Donaldson 1982.

Methods	RCT Parallel design Multi‐centre ‐ 3 paediatric oncology centres USA
Participants	AGE PN ‐ median = 5.8 years, EN ‐ median = 13.4 years SEX Not reported DISEASE STATUS Wilms' tumour Neuroblastoma Soft tissue sarcoma including rhabdomyosarcoma Bone sarcoma including pelvic, Ewing's sarcoma and osteogenic sarcoma Abdominal non‐Hodgkin's lymphoma Ovarian cancer Only well‐nourished patients randomised CHEMOTHERAPY/RADIOTHERAPY All patients received radiotherapy and all but 2 patients received additional chemotherapy
Interventions	25 patients randomly assigned: PN ‐ n = 12, EN (usual food intake) ‐ n = 13 PN ‐ consisted of crystalline amino acid solution, dextrose, minerals and vitamins. PN patients were fed by central line throughout the period of the study. PN patients were allowed only non‐calorie oral intake during the period of PN. EN (usual food intake) ‐ consisted of no restriction in terms of dietary intake Both groups profited from regular dietary counselling throughout the period of the study
Outcomes	Outcomes measured at the following time points: ‐ End of radiotherapy/chemotherapy ‐ 3‐month follow up Primary outcomes Change in nutritional indices ‐ Median percentage change in triceps skinfold ‐ Median percentage change in arm circumference ‐ Median percentage change in arm muscle circumference ‐ Median change in serum albumin Adverse events ‐ Number of patients with nausea and vomiting ‐ Number of patients with diarrhoea Secondary outcomes ‐ Number of deaths ‐ Performance status (activity compared to baseline)
Notes	PN was begun simultaneously with irradiation and continued throughout radiotherapy and for a 3‐ to 5‐day post‐therapy period as a PN "weaning" period. Patients received PN or EN treatment throughout the course of the radiotherapy treatment which lasted for approximately 6 weeks. All but 2 randomised patients received chemotherapy in addition to radiotherapy.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: "patients randomized to receive or not to receive PN", "patients were stratified on the basis of cancer therapy, i.e. radiation dose and volume, with/or without chemotherapy" Comment: actual method of randomisation not described
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 possible (PN infusion obvious) Participants: not possible (PN infusion 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: PN n = 12 EN n = 13 Withdrawals described PN n = 1 One randomised PN patient complained of hunger and withdrew from the study after only 4 days of PN EN n = 4 Four of the children randomised to the EN arm of the study received PN at some point during their oncologic treatment. One of these 4 requested PN; the other 3 became malnourished and were crossed‐over to PN per protocol. Patients analysed: PN n = 11 EN n = 12 In the evaluation of the effect of PN on nutritional status during oncologic treatment, the EN patient who requested PN and the PN patient who withdrew from the study were excluded from the analysis, making 11 PN and 12 EN patients. Intention‐to‐treat analysis was therefore not conducted.
Selective reporting (reporting bias)	High risk	Comment: Outcomes described in trial as being measured but results not reported: Dietary history Skin tests for delayed hypersensitivity Detailed laboratory studies ‐ haematologic studies Time points: States children were evaluated at 3‐weekly intervals following radiotherapy ‐ but only end of radiotherapy and 3‐month follow up endpoints reported
Other bias	Unclear risk	Nausea scale ‐ appears unvalidated, so unclear as to how well it performs.