We discuss the gain in signal-to-noise ratio (SNR) recently reported by Liu et al. [Phys. Rev. E 63, 051912 (2001)] in the Hodgkin-Huxley neuronal model. We show first that the possibility of signal-to-noise ratio enhancements can be checked by consideration of the statistical characteristics of switching between the system states, and we examine how the SNR depends on the shape of a periodic signal. Second, we attempt to verify the SNR gain reported by Liu et al.: based on spectral calculations and analyses of switching statistics, we are unable to find any SNR gain in the Hodgkin-Huxley model.