In this line of research, the authors have developed an advanced PRA methodology, the Integrated PRA (I-PRA) framework, which explicitly incorporates the underlying failure mechanisms into PRA scenarios by integrating the spatio-temporal simulation of underlying physical and social phenomena with classical PRA. The focus of this paper is on developing an Importance Measure (IM) method for I-PRA. The classical IM methods (e.g., Fussell-Vesely IM and Risk Achievement Worth), which are common in the PRA field, are not adequate for I-PRA because they only focus on the risk ranking of components. In I-PRA, the risk importance ranking of input parameters within the simulation models needs to be analyzed and, for that purpose, a moment-independent Global IM, the cdf-based sensitivity indicator Si(CDF), is selected and tailored for the I-PRA framework. This IM method can capture three key aspects of the I-PRA model: (i) uncertainty associated with the input parameters, (ii) uncertainty of risk outputs, and (iii) non-linearity and interactions among input parameters within the simulation model. This paper shows the progress of the ongoing research, and a case study using a reduced-order I-PRA to demonstrate the feasibility of implementing the Global IM method in a realistic PRA application, is presented.