Which information derived from the Coma Recovery Scale-Revised provides the most reliable prediction of clinical diagnosis and recovery of consciousness? A comparative study using machine learning techniques.

BACKGROUND: The Coma Recovery Scale-Revised (CRS-R) is the most recommended clinical tool to examine the neurobehavioral condition of individuals with disorders of consciousness (DOCs). Different studies have investigated the prognostic value of the information provided by the conventional administration of the scale, while other measures derived from the scale have been proposed to improve the prognosis of DOCs. However, the heterogeneity of the data used in the different studies prevents a reliable comparison of the identified predictors and measures. AIM: This study investigates which information derived from the CRS-R provides the most reliable prediction of both the clinical diagnosis and recovery of consciousness at the discharge of a long-term neurorehabilitation program. DESIGN: Retrospective observational multisite study. SETTING: The enrollment was performed in three neurorehabilitation facilities of the same hospital network. POPULATION: A total of 171 individuals with DOCs admitted to an inpatient neurorehabilitation program for a minimum of 3 months were enrolled. METHODS: Machine learning classifiers were trained to predict the clinical diagnosis and recovery of consciousness at discharge using clinical confounders and different metrics extracted from the CRS-R scale. RESULTS: Results showed that the neurobehavioral state at discharge was predicted with acceptable and comparable predictive value with all the indices and measures derived from the CRS-R, but for the clinical diagnosis and the Consciousness Domain Index, and the recovery of consciousness was predicted with higher accuracy and similarly by all the investigated measures, with the exception of initial clinical diagnosis. CONCLUSIONS: Interestingly, the total score in the CRS-R and, especially, the total score in its subscales provided the best overall results, in contrast to the clinical diagnosis, which could indicate that a comprehensive measure of the clinical diagnosis rather than the condition of the individuals could provide a more reliable prediction of the neurobehavioral progress of individuals with prolonged DOC. CLINICAL REHABILITATION IMPACT: The results of this work have important implications in clinical practice, offering a more accurate prognosis of patients and thus giving the possibility to personalize and optimize the rehabilitation plan of patients with DoC using low-cost and easily collectable information.

Medical comorbidities in patients with prolonged disorder of consciousness: A narrative review.

BACKGROUND: A few studies specifically addressed medical comorbidities (MCs) in patients with severe acquired traumatic or non-traumatic brain injury and prolonged disorders of consciousness (pDoC; i.e., patients in vegetative state/unresponsive wakefulness syndrome, VS/UWS, or in minimally conscious state, MCS). OBJECTIVE: To provide an overview on incidence of MCs in patients with pDoC. METHODS: Narrative review on most impacting MCs in patients with pDoC, both those directly related to brain damage (epilepsy, neurosurgical complications, spasticity, paroxysmal sympathetic hyperactivity, PSH), and those related to severe disability and prolonged immobility (respiratory comorbidities, endocrine disorders, metabolic abnormalities, heterotopic ossifications). RESULTS: Patients with pDoC are at high risk to develop at least one MC. Moderate or severe respiratory and musculoskeletal comorbidities are the most common MCs. Epilepsy and PSH seem to be more frequent in patients in VS/UWS compared to patients in MCS, likely because of higher severity in the brain damage in VS. Endocrine metabolic, PSH and respiratory complications are less frequent in traumatic etiology, whereas neurogenic heterotopic ossifications are more frequent in traumatic etiology. Spasticity did not significantly differ between VS/UWS and MCS and in the three etiologies. MCs are associated with higher mortality rates, worse clinical improvement and can impact accuracy in the clinical diagnosis. CONCLUSIONS: The frequent occurrence of several MCs requires a specialized rehabilitative setting with high level of multidisciplinary medical expertise to prevent, appropriately recognize and treat them. Comprehensive rehabilitation could avoid possible progression to more serious complications that can negatively impact clinical outcomes.

International survey on the implementation of the European and American guidelines on disorders of consciousness.

Diagnostic, prognostic, and therapeutic procedures for patients with prolonged disorders of consciousness (pDoCs) vary significantly across countries and clinical settings, likely due to organizational factors (e.g., research vs. non-academic hospitals), expertise and availability of resources (e.g., financial and human). Two international guidelines, one from the European Academy of Neurology (EAN) and one from the American Academy of Neurology (AAN) in collaboration with the American Congress of Rehabilitation Medicine (ACRM) and the National Institute on Disability, Independent Living, and Rehabilitation Research (NIDILRR), were developed to facilitate consistent practice among professionals working with this challenging patient population. While the recommendations of both guidelines agree in principle, it remains an open issue how to implement them into clinical practice in the care pathway for patients with pDoCs. We conducted an online survey to explore health professional clinical practices related to the management of patients with pDoCs, and compare said practices with selected recommendations from both the guidelines. The survey revealed that while some recommendations are being followed, others are not and/or may require more honing/specificity to enhance their clinical utility. Particular attention should be given to the implementation of a multimodal assessment of residual consciousness, to the detection and treatment of pain, and to the impact of restrictions imposed by COVID-19 pandemics on the involvement of patients' families/representatives.

Investigating microRNAs as biomarkers in disorders of consciousness: a longitudinal multicenter study.

MicroRNAs (miRNAs) are involved in gene regulation and may affect secondary brain injury and recovery in patients with disorders of consciousness (DoC). This study investigated the role of five miRNAs (150-5p, 132-3p, 23b-3p, 451a, and 16-5p) in prolonged DoC. miRNA levels were assessed in serum samples from 30 patients with unresponsive wakefulness syndrome or minimally conscious state due to traumatic or hypoxic-ischemic brain injury (TBI, HIBI) at baseline (1-3 months) and 6 months post-injury. Patients' diagnoses were determined using the Coma Recovery Scale revised, and functional outcomes were evaluated 6 months after injury with the Glasgow Outcome Scale Extended (GOSE) and the Functional Independence Measure (FIM). Compared to healthy controls, patients with TBI had lower levels of miRNAs 150-5p, 132-3p, and 23b-3p at baseline, while patients with HIBI had lower levels of miRNA 150-5p at baseline and 6 months post-injury and a reduction of miRNA 451a at baseline. Higher levels of miRNAs 132-3p and 23b-3p were associated with better outcomes in TBI patients as indicated by GOSE and FIM scores. This study highlights distinct miRNA dysregulated patterns in patients with prolonged DoC, dependent on etiology and post-injury time, and suggests that miRNAs 132-3p and 23b-3p may serve as prognostic biomarkers.

Spontaneous eye blinking during an auditory, an interoceptive and a visual task: The role of the sensory modality and the attentional focus.

Previous evidence suggested that spontaneous eye blinking changes as a function of the attentional focus. In particular, eye blink rate (EBR) tends to increase when attention is directed to internal versus environmental processing. Most studies on this issue compared eye blinking during visual and mental imagery tasks, and interpreted the increase in EBR as a mechanism to focus cognitive resources on internal processing by disengaging attention from interfering information. However, since eye blinking also depends on the sensory modality of the task, the findings might be influenced by a modality-specific effect. In the present Registered Report we aim at investigating whether the environmental versus internal attentional focus can affect spontaneous blinking behaviour in non-visual tasks as well, in conditions where visual stimuli are not relevant. In a within-subject design, healthy participants performed an interoceptive task (i.e., heartbeat counting) and an auditory task in which pre-recorded heartbeats were presented aurally; during both tasks irrelevant visual stimuli were also presented. In a further control condition with the same auditory and visual stimuli, the participants were required to focus their attention on visual stimuli. Participants' EBR was recorded during each task by means of an eye-tracking system. We found that, although the interoceptive task was more difficult than the auditory and visual tasks, participants' EBR decreased by a comparable level in all tasks with respect to a rest condition, with no differences between internal versus environmental conditions. The present findings do not support the idea that EBR is modulated by an internal versus external focus of attention, at least in presence of controlled visual stimulation.

The structure of the Early Rehabilitation Barthel Index (ERBI) should be modified: evidence from a Rasch analysis study.

BACKGROUND: The Early Rehabilitation Barthel Index (ERBI) comprises seven items of the Early Rehabilitation Index and ten items of the Barthel Index. The ERBI is usually used to measure functional changes in patients with severe acquired brain injury (sABI), but its measurement properties have yet to be extensively assessed. AIM: To study the unidimensionality and internal construct validity (ICV) of the ERBI through Confirmatory Factor Analysis (CFA), Mokken Analysis (MA), and Rasch Analysis (RA). DESIGN: Multicenter prospective study. SETTING: Inpatients from five intensive rehabilitation centers. POPULATION: Two hundred and forty-seven subjects with sABI. METHODS: ERBI was administered on admission and discharge to study its unidimensionality through CFA and MA and its ICV, reliability, and targeting through RA. RESULTS: The preliminary analyses showed a lack of unidimensionality (RMSEA=0.460 >0.06; SRMR=0.176 >0.06; CFI=1.000 >0.950; TLI=1.000 >0.950). According to CFA, "Confusional state" and "Behavioral disturbance" items showed low factor loadings (<0.40), whereas these two items composed a separate scale within the MA. Furthermore, the baseline RA showed that three items misfitted ("Mechanical ventilation," "Confusional state," "Behavioral disturbances") and a lack of conformity of several ICV requirements. After deletion of three misfitting items and further non-structural modifications (i.e., testlets creation to absorb local dependence between items and item misfit), the solution obtained showed adequate ICV, adequate reliability for measurements at the individual level (PSI>0.85), although with a frank floor effect. This final solution was successfully replicated in a total sample of the subjects. After post-hoc modifications of the score structure of two out of three misfitting items, the subsequent CFA (RMSEA=0.044 <0.06; SRMR=0.056 <0.06; CFI=1.000 >0.950 TLI=1.000 >0.950) and MA showed the resolution of the unidimensional issues. CONCLUSIONS: Although the ERBI is a potentially valuable tool for measuring functioning in the coma-to-community continuum, our analyses suggested its lack of ICV, partly due to an incorrect scoring design of some items. A new perspective multicenter study is proposed to validate a modified version of the ERBI that overcomes the problems highlighted in this analysis. CLINICAL REHABILITATION IMPACT: Our results do not support the use of the original structure of the ERBI in clinical practice and research, as a lack of ICV was highlighted.

Sham-controlled randomized multicentre trial of transcranial direct current stimulation for prolonged disorders of consciousness.

BACKGROUND AND PURPOSE: Transcranial direct current stimulation (tDCS) has been shown to improve signs of consciousness in a subset of patients with disorders of consciousness (DoC). However, no multicentre study confirmed its efficacy when applied during rehabilitation. In this randomized controlled double-blind study, the effects of tDCS whilst patients were in rehabilitation were tested at the group level and according to their diagnosis and aetiology to better target DoC patients who might repond to tDCS. METHODS: Patients received 2 mA tDCS or sham applied over the left prefrontal cortex for 4 weeks. Behavioural assessments were performed weekly and up to 3 months' follow-up. Analyses were conducted at the group and subgroup levels based on the diagnosis (minimally conscious state [MCS] and unresponsive wakefulness syndrome) and the aetiology (traumatic or non-traumatic). Interim analyses were planned to continue or stop the trial. RESULTS: The trial was stopped for futility when 62 patients from 10 centres were enrolled (44 ± 14 years, 37 ± 24.5 weeks post-injury, 18 women, 32 MCS, 39 non-traumatic). Whilst, at the group level, no treatment effect was found, the subgroup analyses at 3 months' follow-up revealed a significant improvement for patients in MCS and with traumatic aetiology. CONCLUSIONS: Transcranial direct current stimulation during rehabilitation does not seem to enhance patients' recovery. However, diagnosis and aetiology appear to be important factors leading to a response to the treatment. These findings bring novel insights into possible cortical plasticity changes in DoC patients given these differential results according to the subgroups of patients.

Scoping Review on the Diagnosis, Prognosis, and Treatment of Pediatric Disorders of Consciousness.

BACKGROUND AND OBJECTIVES: Comprehensive guidelines for the diagnosis, prognosis, and treatment of disorders of consciousness (DoC) in pediatric patients have not yet been released. We aimed to summarize available evidence for DoC with >14 days duration to support the future development of guidelines for children, adolescents and young adults aged 6 months-18 years. METHODS: This scoping review was reported based on Preferred Reporting Items for Systematic reviews and Meta-Analyses-extension for Scoping Reviews guidelines. A systematic search identified records from 4 databases: PubMed, Embase, Cochrane Library, and Web of Science. Abstracts received 3 blind reviews. Corresponding full-text articles rated as "in-scope" and reporting data not published in any other retained article (i.e., no double reporting) were identified and assigned to 5 thematic evaluating teams. Full-text articles were reviewed using a double-blind standardized form. Level of evidence was graded, and summative statements were generated. RESULTS: On November 9, 2022, 2,167 documents had been identified; 132 articles were retained, of which 33 (25%) were published over the past 5 years. Overall, 2,161 individuals met the inclusion criteria; female patients were 527 of 1,554 (33.9%) cases included, whose sex was identifiable. Of 132 articles, 57 (43.2%) were single case reports and only 5 (3.8%) clinical trials; the level of evidence was prevalently low (80/132; 60.6%). Most studies included neurobehavioral measures (84/127; 66.1%) and neuroimaging (81/127; 63.8%); 59 (46.5%) were mainly related to diagnosis, 56 (44.1%) to prognosis, and 44 (34.6%) to treatment. Most frequently used neurobehavioral tools included the Coma Recovery Scale-Revised, Coma/Near-Coma Scale, Level of Cognitive Functioning Assessment Scale, and Post-Acute Level of Consciousness scale. EEG, event-related potentials, structural CT, and MRI were the most frequently used instrumental techniques. In 29/53 (54.7%) cases, DoC improvement was observed, which was associated with treatment with amantadine. DISCUSSION: The literature on pediatric DoCs is mainly observational, and clinical details are either inconsistently presented or absent. Conclusions drawn from many studies convey insubstantial evidence and have limited validity and low potential for translation in clinical practice. Despite these limitations, our work summarizes the extant literature and constitutes a base for future guidelines related to the diagnosis, prognosis, and treatment of pediatric DoC.

When the practice does not meet the theory: results from an Italian survey on the clinical and pathway management of inpatients with decompressive craniectomy or cranioplasty admitted to rehabilitation.

BACKGROUND: Cranioplasty (CP) is supposed to improve the functional outcome of severe acquired brain injury (sABI) patients with decompressive craniectomy (DC). However, ongoing controversies exist regarding its indications, optimum materials, timing, complications, and relationships with hydrocephalus (HC). For these reasons, an International Consensus Conference (ICC) on CP in traumatic brain injury (TBI) was held in June 2018 to issue some recommendations. AIM: The aims of this study were: to investigate cross-sectionally before the ICC the prevalence of DC/CP in sABI inpatients admitted to neurorehabilitation units in Italy; to assess the perception of Italian clinicians working in the sABI neurorehabilitation settings on the management of inpatients with DC/CP during their rehabilitation stay. DESIGN: Cross-sectional. SETTING AND POPULATION: Physiatrists or neurologists working in 38 Italian rehabilitation centers involved in the care of sABI, giving a pooled sample of 599 inpatients. METHODS: Survey questionnaire consisting of 21 closed-ended questions with multiple-choice answers. Sixteen questions regarded the respondents' opinions and experiences regarding the clinical and management aspects of patients. Survey data were collected via e-mail between April and May 2018. RESULTS: About 1/3 of the 599 inpatients had either a DC (18.9%) or a CP (13.5%). TBI and cerebral hemorrhage were strongly associated with DC/CP, although the association was much stronger for TBI. Significant discrepancies were uncovered between some of the recommendations of the ICC and the corresponding perceptions of the respondents, especially regarding the clinical management of patients (i.e., CP timing). Clear guidelines were perceived as the most crucial factor in improving clinical pathways. CONCLUSIONS: Early collaboration between the neurosurgical and the neurorehabilitation teams is crucial to optimize all clinical and organizational factors, which could expedite CP and minimize the risk of complications, such as infections and HC, to ensure the best possible outcome for DC patients, regardless of the etiology of the sABI. CLINICAL REHABILITATION IMPACT: There may be different attitudes and perceptions, if not controversies, between neurorehabilitation physicians and neurosurgeons regarding the optimal clinical and care pathway management of patients with DC/CP in Italy. Therefore, an Italian Consensus Conference involving all stakeholders on the clinical and management pathways of DC/CP patients admitted to neurorehabilitation is advocated.

Long-term clinical evolution of patients with prolonged disorders of consciousness due to severe anoxic brain injury: A meta-analytic study.

BACKGROUND AND PURPOSE: The prognosis of prolonged (28 days to 3 months post-onset) disorders of consciousness (pDoC) due to anoxic brain injury is uncertain. The present study aimed to evaluate the long-term outcome of post-anoxic pDoC and identify the possible predictive value of demographic and clinical information. METHOD: This is a systematic review and meta-analysis. The rates of mortality, any improvement in clinical diagnosis, and recovery of full consciousness at least 6 months after severe anoxic brain injury were evaluated. A cross-sectional approach searched for differences in baseline demographic and clinical characteristics between survivors and non-survivors, patients improved versus not improved, and patients who recovered full consciousness versus not recovered. RESULTS: Twenty-seven studies were identified. The pooled rates of mortality, any clinical improvement and recovery of full consciousness were 26%, 26% and 17%, respectively. Younger age, baseline diagnosis of minimally conscious state versus vegetative state/unresponsive wakefulness syndrome, higher Coma Recovery Scale Revised total score, and earlier admission to intensive rehabilitation units were associated with a significantly higher likelihood of survival and clinical improvement. These same variables, except time of admission to rehabilitation, were also associated with recovery of full consciousness. CONCLUSIONS: Patients with anoxic pDoC might improve over time up to full recovery of consciousness and some clinical characteristics can help predict clinical improvement. These new insights could support clinicians and caregivers in the decision-making on patient management.

The Consciousness Domain Index: External Validation and Prognostic Relevance of a Data-Driven Assessment.

The prognosis of neurological outcomes in patients with prolonged Disorders of Consciousness (pDoC) has improved in the last decades. Currently, the level of consciousness at admission to post-acute rehabilitation is diagnosed by the Coma Recovery Scale-Revised (CRS-R) and this assessment is also part of the used prognostic markers. The consciousness disorder diagnosis is based on scores of single CRS-R sub-scales, each of which can independently assign or not a specific level of consciousness to a patient in a univariate fashion. In this work, a multidomain indicator of consciousness based on CRS-R sub-scales, the Consciousness-Domain-Index (CDI), was derived by unsupervised learning techniques. The CDI was computed and internally validated on one dataset (N=190) and then externally validated on another dataset (N=86). Then, the CDI effectiveness as a short-term prognostic marker was assessed by supervised Elastic-Net logistic regression. The prediction accuracy of the neurological prognosis was compared with models trained on the level of consciousness at admission based on clinical state assessments. CDI-based prediction of emergence from a pDoC improved the clinical assessment-based one by 5.3% and 3.7%, respectively for the two datasets. This result confirms that the data-driven assessment of consciousness levels based on multidimensional scoring of the CRS-R sub-scales improve short-term neurological prognosis with respect to the classical univariately-derived level of consciousness at admission.

A retrospective multicentric study on the effectiveness of intravenous brivaracetam in seizure clusters: Data from the Italian experience.

OBJECTIVE: Nearly half of people with epilepsy (PWE) are expected to develop seizure clusters (SC), with the subsequent risk of hospitalization. The aim of the present study was to evaluate the use, effectiveness and safety of intravenous (IV) brivaracetam (BRV) in the treatment of SC. METHODS: Retrospective multicentric study of patients with SC (≥ 2 seizures/24 h) who received IV BRV. Data collection occurred from January 2019 to April 2022 in 25 Italian neurology units. Primary efficacy outcome was seizure freedom up to 24 h from BRV administration. We also evaluated the risk of evolution into Status Epilepticus (SE) at 6, 12 and 24 h after treatment initiation. A Cox regression model was used to identify outcome predictors. RESULTS: 97 patients were included (mean age 62 years), 74 (76%) of whom had a history of epilepsy (with drug resistant seizures in 49% of cases). BRV was administered as first line treatment in 16% of the episodes, while it was used as first or second drug after benzodiazepines failure in 49% and 35% of episodes, respectively. On the one hand, 58% patients were seizure free at 24 h after BRV administration and no other rescue medications were used in 75 out of 97 cases (77%) On the other hand, SC evolved into SE in 17% of cases. A higher probability of seizure relapse and/or evolution into SE was observed in patients without a prior history of epilepsy (HR 2.0; 95% CI 1.03 - 4.1) and in case of BRV administration as second/third line drug (HR 3.2; 95% CI 1.1 - 9.7). No severe treatment emergent adverse events were observed. SIGNIFICANCE: In our cohort, IV BRV resulted to be well tolerated for the treatment of SC and it could be considered as a treatment option, particularly in case of in-hospital onset. However, the underlying etiology seems to be the main outcome predictor.

Modeling outcome trajectories in patients with acquired brain injury using a non-linear dynamic evolution approach.

This study describes a dynamic non-linear mathematical approach for modeling the course of disease in acquired brain injury (ABI) patients. Data from a multicentric study were used to evaluate the reliability of the Michaelis-Menten (MM) model applied to well-known clinical variables that assess the outcome of ABI patients. The sample consisted of 156 ABI patients admitted to eight neurorehabilitation subacute units and evaluated at baseline (T0), 4 months after the event (T1) and at discharge (T2). The MM model was used to characterize the trend of the first Principal Component Analysis (PCA) dimension (represented by the variables: feeding modality, RLAS, ERBI-A, Tracheostomy, CRS-r and ERBI-B) in order to predict the most plausible outcome, in terms of positive or negative Glasgow outcome score (GOS) at discharge. Exploring the evolution of the PCA dimension 1 over time, after day 86 the MM model better differentiated between the time course for individuals with a positive and negative GOS (accuracy: 85%; sensitivity: 90.6%; specificity: 62.5%). The non-linear dynamic mathematical model can be used to provide more comprehensive trajectories of the clinical evolution of ABI patients during the rehabilitation period. Our model can be used to address patients for interventions designed for a specific outcome trajectory.

Consciousness-Domain Index: a data-driven clustering-based consciousness labeling.

Assessing consciousness results in one of the most complex neurological diagnosis. Even more complex and uncertain is prognosticating on consciousness recovery. Currently, consciousness is assessed by using a six-items scale, the Coma Recovery Scale-Revised. Namely, scores on the sub-items can individually assign or not a specific level of consciousness to a patient. In our work, by using solely the six sub-items of the CRS-R, we implemented a clustering algorithm labeling patients with the Consciousness-Domains Index (CDI) starting from a dataset of 190 patients with a Disorder of Consciousness (DoC). Then, the CDI is compared with the clinical state at admission and at six months via univariate analysis. The number of clusters best dividing the groups resulted equal to two and the most influencing sub-items resulted the visual and motor one. The CDI closely resembles the clinical state at admission (CSA) (Cohen's k=0.85). On the other hand, when comparing CDI and CSA, a net improvement was found in the prognostic power of the neurological outcome at six months, targeted as presence/absence of a DoC ( ). Data-driven techniques pave the way for automated and model-based search of prognostic factors, together with the use of such prognostic factors in multivariate prognostic models. Future works will address the external validation of the CDI, together with the inclusion of the CDI in a multivariate supervised model, in order to assess the true potential of such novel index. Clinical Relevance- A completely data-driven index was derived from a clustering of CRS-R sub-items. It correlates with the neurological outcome at six months better than the state of consciousness at admission.

What names for covert awareness? A systematic review.

Background: With the emergence of Brain Computer Interfaces (BCI), clinicians have been facing a new group of patients with severe acquired brain injury who are unable to show any behavioral sign of consciousness but respond to active neuroimaging or electrophysiological paradigms. However, even though well documented, there is still no consensus regarding the nomenclature for this clinical entity. Objectives: This systematic review aims to 1) identify the terms used to indicate the presence of this entity through the years, and 2) promote an informed discussion regarding the rationale for these names and the best candidates to name this fascinating disorder. Methods: The Disorders of Consciousness Special Interest Group (DoC SIG) of the International Brain Injury Association (IBIA) launched a search on Pubmed and Google scholar following PRISMA guidelines to collect peer-reviewed articles and reviews on human adults (>18 years) published in English between 2006 and 2021. Results: The search launched in January 2021 identified 4,089 potentially relevant titles. After screening, 1,126 abstracts were found relevant. Finally, 161 manuscripts were included in our analyses. Only 58% of the manuscripts used a specific name to discuss this clinical entity, among which 32% used several names interchangeably throughout the text. We found 25 different names given to this entity. The five following names were the ones the most frequently used: covert awareness, cognitive motor dissociation, functional locked-in, non-behavioral MCS (MCS*) and higher-order cortex motor dissociation. Conclusion: Since 2006, there has been no agreement regarding the taxonomy to use for unresponsive patients who are able to respond to active neuroimaging or electrophysiological paradigms. Developing a standard taxonomy is an important goal for future research studies and clinical translation. We recommend a Delphi study in order to build such a consensus.

Predicting outcome of patients with prolonged disorders of consciousness using machine learning models based on medical complexity.

Patients with severe acquired brain injury and prolonged disorders of consciousness (pDoC) are characterized by high clinical complexity and high risk to develop medical complications. The present multi-center longitudinal study aimed at investigating the impact of medical complications on the prediction of clinical outcome by means of machine learning models. Patients with pDoC were consecutively enrolled at admission in 23 intensive neurorehabilitation units (IRU) and followed-up at 6 months from onset via the Glasgow Outcome Scale-Extended (GOSE). Demographic and clinical data at study entry and medical complications developed within 3 months from admission were collected. Machine learning models were developed, targeting neurological outcomes at 6 months from brain injury using data collected at admission. Then, after concatenating predictions of such models to the medical complications collected within 3 months, a cascade model was developed. One hundred seventy six patients with pDoC (M: 123, median age 60.2 years) were included in the analysis. At admission, the best performing solution (k-Nearest Neighbors regression, KNN) resulted in a median validation error of 0.59 points [IQR 0.14] and a classification accuracy of dichotomized GOS-E of 88.6%. Coherently, at 3 months, the best model resulted in a median validation error of 0.49 points [IQR 0.11] and a classification accuracy of 92.6%. Interpreting the admission KNN showed how the negative effect of older age is strengthened when patients' communication levels are high and ameliorated when no communication is present. The model trained at 3 months showed appropriate adaptation of the admission prediction according to the severity of the developed medical complexity in the first 3 months. In this work, we developed and cross-validated an interpretable decision support tool capable of distinguishing patients which will reach sufficient independence levels at 6 months (GOS-E > 4). Furthermore, we provide an updated prediction at 3 months, keeping in consideration the rehabilitative path and the risen medical complexity.