Cascading outages in power systems can lead to major power disruptions and blackouts in power systems. By taking Manual Corrective Actions (MCAs), operators could be able to mitigate a cascading outage following an initiating event. However, due to stressful and time-constrained situations, they might not be able to take appropriate corrective actions in time and might even take counter-productive actions. Although numerous approaches have been developed to assess the risk of cascading outages in a probabilistic way, they generally do not consider in a realistic manner MCAs, including their imperfection. This paper aims to address that gap by proposing a Human Reliability Analysis-Optimal Power Flow (HRA-OPF) framework. The developed approach is applied to the New England Test System (NETS) and to the Reliability Test System (RTS). The risks of loss of supplied power are compared for different possibilities: no MCAs, perfect MCAs, and imperfect MCAs.