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

ABSTRACT

Objective: This study aimed to establish and validate a prediction model for clinical outcomes in patients with prolonged disorders of consciousness (pDOC). Methods: A total of 170 patients with pDOC enrolled in our rehabilitation unit were included and divided into training (n = 119) and validation sets (n = 51). Independent predictors for improved clinical outcomes were identified by univariate and multivariate logistic regression analyses, and a nomogram model was established. The nomogram performance was quantified using receiver operating curve (ROC) and calibration curves in the training and validated sets. A decision curve analysis (DCA) was performed to evaluate the clinical usefulness of this nomogram model. Results: Univariate and multivariate logistic regression analyses indicated that age, diagnosis at entry, serum albumin (g/L), and pupillary reflex were the independent prognostic factors that were used to construct the nomogram. The area under the curve in the training and validation sets was 0.845 and 0.801, respectively. This nomogram model showed good calibration with good consistency between the actual and predicted probabilities of improved outcomes. The DCA demonstrated a higher net benefit in clinical decision-making compared to treating all or none. Conclusion: Several feasible, cost-effective prognostic variables that are widely available in hospitals can provide an efficient and accurate prediction model for improved clinical outcomes and support clinicians to offer suitable clinical care and decision-making to patients with pDOC and their family members.