It is shown that the result of calculation of the probability of single-photon emission by a two-dimensional hydrogen-like atom can depend on the choice of the spin operator. Specifically, if it is chosen to be the projection operator of the spin onto the plane of motion, then the result, as expected, agrees with the classical theory of emission of a three-dimensional atom, as was demonstrated by the author in one of his previous works. If, however, it is chosen as the projection operator onto the perpendicular to the plane of motion, the result does not agree with the classical theory. This puts in doubt the results presented in [S. H. Guo, X. L. Yang, F. T. Chan, Phys. Rev. A, 43, No. 3, 1197 (1991)].