Prognostic value of a three-dimensional dynamic quantitative analysis system to measure facial motion in acute facial paralysis patients

Background To investigate the prognostic value of a three-dimensional dynamic quantitative analysis system to measure facial motion (3D ASFM) in acute facial palsy patients and compare it with subjective grading methods and electroneurography. Methods We continuously recruited 37 patients with acute (< 1 month) Bell's palsy. An integrated evaluation of facial palsy was performed for each patient. The integrated evaluation included the House-Brackmann grading system (H-BGS), Sunnybrook Facial Grading System (SFGS), electroneurography and three-dimensional objective measurements. Then, the entire set of evaluations were repeated for each patient 1 month later. The patients were followed up monthly until recovery or for up to more than 6 months. We adopted the SFGS and H-BGS as the representative subjective grading system and final criteria for recovery. Poor recovery was defined as an SFGS score less than 70 or H-BGS score higher than II. Results Multiple regression analysis was performed to find the best prognostic indicators. In less than 1 month from onset, ENoG had the highest prognostic value. However, in the second month from onset, the results of SFGS and 3D ASFM were identified as the best prognostic parameters, and a prediction formula with a determination coefficient of 0.673 was established. The receiver operating characteristic curves revealed that a gross score of the 3D ASFM less than 31 in the first evaluation and 49 in the second evaluation had higher sensitivity and specificity to predict poor recovery. Conclusions In different phases of Bell's palsy, the best predictor of prognosis is different. ENOG is the most effective predictor of the prognosis in the first month after onset. In the second month after onset, the combination of SFGS and 3D ADSM is considered to be the best prognostic predictor.