Weekes 2009.

Methods	Randomised controlled trial. Duration 12 months Six months intervention and follow‐up to 12 months.
Participants	Adults (n = 66; 35 males and 31 females) outpatients with severe COPD. Mean age 69 years. 66 randomised; 37 completed; 5 deaths in each group; all patients malnourished defined as unintentional weight loss and poor intake or BMI <18.5 kg/m2. 11% dropped out before baseline assessment.
Interventions	Dietary counselling to increase intake and advice on food fortification (n = 31) versus usual care (n = 28).
Outcomes	Survival*, weight*, BMI*, triceps skinfold*, MAC*, MAMC*, grip strength*, energy intake*, cost*, respiratory function, respiratory muscle function, QoL.
Notes	Usual care consisted of leaflet. Data have been obtained from the author. Paper submitted, awaiting decision re acceptance. Post hoc analysis of cost data will be available at the next update.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated randomisation sequence.
Allocation concealment (selection bias)	Low risk	Sealed opaque envelopes.
Blinding (performance bias and detection bias) Clinical outcomes	High risk	All assessments made by the lead investigator who was not blinded to group allocation.
Blinding (performance bias and detection bias) Functional outcomes	High risk	All assessments made by the lead investigator who was not blinded to group allocation.
Blinding (performance bias and detection bias) Nutritional outcomes	High risk	All assessments made by the lead investigator who was not blinded to group allocation.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	66 randomised; 37 completed; 5 deaths in each group; 11% dropped out before baseline assessment.
Selective reporting (reporting bias)	Low risk	All specified outcomes reported but not as mean change (SD) at 6 months (end of intervention) for the outcomes of interest therefore not usable for meta‐analysis. Additional data and information obtained from the author.
Other bias	Low risk	Baseline characteristics reported and no differences between groups.