Background
The Australian Stroke Foundation clinical guidelines strongly recommend all stroke and TIA patients have their swallow screened within four hours of admission prior to any oral medications, diet or fluids. The Acute Screening of Swallow in Stroke and Transient Ischaemic Attack (ASSIST) is a commonly used bedside nursing screening tool to determine the risk of dysphagia and resultant aspiration after stroke. It comprises of four stages of assessment; if all stages are passed, the patient can recommence their premorbid diet and oral medications. Failure at any stage deems the patient nil by mouth until a formal speech pathology assessment is completed.
Aims
To assess the positive predictive value of the ASSIST tool against gold-standard, bedside speech pathology assessment.
Methods
We retrospectively reviewed demographics, ASSIST and speech pathology assessments for all patients with a diagnosis of acute stroke admitted to The Alfred Hospital, Melbourne, Australia between April 2021 and March 2022.
Results
Of the 390 patients who had an ASSIST completed, 249/390 (64%) passed the ASSIST and recommenced their premorbid diet. 141/390 (36%) failed the ASSIST.
Of the 141 who failed the ASSIST, only 75/141 (53%) were diagnosed with dysphagia by the speech pathologist assessment. 66/141 (47%) did not have dysphagia. The positive predictive value is 53%.
Conclusion
The ASSIST may result in a prolonged period of inadvertent fasting for patients who could safely commence oral medications, a diet and fluids.
Further investigation is warranted to understand the clinical utility of the ASSIST, including nurse training and inter-rater reliability.
A prospective study of sensitivity and specificity using a modified ASSIST pilot tool would be the next step to improve dysphagia screening tools in acute stroke.