This study was designed to compare the percutaneous endoscopic gastrostomy(PEG) tube feeding with the nasogastric(NG) tube feeding for the patients with dysphagia after the stroke, and to find out the most optimal timing for the PEG tube feeding.

We monitored the nutritional parameters, the frequency and the timing of complications, and other risk factors in 54 stroke patients with dysphagia. In the group of patients with the nasogastric(NG) tube feeding, a reduction in nutritional parameters was greater than in the group of patients with PEG tube feeding. Especially the reduction in serum hemoglobin and albumin level was statistically significant. Thirteen cases of aspiration pneumonia who had frequent self removal of feeding tubes developed in the group with NG tube feeding. Most cases of aspiration pneumonia in the NG tube feeding group developed within the first 2 weeks. Complications from the PEG tube feeding group were three cases of upper gastrointestinal bleeding and three cases of local infection. There were no correlations between the duration of dysphagia and the location of brain lesions, the history of tracheostomy, the age, the initial mental status, or the artificial ventilation. But, there was a significant prolongation of duration of dysphagia in the group of patients who had a vocal cord palsy, an absence of gag reflex, a paralytic dysarthria and a prolonged intensive medical care.

We conclude that the PEG tube feeding is a safer and the more effective method to provide a long term enteral nutrition to patients with neurological dysphagia than the NG tube feeding. Since the most complications developed in the first 2 weeks, the PEG tube feeding should be applied within the initial 2nd to 3rd week for the stroke patients with dysphagia and aspiration risks. Further prospective study will be needed to decide an ideal timing of PEG tube feeding after an acute stroke.