When calculating the detection efficiency of the gravitational microlensing effect, the blending effects have so far been estimated without considering the binary systems that the HST cannot resolve at the distance of the LMC. The components of such systems, if sufficiently separated, are indeed not amplified in the same way, and these configurations must be taken into account in the study of blending. We used the GAIA catalog of nearby stars to estimate the rate of unresolved binary systems by HST in the LMC. We show that, in the case of the search for lenses heavier than 10 times the mass of the Sun, the rate of distant binaries is small enough that its impact on the detection efficiency of microlensing can be neglected.