The efficient management of ambulance routing for emergency requests is vital to save patients when a disaster response scenario occurs. In todays road traffic, the transportation of patients from emergency points becomes more difficult in such tragic situation. We consider an urgent situation where a lot of hurts/patients require simultaneously an urgent medical care. Patients are either (1) slightly injured which can be assisted on the spot or (2) seriously injured which should be transferred to hospitals. In this paper, we aim to enhance the response-time performance of emergency medical service providers by handling the ambulance routing problem (ARP). The problem can be modeled as either the Open Vehicle Routing Problem (OVRP) or a Vehicle Routing Problem with Pickup and Delivery (VRPPD). We propose a cluster-first route-second algorithm based on the Petal algorithm and the particle swarm optimization (PSO) approach in order to handle efficiently the ARP.