Nomogram to Predict Cognitive State Improvement after Deep Brain Stimulation for Parkinson’s Disease

Abstract

Purpose: Parkinson’s disease (PD) is a common neurodegenerative disease, for which cognitive impairment is a non‐motor symptom (NMS). Bilateral subthalamic nucleus deep brain stimulation (STN‐DBS) is an effective treatment for PD. This study established a nomogram to predict cognitive improvement rate after STN‐DBS in PD patients. Methods: We retrospectively analyzed 103 PD patients who underwent STN‐DBS. Patients were followed up to measure improvement in MoCA scores one year after surgery. Univariate and multivariate logistic regression analyses were used to identify factors affecting improvement in cognitive status. A nomogram was developed to predict this factor. The discrimination and fitting performance were evaluated by receiver operating characteristics (ROC) analysis, calibration diagram, and decision curve analysis (DCA). Results: Among 103 patients, the mean improvement rate of the MoCA score was 37.3% and the median improvement rate was 27.3%, of which 64% improved cognition, 27% worsened cognition, and 8.7% remained unchanged. Logistic multivariate regression analysis showed that years of education, UPDRSIII drug use, MoCA Preop, and MMSE Preop scores were independent factors affecting the cognitive improvement rate. A nomogram model was subsequently developed. The C‐index of the nomogram was 0.98 (95%CI, 0.97–1.00), and the area under the ROC was 0.98 (95%CI 0.97–1.00). The calibration plot and DCA demonstrated the goodness‐of‐fit between nomogram predictions and actual observations. Conclusion: Our nomogram could effectively predict the possibility of achieving good cognitive improvement one year after STN‐DBS in patients with PD. This model has value in judging the expected cognitive improvement of patients with PD undergoing STN‐DBS.