Smith 1992.

Methods	RCT Parallel design Single‐centre UK
Participants	AGE Median age of 12 patients = 3.0 years EN (nasogastric) ‐ age range 2.0 to 5.5 years EN (usual food intake) ‐ age range 1.9 to 9.3 years SEX EN (nasogastric) ‐ 4 males, 2 females EN (usual food intake) ‐ 4 males, 2 females DISEASE STATUS Juvenile chronic myeloid leukaemia (1) Wilms' tumour (4) Neuroblastoma (4) Rhabdomyosarcoma (1) Liver sarcoma (1) Megakaryocytic leukaemia (1) Patients were malnourished CHEMOTHERAPY/RADIOTHERAPY All patients received chemotherapy
Interventions	12 patients randomly assigned: EN (nasogastric) ‐ n = 6, EN (usual food intake) ‐ n = 6 EN (nasogastric) ‐ consisted of high energy density feed (1.5 kcal/ml) Fortisip Energy Plus (Cow and Gate). Seravit (SHS) was added to provide standard requirements of vitamins and minerals. The aim of the supplementation was to provide at least 100% of recommended daily allowance of energy or 100% of previous daily intake, whichever was higher. Feeds were administered via a Kangaroo 330 pump. Oral feeding was permitted ad libitum but did not influence the quantity of supplementary feed administered. EN (usual food intake) ‐ consisted of regular dietary counselling, the emphasis of which was to promote as high an energy intake as possible.
Outcomes	Primary outcomes Change in nutritional indices ‐ Mean change in mid‐upper arm circumference Adverse events ‐ Number of patients with recurrent vomiting Calorie and nutritional intake ‐ Median percentage of RDA of energy intake Secondary outcomes ‐ Number of deaths ‐ Acceptability of nasogastric feeding ‐ parental questionnaire ‐ Activity ‐ median play scores
Notes	EN (nasogastric) ‐ median duration = 4 weeks, range 3 to 6 weeks Subjects commenced supplementary nasogastric feeding as soon after diagnosis as possible. The feeding routine for nasogastric was flexible, with the expectation that most children would eventually receive nasogastric feeding overnight at home. Feeding was usually commenced at a continuous infusion of 10 ml/h and was increased by 10 ml/h/day, if tolerated. If children were in bed throughout most of each 24‐hour period, nasogastric feeding was continued throughout the 24 hours. In children were more active by day, feeding was restricted to 10 to 14 hours overnight. Once MAC was documented to be increasing, no attempt was made to increase total daily intake any further. Nasogastric feeding was continued until MAC was above the tenth percentile for at least 2 weeks. Follow up was for a period of 3 months from diagnosis. This permitted study during the most intensive phase of disease specific treatment.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "patients were randomized to study or control groups using a computer generated random sequence"
Allocation concealment (selection bias)	Unclear risk	Comment: methods not described
Blinding of participants and personnel (performance bias) All outcomes	High risk	Comment: Clinician/person delivering treatment: not discussed (not possible as nasogastric tube obvious) Participants: not discussed (not possible as nasogastric tube obvious)
Blinding of outcome assessment (detection bias) All outcomes	High risk	Comment: not discussed, Some outcomes were subjective (e.g. nausea) so lack of blinding may have influenced outcome measurement.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Comment: Patients randomised: EN (nasogastric) n = 6 EN (usual food intake) n = 6 Withdrawals described: EN (nasogastric) n = 1 One nasogastric patient withdrew from the study before nasogastric feeding established ‐ remaining 5 patients completed the study EN (usual food intake) n = 3 Two of the 6 EN (usual food intake) patients received nasogastric feeding; one 4 months from diagnosis by which time the study was completed; one 2 weeks from diagnosis, but data kept within EN (usual food intake) group up to 2 weeks. One EN (usual food intake) patient died 2 months following diagnosis. Patients analysed: EN (nasogastric) n = 5 EN (usual food intake) n = 6 Intention‐to‐treat analysis was not undertaken as described in the Cochrane handbook, as an eligible patient dropped out and was not included in the analysis.
Selective reporting (reporting bias)	Low risk	Comment: all outcomes described in the trial as being measured were reported.on
Other bias	High risk	The reproducibility of data from the parental report sheets is unclear.