Background: The RailCorp Lantern (RL) is a simulation of light emitting diode railway signals developed as a practical and consistent test for colour vision deficient workers.
Aims: To analyse errors made on the RL and correlate with diagnosis and results on other colour vision tests.
Methods: Retrospective audit of RL tests conducted between February 2006 and December 2008 and comparison between results on the RL conducted at 3 and 6 m, the Standard Farnsworth-Munsell D15 (D15) and the Farnsworth Lantern (FL).
Results: Two-hundred and seven tests were available for analysis. There were significant differences between pass rates by test with 57% passing the D15, 14% the FL and 26% for the RL at 6 m (RL6) and 47% for the RL at 3 m (RL3) (P < 0.001). Both deutans and protans had most difficulty identifying the white light of the FL followed by the yellow of the RL. Seventy-nine percent of protans made red omissions at 6 m compared with 33% of deutans (P < 0.01), and 23% of protans made red omissions at 3 m compared with 3% of deutans (P < 0.001).
Conclusions: The RL identifies individuals who can safely read railway signals and who would have been excluded from working had the FL been the sole test. It is proposed that the RL be considered for use by other rail operators.