Measurement of dose or dose perturbation factors at high atomic number interfaces are usually performed with a thin-window parallel-plate ion chamber. In a transition region, under nonequilibrium conditions, accuracy of ion chamber readings for the dose measurements has often been questioned. This paper critically analyzes the factors (stopping power ratio and charge collection) for the dose measurements at interfaces. Monte Carlo simulations were performed to investigate the secondary electron spectrum produced by photon beams and to calculate the stopping power ratios at the point of measurement. The validity of dose measurements was studied for the photon beams in the range of Co-60 gamma rays to 24-MV x rays at bone and lead interfaces with polystyrene, using thermoluminescent dosimeters, extrapolation chamber and several types of commercially available parallel-plate ion chambers. It is observed that for energies greater than 10 MV most parallel-plate chambers can be used to measure dose accurately. At lower energies, however significant differences between measured doses with different detectors were noticed. It is suggested that at high-Z interfaces and lower energies, the dose measurements should be performed with ultrathin-window parallel-plate ion chambers or extrapolation chambers.