Erratum to: Eur. Phys. J. C (2013) 73:2534 DOI 10.1140/epjc/s10052-013-2534-3

The final conclusion of this paper is not correct. It was based on our misunderstanding of the treatment of the subtraction term to the NLO ladder diagram given in column 4 of Table 1 of the original calculation of Curci, Furmanski and Petronzio [11]. Contrary to our treatment, Curci et al. do not use the ‘+’ prescription to calculate the NLO contribution in 4+2ϵ dimensions. Instead they explicitly trace the cancellation of the singularities caused by the soft gluon emissions between the real and virtual contributions. Thus, it turns out, that the non-singlet structure function and the DGLAP evolution in the physical scheme are exactly the same as in the \(\overline{\mathrm{MS}}\) scheme. Therefore the physical approach may be considered as an alternative factorisation scheme.