Clinical Question: Do percutaneous endoscopic gastrostomy (PEG) tubes improve outcomes in stroke patients who have dysphagia?

Setting: Inpatient (any location) with outpatient follow-up

Study Design: Randomized controlled trial (single-blinded)

Allocation: Concealed

Synopsis: In this report of two FOOD (Feed Or Ordinary Diet) trials that shared the same protocols, any patient with dysphagia who was admitted with a recent ischemic stroke and whose physician was uncertain of the best approach to feeding was eligible to participate. In the first FOOD trial (early versus avoid), the patients were assigned randomly to start enteral tube feeding as soon as possible (n = 429) or to avoid enteral tube feeding for at least seven days (n = 430). In the second trial (PEG versus nasogastric feeding), patients were randomized to enteral feeding via PEG tube (n = 162) or nasogastric tube (n = 159) within three days of enrollment. The duration of feeding was not predetermined but was based on practicality and the patient’s condition.

The main outcomes, assessed by intention to treat, were death and a poor outcome after six months of follow-up (defined as a modified Rankin score of 4 or 5). The research staff evaluating the outcomes were unaware of how the patients were fed. Only one patient was lost to follow-up. By the end of the study, nearly one half of all patients had died. In the early versus avoid trial, the rates of mortality and poor outcomes were not significantly affected, although ending the study early severely reduced the sample size, thereby potentially missing meaningful differences. In the PEG versus nasogastric trial, mortality rates were not significantly affected, although there was an absolute increase in the number of patients who had a poor outcome (number needed to treat to harm = 13; 95% confidence interval, 6.4 to infinity;P = .05; lower confidence level for adjusted risk ratio = zero).

Bottom Line: In this underpowered study of patients with acute ischemic stroke and dysphagia, early use of enteral tube feeding achieved no better (or worse) outcomes than avoiding it. Using PEG tubes produced a nonsignificant reduction in mortality rates but significantly increased the number of patients who had poor outcomes six months after the stroke. (Level of Evidence: 1b–)