Establishment of a Multi-parameter Evaluation Model for Risk of Aspiration in Dysphagia: A Pilot Study

It's difficult for clinical bedside evaluations to accurately determine the occurrence of aspiration in patients. Although VFSS and FEES are the gold standards for clinical diagnosis of dysphagia, which are mainly used to evaluate people at high risk of dysphagia found by bedside screening, the operation is complicated and time-consuming. The aim of this pilot study was to present an objective measure based on a multi-parameter approach to screen for aspiration risk in patients with dysphagia. Objective evaluation techniques based on speech parameters were used to assess the oral motor function, vocal cord function, and voice changes before and after swallowing in 32 patients with dysphagia (16 low-risk aspiration group, 16 high-risk aspiration group). Student's t test combined with stepwise logistic regression were used to determine the optimal index. The best model consists of three parameters, and the equation is: logit(P) = - 3.824 - (0.504 × maximum phonation time) + (0.008 × second formant frequency of /u/) - 0.085 × (fundamental frequency difference before and after swallowing). An additional eight patients with dysphagia were randomly selected as the validation group of the model. When applied to validation, this model can accurately identify the risk of aspiration in 87.5% of patients, and the sensitivity is as high as 100%. Therefore, it has certain clinical practical value that may help clinicians to assess the risk of aspiration in patients with dysphagia, especially for silent aspiration.