients were assessed daily for delirium 2941days using the Confusion Assessment Method for the ICU (CAM-ICU). Hospitalization delirium burden (DB) was quantified as number of hospital days with delirium divided by total days at risk. Survival status up to 2.5 years and neurologic outcomes using the Glasgow Outcome Scale were recorded at discharge 3, 6, and 12 months post-discharge.

Results

Of 178 patients, 19 (10.7%) were excluded from outcome analyses due to persistent coma. Among the remaining 159, 123 (77.4%) experienced delirium. DB was independently associated with >4-fold increased mortality at 2.5 years following ICU admission (adjusted hazard ratio [aHR], 4.77; 95% CI, 2.10–10.83; P < .001), and worse neurologic outcome at discharge (adjusted odds ratio [aOR], 0.02; 0.01–0.09; P < .001), 3 (aOR, 0.11; 0.04–0.31; P < .001), 6 (aOR, 0.10; 0.04–0.29; P < .001), and 12 months (aOR, 0.19; 0.07–0.52; P = .001). DB in the ICU alone was not associated with mortality (HR, 1.79; 0.93–3.44; P = .082) and predicted neurologic outcome less strongly than entire hospital stay DB. Similarly, the number of delirium days in the ICU and for whole hospitalization were not associated with mortality (HR, 1.00; 0.93–1.08; P = .917 and HR, 0.98; 0.94–1.03, P = .535) nor with neurological outcomes, except for the association between ICU delirium days and neurological outcome at discharge (OR, 0.90; 0.81–0.99, P = .038).

Conclusions

Delirium burden throughout hospitalization independently predicts long term neurologic outcomes and death up to 2.5 years after critical illness, and is more predictive than delirium burden in the ICU alone and number of delirium days.

Ordinal Prediction Model of 90-Day Modified Rankin Scale in Ischemic Stroke

Background and Purpose: Prediction models for functional outcomes after ischemic stroke are useful for statistical analyses in clinical trials and guiding patient expectations. While there are models predicting dichotomous functional outcomes after ischemic stroke, there are no models that predict ordinal mRS outcomes. We aimed to create a model that predicts, at the time of hospital discharge, a patient's modified Rankin Scale (mRS) score on day 90 after ischemic stroke.

Methods: We used data from three multi-center prospective studies: CRISP, DEFUSE 2, and DEFUSE 3 to derive and validate an ordinal logistic regression model that predicts the 90-day mRS score based on variables available during the stroke hospitalization. Forward selection was used to retain independent significant variables in the multivariable model.

Results: The prediction model was derived using data on 297 stroke patients from the CRISP and DEFUSE 2 studies. National Institutes of Health Stroke Scale (NIHSS) at discharge and age were retained as significant (p < 0.001) independent predictors of the 90-day mRS score. When applied to the external validation set (DEFUSE 3, n = 160), the model accurately predicted the 90-day mRS score within one point for 78% of the patients in the validation cohort.

Conclusions: A simple model using age and NIHSS score at time of discharge can predict 90-day mRS scores in patients with ischemic stroke. This model can be useful for prognostication in routine clinical care and to impute missing data in clinical trials.

Fit-for-Discharge Criteria after Esophagectomy: An International Expert Delphi Consensus

There are no internationally recognized criteria available to determine preparedness for hospital discharge after esophagectomy. This study aims to achieve international consensus using Delphi methodology. The expert panel consisted of 40 esophageal surgeons spanning 16 countries and 4 continents. During a 3-round, web-based Delphi process, experts voted for discharge criteria using 5-point Likert scales. Data were analyzed using descriptive statistics. Consensus was reached if agreement was ≥75% in round 3. Consensus was achieved for the following basic criteria: nutritional requirements are met by oral intake of at least liquids with optional supplementary nutrition via jejunal feeding tube. The patient should have passed flatus and does not require oxygen during mobilization or at rest. Central venous catheters should be removed. Adequate analgesia at rest and during mobilization is achieved using both oral opioid and non-opioid analgesics. All vital signs should be normal unless abnormal preoperatively. Inflammatory parameters should be trending down and close to normal (leucocyte count ≤12G/l and C-reactive protein ≤80 mg/dl). This multinational Delphi survey represents the first expert-led process for consensus criteria to determine 'fit-for-discharge' status after esophagectomy. Results of this Delphi survey may be applied to clinical outcomes research as an objective measure of short-term recovery. Furthermore, standardized endpoints identified through this process may be used in clinical practice to guide decisions regarding patient discharge and may help to reduce the risk of premature discharge or prolonged admission.

Ordinal outcome analysis improves the detection of between-hospital differences in outcome

Background

There is a growing interest in assessment of the quality of hospital care, based on outcome measures. Many quality of care comparisons rely on binary outcomes, for example mortality rates. Due to low numbers, the observed differences in outcome are partly subject to chance. We aimed to quantify the gain in efficiency by ordinal instead of binary outcome analyses for hospital comparisons. We analyzed patients with traumatic brain injury (TBI) and stroke as examples.

Methods

We sampled patients from two trials. We simulated ordinal and dichotomous outcomes based on the modified Rankin Scale (stroke) and Glasgow Outcome Scale (TBI) in scenarios with and without true differences between hospitals in outcome. The potential efficiency gain of ordinal outcomes, analyzed with ordinal logistic regression, compared to dichotomous outcomes, analyzed with binary logistic regression was expressed as the possible reduction in sample size while keeping the same statistical power to detect outliers.

Results

In the IMPACT study (9578 patients in 265 hospitals, mean number of patients per hospital = 36), the analysis of the ordinal scale rather than the dichotomized scale (‘unfavorable outcome’), allowed for up to 32% less patients in the analysis without a loss of power. In the PRACTISE trial (1657 patients in 12 hospitals, mean number of patients per hospital = 138), ordinal analysis allowed for 13% less patients. Compared to mortality, ordinal outcome analyses allowed for up to 37 to 63% less patients.

Conclusions

Ordinal analyses provide the statistical power of substantially larger studies which have been analyzed with dichotomization of endpoints. We advise to exploit ordinal outcome measures for hospital comparisons, in order to increase efficiency in quality of care measurements.

Trial registration

We do not report the results of a health care intervention.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12874-020-01185-7.

CompLement C5 Antibodies for decreasing brain injury after aneurysmal Subarachnoid Haemorrhage (CLASH): study protocol for a randomised controlled phase II clinical trial

Background

The inflammatory response after aneurysmal subarachnoid haemorrhage (aSAH) has been associated with early brain injury, delayed cerebral ischaemia, poor functional outcome, and case fatality. In experimental SAH studies, complement C5 antibodies administered shortly after SAH reduced brain injury with approximately 40%. Complement component C5 may be a new therapeutic target to reduce brain injury and hereby improve the outcome after aSAH. We aim to investigate the pharmacodynamic efficacy and safety of eculizumab (complement C5 antibody) in patients with aSAH.

Methods

A randomised, controlled, open-label, phase II clinical trial with blinded outcome assessment. Eculizumab (1200 mg) is administered intravenously < 12 h, on day 3 and on day 7 after ictus. Patients in the intervention group receive prophylactic antibiotics for 4 weeks, and those with a central line or an external ventricular shunt and a positive fungal or yeast culture also receive prophylactic antifungal therapy for 4 weeks. The primary outcome is C5a concentration in the cerebrospinal fluid (CSF) on day 3 after ictus. Secondary outcomes include the occurrence of adverse events, inflammatory parameters in the blood and CSF, cerebral infarction on magnetic resonance imaging, and clinical and cognitive outcomes. We aim to evaluate 26 patients with CSF assessments, 13 in the intervention group and 13 in the comparator group. To compensate for early case fatality and inability to obtain CSF, we will include 20 patients per group.

Discussion

The CLASH trial is the first trial to investigate the pharmacodynamic efficacy and safety of eculizumab in the early phase after aSAH.

Trial registration

Netherlands Trial Register NTR6752. Registered on 27 October 2017

European Clinical Trials Database (EudraCT) 2017-004307-51

Randomized, placebo-controlled, double-blind, pilot trial to investigate safety and efficacy of Cerebrolysin in patients with aneurysmal subarachnoid hemorrhage

Background

There are limited neuroprotective treatment options for patients with aneurysmal subarachnoid hemorrhage (SAH). Cerebrolysin, a brain-specific proposed pleiotropic neuroprotective agent, has been suggested to improve global functional outcomes in ischemic stroke. We investigated the efficacy, safety and feasibility of administering Cerebrolysin for SAH patients.

Methods

This was a prospective, randomized, double-blind, placebo-controlled, single-center, parallel-group pilot study. Fifty patients received either daily Cerebrolysin (30 ml/day) or a placebo (saline) for 14 days (25 patients per study group). The primary endpoint was a favorable Extended Glasgow Outcome Scale (GOSE) of 5 to 8 (moderate disability to good recovery) at six-months. Secondary endpoints included the modified Ranking Scale (mRS), the Montreal Cognitive Assessment (MOCA) score, occurrence of adverse effects and the occurrence of delayed cerebral ischemia (DCI).

Results

No severe adverse effects or mortality attributable to Cerebrolysin were observed. No significant difference was detected in the proportion of patients with favorable six-month GOSE in either study group (odds ratio (OR): 1.49; 95% confidence interval (CI): 0.43–5.17). Secondary functional outcome measures for favorable six-month recovery i.e. a mRS of 0 to 3 (OR: 3.45; 95% CI 0.79–15.01) were comparable for both groups. Similarly, there was no difference in MOCA neurocognitive performance (p-value: 0.75) and in the incidence of DCI (OR: 0.85 95% CI: 0.28–2.59).

Conclusions

Use of Cerebrolysin in addition to standard-of-care management of aneurysmal SAH is safe, well tolerated and feasible. However, the neutral results of this trial suggest that it does not improve the six-month global functional performance of patients.

Clinical trial registration

Name of Registry: ClinicalTrials.gov

Trial Registration Number: NCT01787123.

Date of Registration: 8th February 2013.

Sliding Scoring of the Glasgow Outcome Scale-Extended as Primary Outcome in Traumatic Brain Injury Trials

The Glasgow Outcome Scale-Extended (GOS-E), an ordinal scale measuring global outcome, is used commonly as the primary outcome measure in clinical trials of traumatic brain injury. Analysis is often based on a dichotomization and thus has inherent statistical limitations, including loss of information related to the collapse of adjacent categories. A fixed dichotomization defines favorable outcome consistently for all subjects, whereas a sliding dichotomy tailors the definition of favorable outcome according to baseline prognosis/severity. Literature indicates that the sliding dichotomy is more statistically efficient than the fixed dichotomy; however, the sliding dichotomy still collapses categories and therefore discards information. We propose an alternative, a sliding scoring system for the GOS-E, intended to address the limitations of the sliding dichotomy. The score is assigned based on the number of levels between the achieved score and the favorable cut-point. The proposed scoring system reflects the magnitude of change, where change is defined according to each subject's baseline prognosis. Because the score is approximately continuous, statistical methods can rely on the normal distribution, both for analysis and study design. Two examples show the corresponding potential for improved power. A sliding score approach allows for quantification of the magnitude of change while still accounting for prognosis. Scientific advantages include increased power and an intuitive interpretation.

Prognostic Validation of the NINDS Common Data Elements for the Radiologic Reporting of Acute Traumatic Brain Injuries: A CENTER-TBI Study

The aim of this study is to investigate the prognostic value of using the National Institute of Neurological Disorders and Stroke (NINDS) standardized imaging-based pathoanatomic descriptors for the evaluation and reporting of acute traumatic brain injury (TBI) lesions. For a total of 3392 patients (2244 males and 1148 females, median age = 51 years) enrolled in the Collaborative European NeuroTrauma Effectiveness Research in Traumatic Brain Injury (CENTER-TBI) study, we extracted 96 Common Data Elements (CDEs) from the structured reports, spanning all three levels of pathoanatomic information (i.e., 20 "basic," 60 "descriptive," and 16 "advanced" CDE variables per patient). Six-month clinical outcome scores were dichotomized into favorable (Glasgow Outcome Scale Extended [GOS-E] = 5-8) versus unfavorable (GOS-E = 1-4). Regularized logistic regression models were constructed and compared using the optimism-corrected area under the curve (AUC). An abnormality was reported for the majority of patients (64.51%). In 79.11% of those patients, there was at least one coexisting pathoanatomic lesion or associated finding. An increase in lesion severity, laterality, and volume was associated with more unfavorable outcomes. Compared with the full set of pathoanatomic descriptors (i.e., all three categories of information), reporting "basic" CDE information provides at least equal discrimination between patients with favorable versus unfavorable outcome (AUC = 0.8121 vs. 0.8155, respectively). Addition of a selected subset of "descriptive" detail to the basic CDEs could improve outcome prediction (AUC = 0.8248). Addition of "advanced" or "emerging/exploratory" information had minimal prognostic value. Our results show that the NINDS standardized-imaging based pathoanatomic descriptors can be used in large-scale studies and provide important insights into acute TBI lesion patterns. When used in clinical predictive models, they can provide excellent discrimination between patients with favorable and unfavorable 6-month outcomes. If further validated, our findings could support the development of structured and itemized templates in routine clinical radiology.

Multivariable and Bayesian Network Analysis of Outcome Predictors in Acute Aneurysmal Subarachnoid Hemorrhage: Review of a Pure Surgical Series in the Post-International Subarachnoid Aneurysm Trial Era

BACKGROUND

Following the International Subarachnoid Aneurysm Trial (ISAT), evolving treatment modalities for acute aneurysmal subarachnoid hemorrhage (aSAH) has changed the case mix of patients undergoing urgent surgical clipping.

OBJECTIVE

To update our knowledge on outcome predictors by analyzing admission parameters in a pure surgical series using variable importance ranking and machine learning.

METHODS

We reviewed a single surgeon's case series of 226 patients suffering from aSAH treated with urgent surgical clipping. Predictions were made using logistic regression models, and predictive performance was assessed using areas under the receiver operating curve (AUC). We established variable importance ranking using partial Nagelkerke R² scores. Probabilistic associations between variables were depicted using Bayesian networks, a method of machine learning.

RESULTS

Importance ranking showed that World Federation of Neurosurgical Societies (WFNS) grade and age were the most influential outcome prognosticators. Inclusion of only these 2 predictors was sufficient to maintain model performance compared to when all variables were considered (AUC = 0.8222, 95% confidence interval (CI): 0.7646-0.88 vs 0.8218, 95% CI: 0.7616-0.8821, respectively, DeLong's P = .992). Bayesian networks showed that age and WFNS grade were associated with several variables such as laboratory results and cardiorespiratory parameters.

CONCLUSION

Our study is the first to report early outcomes and formal predictor importance ranking following aSAH in a post-ISAT surgical case series. Models showed good predictive power with fewer relevant predictors than in similar size series. Bayesian networks proved to be a powerful tool in visualizing the widespread association of the 2 key predictors with admission variables, explaining their importance and demonstrating the potential for hypothesis generation.

Escitalopram for agitation in Alzheimer's disease (S-CitAD): Methods and design of an investigator-initiated, randomized, controlled, multicenter clinical trial

INTRODUCTION

Alzheimer's disease (AD) is a disabling, common cause of dementia, and agitation is one of the most common and distressing symptoms for patients with AD. Escitalopram for agitation in Alzheimer's disease (S-CitAD) tests a novel, clinically derived therapeutic approach to treat agitation in patients with AD.

METHODS

S-CitAD is a NIH-funded, investigator-initiated, randomized, multicenter clinical trial. Participants receive a structured psychosocial intervention (PSI) as standard of care. Participants without sufficient response to PSI are randomized to receive 15 mg escitalopram/day or a matching placebo in addition to PSI. Primary outcome is the Modified Alzheimer's Disease Cooperative Study - Clinical Global Impression of Change (mADCS-CGIC).

DISCUSSION

S-CitAD will provide information about a practical, immediately available approach to treating agitation in patients with AD. S-CitAD may become a model of how to evaluate and predict treatment response in patients with AD and agitation as a neuropsychiatric symptom (ClinicalTrials.gov Identifier: NCT03108846).

Effect size measures and their benchmark values for quantifying benefit or risk of medicinal products

The standardized mean difference is a well-known effect size measure for continuous, normally distributed data. In this paper we present a general basis for important other distribution families. As a general concept, usable for every distribution family, we introduce the relative effect, also called Mann-Whitney effect size measure of stochastic superiority. This measure is a truly robust measure, needing no assumptions about a distribution family. It is thus the preferred tool for assumption-free, confirmatory studies. For normal distribution shift, proportional odds, and proportional hazards, we show how to derive many global values such as risk difference average, risk difference extremum, and odds ratio extremum. We demonstrate that the well-known benchmark values of Cohen with respect to group differences-small, medium, large-can be translated easily into corresponding Mann-Whitney values. From these, we get benchmarks for parameters of other distribution families. Furthermore, it is shown that local measures based on binary data (2 × 2 tables) can be associated with the Mann-Whitney measure: The concept of stochastic superiority can always be used. It is a general statistical value in every distribution family. It therefore yields a procedure for standardizing the assessment of effect size measures. We look at the aspect of relevance of an effect size and-introducing confidence intervals-present some examples for use in statistical practice.

Ranking hospital performance based on individual indicators: can we increase reliability by creating composite indicators?

BACKGROUND

Report cards on the health care system increasingly report provider-specific performance on indicators that measure the quality of health care delivered. A natural reaction to the publishing of hospital-specific performance on a given indicator is to create 'league tables' that rank hospitals according to their performance. However, many indicators have been shown to have low to moderate rankability, meaning that they cannot be used to accurately rank hospitals. Our objective was to define conditions for improving the ability to rank hospitals by combining several binary indicators with low to moderate rankability.

METHODS

Monte Carlo simulations to examine the rankability of composite ordinal indicators created by pooling three binary indicators with low to moderate rankability. We considered scenarios in which the prevalences of the three binary indicators were 0.05, 0.10, and 0.25 and the within-hospital correlation between these indicators varied between - 0.25 and 0.90.

RESULTS

Creation of an ordinal indicator with high rankability was possible when the three component binary indicators were strongly correlated with one another (the within-hospital correlation in indicators was at least 0.5). When the binary indicators were independent or weakly correlated with one another (the within-hospital correlation in indicators was less than 0.5), the rankability of the composite ordinal indicator was often less than at least one of its binary components. The rankability of the composite indicator was most affected by the rankability of the most prevalent indicator and the magnitude of the within-hospital correlation between the indicators.

CONCLUSIONS

Pooling highly-correlated binary indicators can result in a composite ordinal indicator with high rankability. Otherwise, the composite ordinal indicator may have lower rankability than some of its constituent components. It is recommended that binary indicators be combined to increase rankability only if they represent the same concept of quality of care.

Comparison of an ordinal endpoint to time-to-event, longitudinal, and binary endpoints for use in evaluating treatments for severe influenza requiring hospitalization

Background/aims

The Food and Drug Administration recommends research into developing well-defined and reliable endpoints to evaluate treatments for severe influenza requiring hospitalization. A novel 6-category ordinal endpoint of patient health status after 7 days that ranges from death to hospital discharge with resumption of normal activities is being used in a randomized placebo-controlled trial of intravenous immunoglobulin (IVIG) for severe influenza (FLU-IVIG). We compare the power of the ordinal endpoint under a proportional odds model to other types of endpoints as a function of various trial parameters.

Methods

We used closed-form analysis and empirical simulation to compare the power of the ordinal endpoint to time-to-event, longitudinal, and binary endpoints. In the simulation setting, we varied the treatment effect and the distribution of the placebo group across the follow-up period with consideration of adjustment for baseline health status.

Results

In the analytic setting, ordinal endpoints of high granularity provided greater power than time-to-event endpoints when most patients in the placebo group had either naturally progressed to the category of hospital discharge by day 7 or were far from hospital discharge on day 7. In the simulation setting, adjustment for baseline health status universally raised power for the proportional odds model. Across different placebo group distributions of the ordinal endpoint regardless of adjustment for baseline health status, only time-to-event endpoints yielded higher power than the ordinal endpoint for certain treatment effects.

Conclusions

In this case study, the FLU-IVIG ordinal endpoint provided greater power than time-to-event, binary, and longitudinal endpoints for most scenarios of the treatment effect and placebo group distribution, including the target population studied for FLU-IVIG. The ordinal endpoint was only surpassed by the time-to-event endpoint when many patients in the placebo group were on the cusp of hospital discharge on day 7 and the follow-up period for the time-to-event endpoint was extended to allow for additional events. Our general approach for evaluating the power of several potential endpoints for an influenza trial can be used for designing other influenza trials with different target populations and for other trials in other disease areas.

Efficient design and analysis of randomized controlled trials in rare neurological diseases: An example in Guillain-Barré syndrome

BACKGROUND

Randomized controlled trials (RCTs) pose specific challenges in rare and heterogeneous neurological diseases due to the small numbers of patients and heterogeneity in disease course. Two analytical approaches have been proposed to optimally handle these issues in RCTs: covariate adjustment and ordinal analysis. We investigated the potential gain in efficiency of these approaches in rare and heterogeneous neurological diseases, using Guillain-Barré syndrome (GBS) as an example.

METHODS

We analyzed two published GBS trials with primary outcome 'at least one grade improvement' on the GBS disability scale. We estimated the treatment effect using logistic regression models with and without adjustment for prognostic factors. The difference between the unadjusted and adjusted estimates was disentangled in imbalance (random differences in baseline covariates between treatment arms) and stratification (change of the estimate due to covariate adjustment). Second, we applied proportional odds regression, which exploits the ordinal nature of the GBS disability score. The standard error of the estimated treatment effect indicated the statistical efficiency.

RESULTS

Both trials were slightly imbalanced with respect to baseline characteristics, which was corrected in the adjusted analysis. Covariate adjustment increased the estimated treatment effect in the two trials by 8% and 18% respectively. Proportional odds analysis resulted in lower standard errors indicating more statistical power.

CONCLUSION

Covariate adjustment and proportional odds analysis most efficiently use the available data and ensure balance between the treatment arms to obtain reliable and valid treatment effect estimates. These approaches merit application in future trials in rare and heterogeneous neurological diseases like GBS.

Final outcome trends in severe traumatic brain injury: a 25-year analysis of single center data

BACKGROUND

Evidence from the last 25 years indicates a modest reduction of mortality after severe traumatic head injury (sTBI). This study evaluates the variation over time of the whole Glasgow Outcome Scale (GOS) throughout those years.

METHODS

The study is an observational cohort study of adults (≥ 15 years old) with closed sTBI (GCS ≤ 8) who were admitted within 48 h after injury. The final outcome was the 1-year GOS, which was divided as follows: (1) dead/vegetative, (2) severely disabled (dependent patients), and (3) good/moderate recovery (independent patients). Patients were treated uniformly according to international protocols in a dedicated ICU. We considered patient characteristics that were previously identified as important predictors and could be determined easily and reliably. The admission years were divided into three intervals (1987-1995, 1996-2004, and 2005-2012), and the following individual CT characteristics were noted: the presence of traumatic subarachnoid or intraventricular hemorrhage (tSAH, IVH), midline shift, cisternal status, and the volume of mass lesions (A × B × C/2). Ordinal logistic regression was performed to estimate associations between predictors and outcomes. The patients' estimated propensity scores were included as an independent variable in the ordinal logistic regression model (TWANG R package).

FINDINGS

The variables associated with the outcome were age, pupils, motor score, deterioration, shock, hypoxia, cistern status, IVH, tSAH, and epidural volume. When adjusting for those variables and the propensity score, we found a reduction in mortality from 55% (1987-1995) to 38% (2005-2012), but we discovered an increase in dependent patients from 10 to 21% and just a modest increase in independent patients of 6%.

CONCLUSIONS

This study covers 25 years of management of sTBI in a single neurosurgical center. The prognostic factors are similar to those in the literature. The improvement in mortality does not translate to better quality of life.

Utility-Weighted Modified Rankin Scale as Primary Outcome in Stroke Trials: A Simulation Study

Supplemental Digital Content is available in the text.

Background and Purpose—

The utility-weighted modified Rankin Scale (UW-mRS) has been proposed as a new patient-centered primary outcome in stroke trials. We aimed to describe utility weights for the mRS health states and to evaluate the statistical efficiency of the UW-mRS to detect treatment effects in stroke intervention trials.

Methods—

We used data of the 500 patients enrolled in the MR CLEAN (Multicenter Randomized Clinical Trial of Endovascular Treatment for Acute Ischemic Stroke in the Netherlands). Utility values were elicited from the EuroQol Group 5-Dimension Self-Report Questionnaire assessed at 90 days after inclusion, simultaneously with the mRS. Utility weights were determined by averaging the utilities of all patients within each mRS category. We performed simulations to evaluate statistical efficiency. The simulated treatment effect was an odds ratio of 1.65 in favor of the treatment arm, similar for all mRS cutoffs. This treatment effect was analyzed using 3 approaches: linear regression with the UW-mRS as outcome, binary logistic regression with a dichotomized mRS (0–1/2–6, 0–2/3–6, and 0–4/5–6), and proportional odds logistic regression with the ordinal mRS. The statistical power of the 3 approaches was expressed as the proportion of 10 000 simulations that resulted in a statistically significant treatment effect (P≤0.05).

Results—

The mean utility values (SD) for mRS categories 0 to 6 were: 0.95 (0.08), 0.93 (0.13), 0.83 (0.21), 0.62 (0.27), 0.42 (0.28), 0.11 (0.28), and 0 (0), respectively, but varied substantially between individual patients within each category. The UW-mRS approach was more efficient than the dichotomous approach (power 85% versus 71%) but less efficient than the ordinal approach (power 85% versus 87%).

Conclusions—

The UW-mRS as primary outcome does not capture individual variation in utility values and may reduce the statistical power of a randomized trial.

Using ordinal outcomes to construct and select biomarker combinations for single-level prediction

BACKGROUND

Biomarker studies may involve an ordinal outcome, such as no, mild, or severe disease. There is often interest in predicting one particular level of the outcome due to its clinical significance.

METHODS

A simple approach to constructing biomarker combinations in this context involves dichotomizing the outcome and using a binary logistic regression model. We assessed whether more sophisticated methods offer advantages over this simple approach. It is often necessary to select among several candidate biomarker combinations. One strategy involves selecting a combination based on its ability to predict the outcome level of interest. We propose an algorithm that leverages the ordinal outcome to inform combination selection. We apply this algorithm to data from a study of acute kidney injury after cardiac surgery, where kidney injury may be absent, mild, or severe.

RESULTS

Using more sophisticated modeling approaches to construct combinations provided gains over the simple binary logistic regression approach in specific settings. In the examples considered, the proposed algorithm for combination selection tended to reduce the impact of bias due to selection and to provide combinations with improved performance.

CONCLUSIONS

Methods that utilize the ordinal nature of the outcome in the construction and/or selection of biomarker combinations have the potential to yield better combinations.

Effect of sex and age on traumatic brain injury: a geographical comparative study

Background

Traumatic brain injury (TBI) is a much researched topic in medical health, which requires additional studies to understand various effects of demographic and geographic factors that can assist in developing the most effective treatments. Thousands of people of different ages are suffering from lifelong disabilities, either mild or severe, from TBI and the number is increasing. This study aims to increase our understanding of the effect of sex and age by applying five different statistical methods to evaluate the effect of these covariates on two independent TBI data sets representing patients from different geographical cohorts. A primary data was collected from Bangladesh and it was compared with CRASH (Corticosteroid Randomisation after Significant Head Injury) data, representing various countries around the world.

Methods

The outcome variable for TBI considered in this paper is Glasgow Outcome Scale, which is a four point scale. It was converted to a binary outcome scale for fitting of Fisher’s exact test, a test of proportions and a binary linear model. For analyzing ordinal outcomes, the proportional odds model and the sliding dichotomy model were fitted. As the sample size of the Bangladeshi data set was small, parametric bootstrapping was applied for the consistency of results.

Results

Females were the worse sufferers of TBI compared to men, according to CRASH data set. The old (aged above 58 years) followed by adults (age 25 to 58) were the most vulnerable victims. Interaction effects concluded that old women tended to endure the worst outcomes of TBI. This conclusion came from the CRASH data set representing the world in general, whereas such effects were not present in the Bangladesh data set. Additional application of parametric bootstrapping for the smaller Bangladesh data set did not result into any significant outcome.

Conclusion

The effect of gender and age could be stronger in some countries than others which is driving the significance in CRASH and was not found in Bangladesh. It reflects the necessity of incorporating geographic patterns as well as demographic features of patients while developing treatments and designing clinical trials.

The rapid and accurate categorisation of critically ill patients (RACE) to identify outcomes of interest for longitudinal studies: a feasibility study

The capacity to measure the impact of an intervention on long-term functional outcomes might be improved if research methodology reflected our clinical approach, which is to individualise goals of care to what is achievable for each patient. The objective of this multicentre inception cohort study was to evaluate the feasibility of rapidly and accurately categorising patients, who were eligible for simulated enrolment into a clinical trial, into unique categories based on premorbid function. Once a patient met eligibility criteria a rapid 'baseline assessment' was conducted to categorise patients into one of eight specified groups. A subsequent 'gold standard' assessment was made by an independent blinded assessor once patients had recovered sufficiently to allow such an assessment to occur. Accuracy was predefined as agreement in >80% of assessments. One hundred and twenty-two patients received a baseline assessment and 104 (85%) were categorised to a unique category. One hundred and six patients survived to have a gold standard assessment performed, with 100 (94%) assigned to a unique category. Ninety-two patients had both a baseline and gold standard assessment, and these agreed in 65 (71%) patients. It was not feasible to rapidly and accurately categorise patients according to premorbid function.

Analysis of an ordinal endpoint for use in evaluating treatments for severe influenza requiring hospitalization

Background/Aims A single best endpoint for evaluating treatments of severe influenza requiring hospitalization has not been identified. A novel six-category ordinal endpoint of patient status is being used in a randomized controlled trial (FLU-Intravenous Immunoglobulin - FLU-IVIG) of intravenous immunoglobulin. We systematically examine four factors regarding the use of this ordinal endpoint that may affect power from fitting a proportional odds model: (1) deviations from the proportional odds assumption which result in the same overall treatment effect as specified in the FLU-IVIG protocol and which result in a diminished overall treatment effect, (2) deviations from the distribution of the placebo group assumed in the FLU-IVIG design, (3) the effect of patient misclassification among the six categories, and (4) the number of categories of the ordinal endpoint. We also consider interactions between the treatment effect (i.e. factor 1) and each other factor. Methods We conducted a Monte Carlo simulation study to assess the effect of each factor. To study factor 1, we developed an algorithm for deriving distributions of the ordinal endpoint in the two treatment groups that deviated from proportional odds while maintaining the same overall treatment effect. For factor 2, we considered placebo group distributions which were more or less skewed than the one specified in the FLU-IVIG protocol by adding or subtracting a constant from the cumulative log odds. To assess factor 3, we added misclassification between adjacent pairs of categories that depend on subjective patient/clinician assessments. For factor 4, we collapsed some categories into single categories. Results Deviations from proportional odds reduced power at most from 80% to 77% given the same overall treatment effect as specified in the FLU-IVIG protocol. Misclassification and collapsing categories can reduce power by over 40 and 10 percentage points, respectively, when they affect categories with many patients and a discernible treatment effect. But collapsing categories that contain no treatment effect can raise power by over 20 percentage points. Differences in the distribution of the placebo group can raise power by over 20 percentage points or reduce power by over 40 percentage points depending on how patients are shifted to portions of the ordinal endpoint with a large treatment effect. Conclusion Provided that the overall treatment effect is maintained, deviations from proportional odds marginally reduce power. However, deviations from proportional odds can modify the effect of misclassification, the number of categories, and the distribution of the placebo group on power. In general, adjacent pairs of categories with many patients should be kept separate to help ensure that power is maintained at the pre-specified level.

Towards personalised intra-arterial treatment of patients with acute ischaemic stroke: a study protocol for development and validation of a clinical decision aid

INTRODUCTION

Overall, intra-arterial treatment (IAT) proved to be beneficial in patients with acute ischaemic stroke due to a proximal occlusion in the anterior circulation. However, heterogeneity in treatment benefit may be relevant for personalised clinical decision-making. Our aim is to improve selection of patients for IAT by predicting individual treatment benefit or harm.

METHODS AND ANALYSIS

We will use data collected in the Multicenter Randomized Clinical Trial of Endovascular Treatment for Acute Ischemic Stroke in the Netherlands (MR CLEAN) trial to analyse the effect of baseline characteristics on outcome and treatment effect. A multivariable proportional odds model with interaction terms will be developed to predict the outcome for each individual patient, both with and without IAT. Model performance will be expressed as discrimination and calibration, after bootstrap resampling and shrinkage of regression coefficients, to correct for optimism. External validation will be conducted on data of patients in the Interventional Management of Stroke III trial (IMS III). Primary outcome will be the modified Rankin Scale (mRS) at 90 days after stroke.

ETHICS AND DISSEMINATION

The proposed study will provide an internationally applicable clinical decision aid for IAT. Findings will be disseminated widely through peer-reviewed publications, conference presentations and in an online web application tool. Formal ethical approval was not required as primary data were already collected.

TRIAL REGISTRATION NUMBERS

ISRCTN10888758; Post-results and NCT00359424; Post-resultsc.

Effect of Right Insular Involvement on Death and Functional Outcome After Acute Ischemic Stroke in the IST-3 Trial (Third International Stroke Trial)

Background and Purpose—

In patients with acute ischemic stroke, whether involvement of the insular cortex influences outcome is controversial. Much of the apparent adverse outcome may relate to such strokes usually being severe. We examined the influence of right and left insular involvement on stroke outcomes among patients from the IST-3 study (Third International Stroke Trial) who had visible ischemic stroke on neuroimaging.

Methods—

We used multiple logistic regression to compare outcomes of left versus right insular and noninsular strokes across strata of stroke severity, on death, proportion dead or dependent, and level of disability (ordinalized Oxford Handicap Score) at 6 months, with adjustment for the effects of age, lesion size, and presence of atrial fibrillation.

Results—

Of 3035 patients recruited, 2099 had visible ischemic strokes limited to a single hemisphere on computed tomography/magnetic resonance scans. Of these, 566 and 714 had infarction of right and left insula. Six months after randomization, right insular involvement was associated with increased odds of death when compared with noninsular strokes on the left side (adjusted odds ratio, 1.83; 95% confidence interval, 1.33−2.52), whereas the adjusted odds ratio comparing mortality after insular versus noninsular strokes on the left side was not significant. Among mild/moderate strokes, outcomes for right insular involvement were worse than for left insular, but among more severe strokes, the difference in outcomes was less substantial.

Conclusions—

We found an association between right insular involvement and higher odds of death and worse functional outcome. The difference between right- and left-sided insular lesions on outcomes seemed to be most evident for mild/moderate strokes.

Clinical Trial Registration—

URL: http://www.isrctn.com . Unique identifier: ISRCTN25765518.

Two-year clinical follow-up of the Multicenter Randomized Clinical Trial of Endovascular Treatment for Acute Ischemic Stroke in The Netherlands (MR CLEAN): design and statistical analysis plan of the extended follow-up study

Background

MR CLEAN was the first randomized trial to demonstrate the short-term clinical effectiveness of endovascular treatment in patients with acute ischemic stroke caused by large vessel occlusion in the anterior circulation. Several other trials confirmed that endovascular treatment improves clinical outcome at three months. However, limited data are available on long-term clinical outcome. We aimed to estimate the effect of endovascular treatment on functional outcome at two-year follow-up in patients with acute ischemic stroke. Secondly, we aimed to assess the effect of endovascular treatment on major vascular events and mortality during two years of follow-up.

Methods

MR CLEAN is a multicenter clinical trial with randomized treatment allocation, open-label treatment, and blinded endpoint evaluation. Patients included were 18 years or older with acute ischemic stroke caused by a proven anterior proximal artery occlusion who could be treated within six hours after stroke onset. The intervention contrast was endovascular treatment and usual care versus no endovascular treatment and usual care. The current study extended the follow-up duration from three months to two years.

The primary outcome is the score on the modified Rankin scale at two years. Secondary outcomes include all-cause mortality and the occurrence of major vascular events within two years of follow-up.

Discussion

The results of our study provide information on the long-term clinical effectiveness of endovascular treatment, which may have implications for individual treatment decisions and estimates of cost-effectiveness.

Trial registration

NTR1804. Registered on 7 May 2009; ISRCTN10888758. Registered on 24 July 2012 (main MR CLEAN trial); NTR5073. Registered on 26 February 2015 (extended follow-up study).

Autoregressive transitional ordinal model to test for treatment effect in neurological trials with complex endpoints

BACKGROUND

A number of potential therapeutic approaches for neurological disorders have failed to provide convincing evidence of efficacy, prompting pharmaceutical and health companies to discontinue their involvement in drug development. Limitations in the statistical analysis of complex endpoints have very likely had a negative impact on the translational process.

METHODS

We propose a transitional ordinal model with an autoregressive component to overcome previous limitations in the analysis of Upper Extremity Motor Scores, a relevant endpoint in the field of Spinal Cord Injury. Statistical power and clinical interpretation of estimated treatment effects of the proposed model were compared to routinely employed approaches in a large simulation study of two-arm randomized clinical trials. A revisitation of a key historical trial provides further comparison between the different analysis approaches.

RESULTS

The proposed model outperformed all other approaches in virtually all simulation settings, achieving on average 14 % higher statistical power than the respective second-best performing approach (range: -1 %, +34 %). Only the transitional model allows treatment effect estimates to be interpreted as conditional odds ratios, providing clear interpretation and visualization.

CONCLUSION

The proposed model takes into account the complex ordinal nature of the endpoint under investigation and explicitly accounts for relevant prognostic factors such as lesion level and baseline information. Superior statistical power, combined with clear clinical interpretation of estimated treatment effects and widespread availability in commercial software, are strong arguments for clinicians and trial scientists to adopt, and further extend, the proposed approach.

The Glasgow Outcome Scale - 40 years of application and refinement

The Glasgow Outcome Scale (GOS) was first published in 1975 by Bryan Jennett and Michael Bond. With over 4,000 citations to the original paper, it is the most highly cited outcome measure in studies of brain injury and the second most-cited paper in clinical neurosurgery. The original GOS and the subsequently developed extended GOS (GOSE) are recommended by several national bodies as the outcome measure for major trauma and for head injury. The enduring appeal of the GOS is linked to its simplicity, short administration time, reliability and validity, stability, flexibility of administration (face-to-face, over the telephone and by post), cost-free availability and ease of access. These benefits apply to other derivatives of the scale, including the Glasgow Outcome at Discharge Scale (GODS) and the GOS paediatric revision. The GOS was devised to provide an overview of outcome and to focus on social recovery. Since the initial development of the GOS, there has been an increasing focus on the multidimensional nature of outcome after head injury. This Review charts the development of the GOS, its refinement and usage over the past 40 years, and considers its current and future roles in developing an understanding of brain injury.

A collaborative sequential meta-analysis of individual patient data from randomized trials of endovascular therapy and tPA vs. tPA alone for acute ischemic stroke: ThRombEctomy And tPA (TREAT) analysis: statistical analysis plan for a sequential meta-analysis performed within the VISTA-Endovascular collaboration

Rationale

Endovascular treatment has been shown to restore blood flow effectively. Second-generation medical devices such as stent retrievers are now showing overwhelming efficacy in clinical trials, particularly in conjunction with intravenous recombinant tissue plasminogen activator.

Aims and Design

This statistical analysis plan utilizing a novel, sequential approach describes a prospective, individual patient data analysis of endovascular therapy in conjunction with intravenous recombinant tissue plasminogen activator agreed upon by the Thrombectomy and Tissue Plasminogen Activator Collaborative Group.

Study outcomes

This protocol will specify the primary outcome for efficacy, as ‘favorable’ outcome defined by the ordinal distribution of the modified Rankin Scale measured at three-months poststroke, but with modified Rankin Scales 5 and 6 collapsed into a single category. The primary analysis will aim to answer the questions: ‘what is the treatment effect of endovascular therapy with intravenous recombinant tissue plasminogen activator compared to intravenous tissue plasmi-nogen activator alone on full scale modified Rankin Scale at 3 months?’ and ‘to what extent do key patient characteristics influence the treatment effect of endovascular therapy?’. Key secondary outcomes include effect of endovascular therapy on death within 90 days; analyses of modified Rankin Scale using dichotomized methods; and effects of endovascular therapy on symptomatic intracranial hemorrhage. Several secondary analyses will be considered as well as expanding patient cohorts to intravenous recombinant tissue plasminogen activator-ineligible patients, should data allow.

Discussion

This collaborative meta-analysis of individual participant data from randomized trials of endovascular therapy vs. control in conjunction with intravenous thrombolysis will demonstrate the efficacy and generalizability of endovascular therapy with intravenous thrombolysis as a concomitant medication.

Comparative study of outcome measures and analysis methods for traumatic brain injury trials

Batteries of functional and cognitive measures have been proposed as alternatives to the Extended Glasgow Outcome Scale (GOSE) as the primary outcome for traumatic brain injury (TBI) trials. We evaluated several approaches to analyzing GOSE and a battery of four functional and cognitive measures. Using data from a randomized trial, we created a "super" dataset of 16,550 subjects from patients with complete data (n=331) and then simulated multiple treatment effects across multiple outcome measures. Patients were sampled with replacement (bootstrapping) to generate 10,000 samples for each treatment effect (n=400 patients/group). The percentage of samples where the null hypothesis was rejected estimates the power. All analytic techniques had appropriate rates of type I error (≤5%). Accounting for baseline prognosis either by using sliding dichotomy for GOSE or using regression-based methods substantially increased the power over the corresponding analysis without accounting for prognosis. Analyzing GOSE using multivariate proportional odds regression or analyzing the four-outcome battery with regression-based adjustments had the highest power, assuming equal treatment effect across all components. Analyzing GOSE using a fixed dichotomy provided the lowest power for both unadjusted and regression-adjusted analyses. We assumed an equal treatment effect for all measures. This may not be true in an actual clinical trial. Accounting for baseline prognosis is critical to attaining high power in Phase III TBI trials. The choice of primary outcome for future trials should be guided by power, the domain of brain function that an intervention is likely to impact, and the feasibility of collecting outcome data.

Alteplase for acute ischemic stroke: outcomes by clinically important subgroups in the Third International Stroke Trial

BACKGROUND AND PURPOSE

Our aim was to identify whether particular subgroups of patients had an unacceptably high risk of symptomatic intracranial hemorrhage or low chance of benefit when treated with alteplase (recombinant tissue-type plasminogen activator).

METHODS

Third International Stroke Trial was an international randomized trial of the intravenous (IV) recombinant plasminogen activator alteplase (0.9 mg/kg) versus control in 3035 (1515 versus 1520) patients. We analyzed the effect of recombinant tissue-type plasminogen activator on 6-month functional outcome, early death, and symptomatic intracranial hemorrhage (both ≤7 days). We tested for any differences in treatment effect between subgroups by a test of interaction. Our 13 protocol prespecified subgroups were time to randomization, age, sex, stroke subtype, atrial fibrillation, early ischemic change (clinician and expert panel), prior antiplatelet use, stroke severity, diastolic and systolic blood pressure at randomization, center's thrombolysis experience, and trial phase. Analyses were adjusted for key baseline prognostic factors.

RESULTS

There were no significant interactions in the subgroups analyzed that were consistent across all 3 outcomes. Treatment with recombinant tissue-type plasminogen activator increased the odds of symptomatic intracranial hemorrhage by a greater amount in patients taking prior antiplatelets than those who were not (P=0.019 for test of interaction), but had no clear detrimental effect on functional outcome at 6 months in this group (P=0.781 for test of interaction).

CONCLUSIONS

Among the types of patient in the Third International Stroke Trial, this secondary analysis did not identify any subgroups for whom treatment should be avoided. Given the limitations of the analysis, we found no clear evidence to avoid treatment in patients with prior ischemic stroke, diabetes mellitus, or hypertension.

CLINICAL TRIAL REGISTRATION URL

http://www.controlled-trials.com. Unique identifier: ISRCTN25765518. http://www.controlled-trials.com/ISRCTN25765518.

Update of the Preventive Antibiotics in Stroke Study (PASS): statistical analysis plan

BACKGROUND

Infections occur in 30% of stroke patients and are associated with unfavorable outcomes. Preventive antibiotic therapy lowers the infection rate after stroke, but the effect of preventive antibiotic treatment on functional outcome in patients with stroke is unknown. The PASS is a multicenter, prospective, phase three, randomized, open-label, blinded end-point (PROBE) trial of preventive antibiotic therapy in acute stroke. Patients are randomly assigned to either ceftriaxone at a dose of 2 g, given every 24 h intravenously for 4 days, in addition to standard stroke-unit care, or standard stroke-unit care without preventive antibiotic therapy. The aim of this study is to assess whether preventive antibiotic treatment improves functional outcome at 3 months by preventing infections. This paper presents in detail the statistical analysis plan (SAP) of the Preventive Antibiotics in Stroke Study (PASS) and was submitted while the investigators were still blinded for all outcomes.

RESULTS

The primary outcome is the score on the modified Rankin Scale (mRS), assessed by ordinal logistic regression analysis according to a proportional odds model. Secondary analysis of the primary outcome is the score on the mRS dichotomized as a favorable outcome (mRS 0 to 2) versus unfavorable outcome (mRS 3 to 6). Secondary outcome measures are death rate at discharge and 3 months, infection rate during hospital admission, length of hospital admission, volume of post-stroke care, use of antibiotics during hospital stay, quality-adjusted life years and costs. Complications of treatment, serious adverse events (SAEs) and suspected unexpected serious adverse reactions (SUSARs) are reported as safety outcomes.

CONCLUSIONS

The data from PASS will establish whether preventive antibiotic therapy in acute stroke improves functional outcome by preventing infection and will be analyzed according to this pre-specified SAP.

TRIAL REGISTRATION

Current controlled trials; ISRCTN66140176. Date of registration: 6 April 2010.

MR CLEAN, a multicenter randomized clinical trial of endovascular treatment for acute ischemic stroke in the Netherlands: study protocol for a randomized controlled trial

BACKGROUND

Endovascular or intra-arterial treatment (IAT) increases the likelihood of recanalization in patients with acute ischemic stroke caused by a proximal intracranial arterial occlusion. However, a beneficial effect of IAT on functional recovery in patients with acute ischemic stroke remains unproven. The aim of this study is to assess the effect of IAT on functional outcome in patients with acute ischemic stroke. Additionally, we aim to assess the safety of IAT, and the effect on recanalization of different mechanical treatment modalities.

METHODS/DESIGN

A multicenter randomized clinical trial with blinded outcome assessment. The active comparison is IAT versus no IAT. IAT may consist of intra-arterial thrombolysis with alteplase or urokinase, mechanical treatment or both. Mechanical treatment refers to retraction, aspiration, sonolysis, or use of a retrievable stent (stent-retriever). Patients with a relevant intracranial proximal arterial occlusion of the anterior circulation, who can be treated within 6 hours after stroke onset, are eligible. Treatment effect will be estimated with ordinal logistic regression (shift analysis); 500 patients will be included in the trial for a power of 80% to detect a shift leading to a decrease in dependency in 10% of treated patients. The primary outcome is the score on the modified Rankin scale at 90 days. Secondary outcomes are the National Institutes of Health stroke scale score at 24 hours, vessel patency at 24 hours, infarct size on day 5, and the occurrence of major bleeding during the first 5 days.

DISCUSSION

If IAT leads to a 10% absolute reduction in poor outcome after stroke, careful implementation of the intervention could save approximately 1% of all new stroke cases from death or disability annually.

TRIAL REGISTRATION

NTR1804 (7 May 2009)/ISRCTN10888758 (24 July 2012).

Outcomes and statistical power in adult critical care randomized trials

RATIONALE

Intensive care unit (ICU)-based randomized clinical trials (RCTs) among adult critically ill patients commonly fail to detect treatment benefits.

OBJECTIVES

Appraise the rates of success, outcomes used, statistical power, and design characteristics of published trials.

METHODS

One hundred forty-six ICU-based RCTs of diagnostic, therapeutic, or process/systems interventions published from January 2007 to May 2013 in 16 high-impact general or critical care journals were studied.

MEASUREMENT AND MAIN RESULTS

Of 146 RCTs, 54 (37%) were positive (i.e., the a priori hypothesis was found to be statistically significant). The most common primary outcomes were mortality (n = 40 trials), infection-related outcomes (n = 33), and ventilation-related outcomes (n = 30), with positive results found in 10, 58, and 43%, respectively. Statistical power was discussed in 135 RCTs (92%); 92 cited a rationale for their power parameters. Twenty trials failed to achieve at least 95% of their reported target sample size, including 11 that were stopped early due to insufficient accrual/logistical issues. Of 34 superiority RCTs comparing mortality between treatment arms, 13 (38%) accrued a sample size large enough to find an absolute mortality reduction of 10% or less. In 22 of these trials the observed control-arm mortality rate differed from the predicted rate by at least 7.5%.

CONCLUSIONS

ICU-based RCTs are commonly negative and powered to identify what appear to be unrealistic treatment effects, particularly when using mortality as the primary outcome. Additional concerns include a lack of standardized methods for assessing common outcomes, unclear justifications for statistical power calculations, insufficient patient accrual, and incorrect predictions of baseline event rates.

Simvastatin in aneurysmal subarachnoid haemorrhage (STASH): a multicentre randomised phase 3 trial

BACKGROUND

The benefit of statins in patients with acute aneurysmal subarachnoid haemorrhage is unclear. We aimed to determine whether simvastatin 40 mg could improve the long-term outcome in patients with this disorder.

METHODS

In this international, multicentre, randomised, double-blind trial, we enrolled patients aged 18-65 years with confirmatory evidence of an aneurysmal subarachnoid haemorrhage and presenting less than 96 h from ictus from 35 acute neurosurgical centres in nine countries. Patients were randomly allocated (1:1) to receive either simvastatin 40 mg or placebo once a day for up to 21 days. We used a computer-generated randomisation code to randomise patients in every centre by blocks of ten (five simvastatin, five placebo). Participants and investigators were masked to treatment assignment. The primary outcome was the distribution of modified Rankin Scale (mRS) score obtained by questionnaire at 6 months. Analyses were done on the intention-to-treat population. This trial has been completed and is registered with Current Controlled Trials, number ISRCTN75948817.

FINDINGS

Between Jan 6, 2007, and Feb 1, 2013, apart from the period between May 15, 2009, and Feb 8, 2011, when recruitment was on hold, 803 patients were randomly assigned to receive either simvastatin 40 mg (n=391) or placebo (n=412). All patients were included in the intention-to-treat population. 782 (97%) patients had outcome data recorded at 6 months, of whom 560 (72%) were classed as having a favourable outcome, mRS 0-2 (271 patients in the simvastatin group vs 289 in the placebo group). The primary ordinal analysis of the mRS, adjusted for age and World Federation of Neurological Surgeons grade on admission, gave a common odds ratio (OR) of 0·97, 95% CI 0·75-1·25; p=0·803. At 6 months, we recorded 37 (10%) deaths in the simvastatin group compared with 35 (9%) in the placebo group (log-rank p=0·592). 70 (18%) serious adverse events were reported in the simvastatin group compared with 74 (18%) in the placebo group. No suspected unexpected serious adverse reactions were reported.

INTERPRETATION

The STASH trial did not detect any benefit in the use of simvastatin for long-term or short-term outcome in patients with aneurysmal subarachnoid haemorrhage. Despite demonstrating no safety concerns, we conclude that patients with subarachnoid haemorrhage should not be treated routinely with simvastatin during the acute stages.

FUNDING

British Heart Foundation.

The risks and rewards of covariate adjustment in randomized trials: an assessment of 12 outcomes from 8 studies

Background

Adjustment for prognostic covariates can lead to increased power in the analysis of randomized trials. However, adjusted analyses are not often performed in practice.

Methods

We used simulation to examine the impact of covariate adjustment on 12 outcomes from 8 studies across a range of therapeutic areas. We assessed (1) how large an increase in power can be expected in practice; and (2) the impact of adjustment for covariates that are not prognostic.

Results

Adjustment for known prognostic covariates led to large increases in power for most outcomes. When power was set to 80% based on an unadjusted analysis, covariate adjustment led to a median increase in power to 92.6% across the 12 outcomes (range 80.6 to 99.4%). Power was increased to over 85% for 8 of 12 outcomes, and to over 95% for 5 of 12 outcomes. Conversely, the largest decrease in power from adjustment for covariates that were not prognostic was from 80% to 78.5%.

Conclusions

Adjustment for known prognostic covariates can lead to substantial increases in power, and should be routinely incorporated into the analysis of randomized trials. The potential benefits of adjusting for a small number of possibly prognostic covariates in trials with moderate or large sample sizes far outweigh the risks of doing so, and so should also be considered.