El‐Bayoumi 2011.

Methods	Randomised controlled trial Open Parallel group
Participants	Children (age not specified) with GBS requiring artificial ventilation
Interventions	IVIg 0.4 g/kg daily for 5 days versus one plasma volume PE daily for 5 days
Outcomes	Duration of mechanical ventilation; length of intensive care unit stay; ability to walk unaided within 4 weeks of intensive care unit discharge
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Done by computer‐generated random tables
Allocation concealment (selection bias)	Unclear risk	Not stated
Blinding (performance bias and detection bias) All outcomes except death	High risk	Although decisions about removal from ventilation were made independently, no attempt was made to mask whether participants had received PE or IVIg
Blinding (performance bias and detection bias) Death	Low risk	Although decisions about removal from ventilation were made independently, no attempt was made to mask whether participants had received PE or IVIg. This would not have affected the assessment of death and in any case there were no deaths
Incomplete outcome data (attrition bias) All outcomes except death	Low risk	Follow‐up complete
Incomplete outcome data (attrition bias) Death	Low risk	No deaths
Selective reporting (reporting bias)	Low risk	All specified outcomes fully reported
Other bias	High risk	Number of participants who did not receive full treatment unknown
Diagnostic criteria	Low risk	Standard diagnostic criteria, requirement for mechanical ventilation and < 14 days of muscle weakness before randomisation
Baseline differences	Low risk	Done by computer‐generated random tables: groups well balanced