Use of percutaneous gastrojejunostomy feeding tubes in the rehabilitation of stroke patients

Objective: To determine the characteristics of and complications in the rehabilitation of stroke patients in whom percutaneous gastrojejunostomy (PGJ) feeding tubes have been placed. Design: Retrospective cohort study. Setting: A rehabilitation unit in a tertiary care hospital. Participants: Stroke patients (n = 563) admitted to a tertiary care hospital over a 10-year period. Intervention: PGJ feeding tubes. Main Outcome Measures: Evidence of aspiration in all videofluoroscopic modified barium swallow (VMBS) studies was noted. For patients with a PGJ feeding tube, the following were recorded: stroke location; results of subsequent VMBS reports length of time from stroke onset to PGJ feeding tube insertion; total time the PGJ feeding tube remained in situ; discharge disposition; and concurrent feeding status. Follow-up was at 1-year poststroke. Complications during the inpatient stay attributable to the PGJ feeding tube were recorded. Results: Thirty-two of all 563 (5.7%) stroke patients admitted and 28 of the 115 (24.3%) proven aspirators, as shown on VMBS studies, had a PGJ feeding tube inserted. Twenty-one of the 563 (3.7%) stroke patients were discharged to the community with PGJ feeding tubes in place. The tubes were inserted on average 37 days after stroke onset. Seventeen of all 88 (19.3%) brainstem stroke patients and 15 of all 29 (51.7%) brainstem stroke patients with documented aspiration had feeding tubes inserted. whereas only 15 of 475 (3.2%) hemispheric stroke rehabilitation patients received a tube. Eleven of 32 (34.3%) patients with a feeding tube were able to resume oral feedings at discharge; within 1 year of discharge, 24 of 32 (75%) had done so. Although there were no serious complications resulting from tube insertions, minor complications were documented in more than 50% of the cases. The tubes were associated with prolonged institutionalization in only 1 case; most patients were discharged on a home tube-feeding program. Conclusions: PGJ feeding tubes were placed in approximately 1 of every 20 of our stroke rehabilitation patients. One third of the tubes were removed before the patients were discharged from rehabilitation and 75% were removed within 1 year. Insertion of the tubes was most common in patients with evidence of aspiration and in patients with brainstem strokes. Complications caused by the tube were minor and all patients but 1 who were discharged with feeding tubes were able to manage the home tube-feeding program.