A prediction model of clinical outcomes in prolonged disorders of consciousness: A prospective cohort study

Development of a nomogram for predicting the outcome in patients with prolonged disorders of consciousness based on the multimodal evaluative information

OBJECTIVE

To establish a nomogram prediction model for the patients with prolonged disorders of consciousness (PDOC) caused by brain injury at six months based on behavioral scale scores, neuroelectro-physiological techniques and hypothalamic-pituitary hormone levels.

METHODS

The clinical data of patients with PDOC who were first diagnosed and hospitalized in the Department of Rehabilitation Medicine of The Affiliated Jiangning Hospital of Nanjing Medical University from March 2023 to July 2024 were collected retrospectively. We performed stratified sampling based on etiology and divided into a training set (121 cases) and a validation set (49 cases) in a ratio of 7:3. After a 6-month follow-up, patients were divided into groups with improved consciousness and those without improved consciousness based on changes in CRS-R scores.Clinical behavioral scores, somatosensory evoked potentials, brainstem auditory evoked potentials, and levels of hypothalamic-pituitary hormones were utilized to identify prognostic factors for prolonged disorders of consciousness. Concurrently, a nomogram prediction model was crafted and validated to forecast the prognosis of patients with prolonged disorders of consciousness. Decision curve analysis (DCA) was subsequently employed to appraise the clinical applicability of this predictive model.

RESULTS

The comparison of clinical data between the training and validation cohorts revealed no significant statistical disparities (P > 0.05). Within the training cohort of 121 PDOC patients, 63 (52.1%)PDOC patients exhibited enhanced consciousness levels. Similarly, in the validation cohort of 49 PDOC patients, 25 (51%) PDOC patients showed improvements in consciousness. Utilizing a combination of random forest analysis, LASSO regression, and multivariate Logistic regression, we identified four key predictive variables: CRS-R score (OR = 1.05, 95%CI 1.02-1.08, P = 0.002), BAEP grading(OR = 0.88, 95%CI 0.79-0.98, P = 0.02), N60 classification (OR = 1.22, 95%CI 1.01-1.48, P = 0.02), and Estradiol (OR = 1.01, 95%CI 1.00-1.02, P = 0.01). The area under the curve (AUC) for the predictive model in the training set was 0.919(95%CI 0.87-0.968),while in the validation set, it was 0.888(95%CI 0.796-0.98). The calibration curves demonstrated a high degree of concordance between predicted probabilities and actual results, suggesting that the model possesses strong discriminative power and calibration accuracy. Furthermore, in the context of clinical decision-making, Decision Curve Analysis indicated a superior net benefit for our predictive model.

CONCLUSION

The nomogram model, which integrates CRS-R score, BAEP grading, N60 classification and Estradiol, provides a comprehensive assessment of short-term prognosis in patients with prolonged disorders of consciousness, demonstrating high accuracy.

Eye tracking to support assessment of patients with prolonged disorder of consciousness - a case series

OBJECTIVE

To investigate if eye tracking can support detection of covert voluntary eye movements and to compare these findings with a simultaneously performed clinical assessment according to the Coma Recovery Scale manual regarding visual stimuli.

DESIGN

Observational case series.

SUBJECTS

Twelve outpatients with prolonged disorders of consciousness recruited from the rehabilitation clinic of a regional rehabilitation unit.

METHOD

Eye movements were recorded with a wearable eye tracker while performing 4 test items from the Coma Recovery Scale Revised. The clinical assessment and recorded eye movement responses were analysed for agreement.

RESULTS

Response data was obtained from 238 out of 288 trials. Eye-tracking data were obtained in median 89.6% of the trials (37.5-100%). The eye tracking assessment judged a significantly higher percentage of trials as a response (46.2%) compared with the clinical assessment (18.1%), mainly in test items "visual pursuit" and "visual fixation".

CONCLUSION

Eye tracking showed potential to be more effective in the detection of putative voluntary eye movements compared with conventional examination. Based on the findings in this and previous studies, eye tracking may serve as a useful complementary tool when examining patients with prolonged disorders of consciousness.

Establishment and validation of a bad outcomes prediction model based on EEG and clinical parameters in prolonged disorder of consciousness

Objective

This study aimed to explore the electroencephalogram (EEG) indicators and clinical factors that may lead to poor prognosis in patients with prolonged disorder of consciousness (pDOC), and establish and verify a clinical predictive model based on these factors.

Methods

This study included 134 patients suffering from prolonged disorder of consciousness enrolled in our department of neurosurgery. We collected the data of sex, age, etiology, coma recovery scales (CRS-R) score, complications, blood routine, liver function, coagulation and other laboratory tests, resting EEG data and follow-up after discharge. These patients were divided into two groups: training set (n = 107) and verification set (n = 27). These patients were divided into a training set of 107 and a validation set of 27 for this study. Univariate and multivariate regression analysis were used to determine the factors affecting the poor prognosis of pDOC and to establish nomogram model. We use the receiver operating characteristic (ROC) and calibration curves to quantitatively test the effectiveness of the training set and the verification set. In order to further verify the clinical practical value of the model, we use decision curve analysis (DCA) to evaluate the model.

Result

The results from univariate and multivariate logistic regression analyses suggested that an increased frequency of occurrence microstate A, reduced CRS-R scores at the time of admission, the presence of episodes associated with paroxysmal sympathetic hyperactivity (PSH), and decreased fibrinogen levels all function as independent prognostic factors. These factors were used to construct the nomogram. The training and verification sets had areas under the curve of 0.854 and 0.920, respectively. Calibration curves and DCA demonstrated good model performance and significant clinical benefits in both sets.

Conclusion

This study is based on the use of clinically available and low-cost clinical indicators combined with EEG to construct a highly applicable and accurate model for predicting the adverse prognosis of patients with prolonged disorder of consciousness. It provides an objective and reliable tool for clinicians to evaluate the prognosis of prolonged disorder of consciousness, and helps clinicians to provide personalized clinical care and decision-making for patients with prolonged disorder of consciousness and their families.

Predicting outcome in disorders of consciousness: A mega-analysis

OBJECTIVE

Assessing recovery potential in patients with disorders of consciousness (DoC) is pivotal for guiding clinical and ethical decisions. We conducted a mega-analysis of individual patient data to understand (1) if a time threshold exists, beyond which regaining consciousness is almost impossible, and (2) how recovery varies based on factors such as diagnosis, etiology, age, sex, and neuropsychological status.

METHODS

A systematic literature search revealed a total of 3290 patients. In this sample, we performed a Cox proportional hazards analysis for interval censored data.

RESULTS

We observed a late saturation of probability to regain consciousness in Kaplan-Meier curves, and the annual rate of recovery was remarkably stable, in that approximately 35% of patients regained consciousness per year. Patients in minimally conscious state (MCS) recovered more frequently than patients in unresponsive wakefulness syndrome (UWS). No significant difference was observed between the recovery dynamics of MCS subgroups: MCS+ and MCS-. Patients with hypoxic brain lesions showed worse recovery rate than patients with traumatic brain injury and patients with vascular brain lesions, while the latter two categories did not differ from each other. Male patients had moderately better chance to regain consciousness. While younger UWS patients recovered more frequently than older patients, it was not the case in MCS.

INTERPRETATION

Our findings highlight the necessity for neurologists to exercise caution when making negative predictions in individual cases, challenge traditional beliefs regarding recovery timelines, and underscore the importance of conducting detailed and prolonged assessments to better understand recovery prospects in DoC.

Uncovering Brain Network Insights for Prognosis in Disorders of Consciousness: EEG Source Space Analysis and Brain Dynamics

Accurate prognostic prediction in patients with disorders of consciousness (DOC) is a core clinical concern and a formidable challenge in neuroscience. Resting-state EEG has shown promise in identifying electrophysiological prognostic markers and may be easily deployed at the bedside. However, the lack of brain dynamic modeling and the spatial mixture of signals in scalp EEG have constrained our exploration of biomarkers and comprehension of the mechanisms underlying consciousness recovery. Here, we introduce EEG source space analysis and brain dynamics to investigate the brain networks of patients with DOC (n = 178) with different outcomes (six-month follow-up), followed by graph theory and high-order topological analysis to explore the relationship between network structure and prognosis, and finally assess the importance of features. We show that a positive prognosis is associated with large-scale lower levels of low-frequency hypersynchrony. Moreover, we provide evidence that this pattern is driven not by all brain states but only by specific states. Analyses reveal that the positive prognosis is attributed to the network retaining lower segregation, higher integration, and stronger stability compared to the negative prognosis. Furthermore, our results highlight the importance of brain networks derived from brain dynamics in prognosis. The prognosis models based on clinical and neural features can achieve acceptable and even excellent performance under different outcome definitions (AUC = 0.714-0.893). Overall, our study offers new perspectives for the identification of prognostic biomarkers and provides avenues for profound insights into the mechanisms underlying consciousness improvement or recovery.