Background: dysphagia is common in acute stroke. Accurate detection of the presence or absence of aspiration by bedside swallowing assessment is difficult without objective methods, tending to over-diagnose aspiration. As a result, some patients suffer restricted oral intake unnecessarily.
Objective: we examined the predictive values of pulse oximetry and speech and language therapy bedside swallowing assessment in the detection of aspiration compared with videofluoroscopy.
Design: a double-blind observational study.
Setting: two university teaching hospitals.
Subjects: we studied 53 patients whose acute strokes were confirmed by computed tomography scan.
Methods: Each subject had initial standard bedside swallowing assessment, closely followed by simultaneous and mutually blinded pulse oximetry, swallowing assessment and videofluoroscopy.
Results: 15 of 53 subjects aspirated. Bedside swallowing assessment and saturation assessment at > or = 2% desaturation gave good sensitivity (80% and 87% respectively), but low positive predictive values (50% and 36% respectively). Both assessments mistook laryngeal penetration for aspiration. Re-analysis with aspiration +/- penetration as a new endpoint improved bedside swallowing assessment positive predictive values to 83% (chi2 =3.59, P=0.032). Sensitivity of saturation assessment was maintained at 86%, positive predictive values of saturation assessment improved to 69% (chi2=6.74, P=0.009). The combination of bedside swallowing assessment and saturation assessment versus aspiration + penetration gave a positive predictive value of 95%.
Conclusions: screening by saturation assessments detects 86% of aspirators/penetrators and should be followed immediately by bedside swallowing assessment, as the combination of the two assessments gives the best positive predictive value. For patients with acute stroke, we advocate a 10 ml water-swallow screening test with simultaneous pulse oximetry by suitably trained medical and nursing staff. Use of this screening test would improve dysphagia detection whilst minimizing unnecessary restriction of oral intake in stroke patients.