Prediction of the prognosis of Bell's palsy using multivariate analyses

OBJECTIVES: This study aimed to establish more reliable formulae for predicting the prognosis for patients with Bell's palsy at initial examination. PATIENTS: One hundred four consecutive patients with Bell's palsy whose facial movement was either completely recovered or unchanged for more than 6 months. INTERVENTION: Diagnostic. MAIN OUTCOME MEASURES: Multivariate analyses were performed using the following variables: Age, House-Brackmann score (H-B), electroneuronography (ENoG), nerve excitability test (NET), blink reflex, and stapedial muscle reflex on initial examination. RESULTS: Duration to maximal recovery can be predicted with the adjusted coefficient of 0.51 using H-B, ENoG, stapedial muscle reflex, and NET. Whether complete recovery would be achieved can be predicted with the Nagelkerke coefficient of determination of 0.72 using age, H-B, ENoG, and NET. CONCLUSION: This study suggested that the prognosis of patients who recovered early can be predicted more precisely; however, the prognosis of patients who recovered late and incomplete could not be predicted easily.