Clinical study Predictors of prolonged dysphagia following acute stroke S. Broadley 1,2 PHD MRCP PHD MRCP, D. Croser 3 MBBS MBBS, J. Cottrell 4 BAPSC BAPSC, M. Creevy 4 BAPSC BAPSC, E. Teo 1 MBBS MBBS, D. Yiu 1 MBBS MBBS, R. Pathi 1 MBBS MBBS, J. Taylor 3 FRACR FRACR, P.D. Thompson 1,2 PHD, FRACP PHD, FRACP 1 Department of Neurology, Royal Adelaide Hospital, North Terrace, SA 5000, Australia, 2 University Department of Medicine, Adelaide University, SA 5005, Australia, 3 Department of Radiology, Royal Adelaide Hospital, North Terrace, SA 5000, Australia, and 4 Speech Pathology Department, Royal Adelaide Hospital, North Terrace, SA 5000, Australia Summary Dysphagia following acute stroke frequently necessitates prolonged enteral feeding. There is evidence that early enteral feeding via percutaneous endoscopic gastrostomy (PEG) is both beneficial and safe. The aim of this study was to identify predictors of prolonged dysphagia. The subjects were 149 consecutive patients admitted with acute stroke. Clinical findings and imaging results were prospectively collected, and subsequent progress recorded. Subjects were divided into 3 groups for analysis: no dysphagia; transient dysphagia ( O14 days); or prolonged dysphagia (>14 days). Validity of the water swallow test as a predictor of aspiration pneumonia was confirmed. Significant associations for prolonged dysphagia were seen with stroke severity, dysphasia and lesions of the frontal and insular cortex on brain imaging. These results indicate that it may be possible to predict patients who will develop prolonged significant dysphagia following acute stroke thereby facilitating referral for insertion of PEG at an earlier time point. ª 2003 Elsevier Science Ltd. All rights reserved. Keywords: acute stroke, dysphagia, prognosis, PEG, aspiration pneumonia INTRODUCTION Despite being recognised for over a hundred years, 1 the frequent occurrence and consequences of dysphagia following acute hemispheric stroke were only highlighted in the last 20 years. 2 Dysphagia occurs in 30–50% of patients with acute stroke 2–5 and is an independent predictor of outcome. 4 The principal reason for the higher morbidity and mortality is aspiration pneumonia. 4;6 There is also increasing evidence that dysphagia contributes sig- nificantly to protein-energy malnutrition 4;7 in an already at risk group 8 and that protein-energy malnutrition is an independent poor prognostic factor in the first weeks after stroke. 7;9 The issue of identifying patients at risk from aspiration has been intensively investigated. Whilst often considered the Ôgold standardÕ, video-fluoroscopy has not found widespread acceptance in the routine assessment of patients following acute stroke. 4;10;11 This is largely because a high proportion of acute stroke patients 12 and elderly asymptomatic controls 13 can be demonstrated to have some degree of aspiration and the correlation with risk of devel- oping significant aspiration pneumonia is not clear. In contrast, a number of simple clinical assessments of swallowing function have been validated as predictors of aspiration. These have the advantages of being easy to perform in the emergency department, and are less invasive and cheaper. 14 An example is the water swallow test of 50–150 ml, which has been demonstrated to pre- dict risk of aspiration with high sensitivity. 15–18 The sensitivity of this test can be increased by careful assessment of bulbar function with abnormalities of pharyngeal sensation (but not the gag re- flex) 12 and reduced cough reflex 19 being particularly useful in detecting silent aspiration. The early identification of those at risk of significant dys- phagia and requiring nutritional support will permit the study of early nutritional supplementation via nasoenteric tubes (NET) or percutaneous gastrostomy (PEG). Whilst previous studies have looked at the natural history of dysphagia after acute stroke, 20 predictive factors for prolonged dysphagia have not been specif- ically addressed. We have therefore, performed a prospective cohort study of patients admitted to the stroke unit at the Royal Adelaide Hospital, looking specifically for factors, which might predict prolonged dysphagia. METHODS Patients The study was conducted in the Royal Adelaide Hospital stroke unit, which admits all patients presenting with a clinical diagnosis of stroke regardless of age or severity, excepting those requiring neurosurgical intervention. The subjects were consecutive ad- missions during a six month period (12 April to 12 October 2000) with a clinical diagnosis of acute stroke (excluding subarachnoid haemorrhage). Patients were excluded if they were subsequently identified as having had a transient ischaemic attack or a diagnosis other than stroke. Neurological assessment Basic demographic details were noted for all patients. All patients were subjected to a standard history and clinical examination within 24 h of admission. Particular attention was paid to bulbar function, with dysarthria, dysphonia and voluntary cough, together with symmetry and speed of tongue and palatal movements being noted. Gag reflexes were recorded as present, reduced or absent and preservation of pharyngeal sensation was recorded. Patient disability shortly after admission was measured using the modi- fied Barthel index. 21 The water swallow test was performed in all those patients who were able to attend sufficiently to follow the instructions. This consisted of a timed swallow of 50 ml of water. With the patient in a sitting position, they were asked first to take a sip of water and any dribbling, delayed swallowing (greater than 5 s to trigger swallow reflex after water touching lips), repeated swallowing or immediate coughing were noted. If significant Journal of Clinical Neuroscience (2003) 10(3), 300–305 ª 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0967-5868(03)00022-5 Received 20 February 2002 Accepted 22 May 2002 Correspondence to: P.D. Thompson PhD FRACP, Department of Neurology, Royal Adelaide Hospital, North Terrace, SA 5000, Australia. Fax: +61-8-8223-3870; E-mail: philip.thompson@adelaide.edu.au 300