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.