In recent years, there has been a notable increase in the frequency and sophistication of attacks on complex networks, which has posed significant threats to various domains, including critical infrastructure, healthcare systems, and financial networks. The severity of targeted attacks on complex networks is commonly measured by the variation of the largest connected component (LCC) after each attack. However, the standard measure of severity in terms of LCC does not take into account the iteration in which these variations occur. We propose a modified measure that incorporates the idea of weighting the LCC variation based on the iteration in which it occurs. Our proposed measure applies a weighted average to the LCC variation based on the iteration in which it occurs. To validate the accuracy of this measure, we conducted a simulation study on scale-free networks. Our results, while indicating a moderate impact on current simulations, reveal potential limitations attributed to the limited variations in centrality values. We posit that the observed trends could be further elucidated in real-world networks characterized by more dynamic changes.