Objective:To explore the factors influencing the recovery of swallowing function after a brainstem hemorrhage and to construct a prediction model.Methods:Clinical data on 134 persons with dysphagia after a brainstem hemorrhage were collected retrospectively. According to their swallowing ability at discharge, the patients were divided into a swallowing recovery group and a non-recovery group. Univariate correlation analysis and multivariate logistic regression analysis were used to explore the independent factors influencing the recovery of swallowing function and to construct a prediction nomogram. The receiver operating characteristics (ROC) curves were evaluated to analyze the nomogram′s predictive value and those of the relevant influencing factors.Results:Sixty-two of the patients (46%) had recovered their swallowing function at discharge, while 72 (54%) had not. Univariate correlation analysis showed that there had been significant differences in tracheal intubation, NIHSS score, FOIS score, Barthel index and Glasgow coma scale (GCS )score between the two groups, on average. The multivariate logistic regressions showed that a low NIHSS score, a high FOIS score and a high GCS score were independent predictors of swallowing function recovery, so they were used in the prediction model. ROC curve analysis showed that the area under the curve (AUC) of the prediction model was 0.953 (95% CI: 0.902~0.982) with a sensitivity of 87% and a specificity of 93%. The model′s predictions were thus better than using an NIHSS score, GCS score or FOIS score alone. Conclusions:NIHSS score, GCS score and FOIS score can independently predict the recovery of swallowing function after a brainstem hemorrhage. A prediction model constructed using all three has good predictive power.

Objective:To explore the factors influencing the recovery of swallowing function after a brainstem hemorrhage and to construct a prediction model.Methods:Clinical data on 134 persons with dysphagia after a brainstem hemorrhage were collected retrospectively. According to their swallowing ability at discharge, the patients were divided into a swallowing recovery group and a non-recovery group. Univariate correlation analysis and multivariate logistic regression analysis were used to explore the independent factors influencing the recovery of swallowing function and to construct a prediction nomogram. The receiver operating characteristics (ROC) curves were evaluated to analyze the nomogram′s predictive value and those of the relevant influencing factors.Results:Sixty-two of the patients (46%) had recovered their swallowing function at discharge, while 72 (54%) had not. Univariate correlation analysis showed that there had been significant differences in tracheal intubation, NIHSS score, FOIS score, Barthel index and Glasgow coma scale (GCS )score between the two groups, on average. The multivariate logistic regressions showed that a low NIHSS score, a high FOIS score and a high GCS score were independent predictors of swallowing function recovery, so they were used in the prediction model. ROC curve analysis showed that the area under the curve (AUC) of the prediction model was 0.953 (95% CI: 0.902~0.982) with a sensitivity of 87% and a specificity of 93%. The model′s predictions were thus better than using an NIHSS score, GCS score or FOIS score alone. Conclusions:NIHSS score, GCS score and FOIS score can independently predict the recovery of swallowing function after a brainstem hemorrhage. A prediction model constructed using all three has good predictive power.