Fatalities due to road crashes are rising globally, with the situation being particularly severe in low‐income countries. Traditional investigation methods have reached a ceiling in terms of reducing road traffic fatalities, hence safety researchers have begun to explore systems‐based approaches. In this study, two bus‐pedestrian collisions occurring in Bangladesh and separated by a time span of 13 years have been analyzed using the Accimap methodology. Building on and going further than the Accimaps of the two collisions, the recommendations that have been proposed to reduce the recurrence of such collision types have been graphically represented in a combined Preventimap. This graphically displays the inter‐connectedness of the countermeasures. Further, the countermeasures have been mapped against six road safety E's, namely engineering, education, enforcement, economics, emergency response, and enablement. Comparing the countermeasures arising from each of the two collisions showed that some preventive strategies are unique to one collision while others are common to both collisions. With a significant proportion of the proposed recommendations being of similar types, we discuss the idea that had prompt and appropriate intervention measures been implemented at different levels of the system after the first collision, the probability of the occurrence of the second collision would have been significantly less. The same would apply to similar collisions in the future. We also discuss the benefits of the Preventimap representation and its potential uses in road safety policy and practice.

Road collision types repeat themselves, especially in low-and middle-income countries (LMICs), where countermeasures are often improvised and implemented with little planning. At the Shahbag intersection in Dhaka, Bangladesh, speed bumps were quickly constructed at the exit of the intersection as an improvised road safety measure following the occurrence of a fatal collision, which eventually contributed to another collision between a truck and a car. The events influencing the improvisation decision, and that action's consequences, have been analysed using the Impromap methodology, a variation of the Accimap approach that focusses specifically on improvisation. The applicability of the Impromap as a systems-based approach to the road safety domain is assessed using the predictions described in Rasmussen's risk management framework, and corresponding countermeasures are proposed. The analysis shows that improvisation in the road safety domain is undesirable irrespective of the economic setting as it is likely to eventually contribute to secondary collisions.