We study the proposal by Kharzeev-Levin to determine entanglement entropy in Deep Inelastic Scattering (DIS) from parton distribution functions (PDFs) and relates the former to the entropy of final state hadrons. We find several uncertainties in the current comparison to data, in particular uncertainties related to the overall normalization, the relation between charged versus total hadron multiplicity in the comparison to experimental results as well as different methods to determine the number of partons in Deep Inelastic Scattering. We further provide a comparison to data based on leading order HERA PDF as well as PDFs obtained from an unintegrated gluon distribution subject to next-to-leading order Balitsky-Fadin-Kuraev-Lipatov and Baltisky-Kovchegov evolution. Within uncertainties we find good agreement with H1 data. We provide also predictions for entropy at lower photon virtualities, where non-linear QCD dynamics is expected to become relevant.