We derive a closed-form expression that accurately predicts the peak frequency-shift and broadening induced by tiny perturbations of plasmonic nanoresonators without critically relying on repeated electrodynamic simulations of the spectral response of nanoresonator for various locations, sizes or shapes of the perturbing objects. The force of the present approach, in comparison with other approaches of the same kind, is that the derivation is supported by a mathematical formalism based on a rigorous normalization of the resonance modes of nanoresonators consisting of lossy and dispersive materials. Accordingly, accurate predictions are obtained for a large range of nanoparticle shapes and sizes, used in various plasmonic nanosensors, even beyond the quasistatic limit. The expression gives quantitative insight, and combined with an open-source code, provides accurate and fast predictions that are ideally suited for preliminary designs or for interpretation of experimental data. It is also valid for photonic resonators with large mode volumes. KEYWORDS: refractive index sensing, surface plasmon resonance, quasi normal mode, nanoparticle, plasmonics Introduction. In recent years, metallic nanoparticles have gained a lot of attention and also witnessed successful applications in various fields of nanosciences. As their near-fields support strong and highly-confined resonances, metallic nanoparticles can effectively convert local changes of refractive index into frequency shifts of the resonance.