Original articleA retrospective review of enteral nutrition support practices at a tertiary pediatric hospital: A comparison of prolonged nasogastric and gastrostomy tube feeding

SummaryBackground & aimsDespite the frequent use of tube-mediated enteral feeding, there is little evidence clarifying best practices pertaining to prolonged nasogastric and gastrostomy tube use in children. At the Montreal Children's Hospital, tube feeding practices are non-standardized and highly variable, with many patients remaining on protracted nasogastric feeds. We aimed to characterize enteral nutrition practices at our institution and to compare prolonged nasogastric and gastrostomy tube use, hypothesizing that earlier gastrostomy improves outcomes, particularly the development of food refusal.MethodsIn this retrospective cohort study, we reviewed the charts of children beginning long-term (>3 months) nasogastric or gastrostomy feeds at our institution between January 2007 and December 2011, with follow-up until May 2013. Patient demographics, anthropometric parameters, swallowing assessment, tube feeding duration and complications were recorded.ResultsAmong 166 patients, the median total tube feeding duration was 24.9 (3.0-75.6) months and varied with underlying disease and swallowing assessment. The median duration of nasogastric tube use was 7.8 (0.7-45.3) months. Food refusal was significantly associated with nasogastric tube exposure >3 months (RR 3.3, p < 0.001, NNT = 3) and anthropometric outcomes were superior in gastrostomy-fed patients. Rates of aspiration pneumonia were similar in both groups. Despite more initial opposition to gastrostomy and a higher complication rate, gastrostomy users appeared more satisfied with their experience, as demonstrated by a much lower discontinuation rate than observed in the nasogastric group.ConclusionsProlonged nasogastric feeding is common at our institution. Its association with increased food refusal and less favorable anthropometric outcomes may warrant earlier gastrostomy placement.