Donzelli 1996.

Methods	Blinding of randomisation: no Blinding of intervention: can't tell but probably impossible Complete follow‐up: yes Blinding of outcome measurement: can't tell
Participants	Premature infants <37 weeks GA. Haemolysis excluded. Qualifying SBR >205.2. Fibreoptic n=71, conventional (white) n=71, conventional (blue) n=71.
Interventions	Fibreoptic: BiliBlanket at 35 microwatts/cm2/nm (measured). Conventional (white): Sylvania Daylight (1 bank with 8 white tubes) at 9 microwatts/cm2/nm (measured). Conventional (blue): Philips Special Blue (1 bank with 8 blue tubes) at 27 microwatts/cm2/nm (measured).
Outcomes	% decline in SBR/day over treatment period. % decline in SBR/hr over treatment period. Duration of phototherapy.
Notes	Further data awaited
Risk of bias
Bias	Authors' judgement	Support for judgement
Allocation concealment?	High risk	C ‐ Inadequate