Abstract

Various light-emitting diode (LED) phototherapy devices have been trialled on the assumption of a more effective spectral distribution of the light emitted. We reviewed the current literature to determine whether LED is more effective than other types of phototherapy. Eligible studies were randomized controlled trials of LED versus other phototherapies. Studies were found to be of medium quality based on a components approach. Data were statistically aggregated within a very homogeneous population (term or late preterm neonates). Results appeared robust at sensitivity analysis. Five hundred and eleven neonates were included in the meta-analysis. LED and other phototherapy devices appeared to be equally effective in reducing total serum bilirubin (TSB) in term or late preterm neonates. The pooled mean TSB rate of decrease was 3.269 μmol/L/h (0.191 mg/dL/h) and 3.074 μmol/L/h (0.18 mg/dL/h) in the LED and conventional arms, respectively [average difference in TSB rate of decrease = 0.194 μmol/L/h (0.011 mg/dL/h) in favour of LED phototherapy; p = 0.378]. CONCLUSION: No significant difference in TSB rate of decrease was detected between LED and other types of phototherapy. Further randomized controlled trials are needed to ascertain whether LED phototherapy may be more effective when increasing the spectral power, or in certain selected subpopulations.

© 2011 The Author(s)/Acta Paediatrica © 2011 Foundation Acta Paediatrica.