Dysphagia is common after a stroke and is associated with a poor outco
me in terms of survival or functional recovery. Percutaneous gastrosto
my (PG) provides reliable and safe nutrition for patients with neurolo
gical dysphagia in the short term but little is known about the subseq
uent outcome in stroke patients. We reviewed the medical records of al
l stroke patients who had required a PG in four West Yorkshire hospita
ls over a 30-month period. All patients alive at the time of the study
were contacted and functional status was recorded. Forty-one stroke p
atients had undergone PG and 37 records were obtained. There were 24 m
en and 13 women with a mean age of 74 years. Thirty-three patients had
had a hemiplegia while four patients presented acutely with dysphagia
but no hemiplegia (all had cerebral infarcts on CT scan). The timing
of PG varied with a median time from stroke of 26 days (range 12-131).
Complications included five chest infections (<1 week after PG), thre
e local infections, two tubes pulled out and one perforation. Three pa
tients died in the first 5 days after the PG. Thirty-one of the 37 pat
ients had died at the time of the assessment, 21 during the original h
ospital admission. The median survival from the time of PG was 53 days
(range 2-528) with only 12 patients surviving for more than 3 months.
Six patients were alive at the time of the study and all but one were
severely disabled (mean modified Barthel Index seven). There is no co
nsensus about patient selection or the timing of PG and our data shoul
d lead to more careful consideration of the risks and benefits of the
procedure in stroke patients.