Emergency Medical Services (EMS) play a critical role in the provision of healthcare across the globe. Access to pre-hospital treatment and timely arrival at emergency departments can be the difference between life and death. Time-sensitive medical emergencies are a particular major health concern in low and middle-income countries (LMICs), comprising one third of all deaths. Indeed, there are many challenges for providing timely pre-hospital and lifesaving emergency care across Indonesia, a country of 275 million people. These include vast geographical areas, heavily congested cities, frequent natural disasters and, when we started our work back in October 2019 funded by an initial EPSRC Mathematical Sciences GCRF grant, inadequate numbers of ambulances and a lack of a co-ordinated service. Consequently, pre-hospital mortality (for example from cardiac arrests, strokes and road traffic accidents) is high and the ability to respond to major events such as earthquakes and volcanic eruptions is also limited and thus the chances of survival are low.
Our research programme is in collaboration with various government authorities and healthcare providers to explore the feasibility, design and implementation of a free to use, co-ordinated emergency ambulance service. Whilst in Indonesia there exist many private providers, albeit with limited capacity, fragmented and not terribly equipped to save lives but rather transport patients, the focus of our research aims to help to reduce health inequalities and provide for the poorest in society who cannot afford to take privately run ambulances.
Although there is a sizeable body of literature relating to mathematical modelling of EMS systems, to-date it has predominantly focussed on high-income countries, for single vehicle types, and performance management associated with response time targets (e.g. eight minutes for the most critical category of incident). The challenges presented in Indonesia suggested that new approaches would be necessary: traffic can be extremely unpredictable, road congestion is common, multiple types of emergency vehicles can potentially be deployed (e.g. motorcycles, tuk-tuk, minivans), patient outcomes are more meaningful than response times, and demand data is not routinely collected and disparate.
Our initial focus was on the capital, Jakarta, with a population of 11 million. We started our work with Ambulans 118, established in 2005 by the Indonesian Surgeons’ Association, the only charitable ambulance service in the country but with limited numbers of vehicles. In the absence of much understanding about pre-hospital demand and medical needs, we carried the first known surveys in Indonesia involving more than 2,000 respondents attending across five major Emergency Departments in Jakarta. Our survey findings [1] shed light on the barriers to the use of EMS such a lack of awareness of available services, high costs of private providers, and lack of awareness on the symptoms that warrant the use of ambulance. It was shocking to understand how even the most critical patients typically travelled to the hospital in taxis, on the back of motorbikes, or even via public transportation.
We subsequently turned our attentions and research work to gather what demand data was available alongside our own survey work, to create geospatial forecasts of emergency demand, and to build interfaced optimisation and simulation tools to help make critical decisions on the types, capacities and location-allocations of emergency vehicles. A particular novelty of our work (building on our previous related EMS modelling research [2, 3, 4, 5]) has been to explicitly consider EMS allocations for both heterogeneous populations (multiple medical needs) using a heterogeneous fleet (consideration of multiple vehicle types each with differing travel speeds using time-dependent TomTom data).
Our work has directly feed into decisions by the health authority in Jakarta to establish a new organisation, PK3D, to provide a unified free to use emergency ambulance service (with a single number, 119, to call), currently comprising of 80 new ambulances and 20 paramedics on motorbikes. Our work has more recently highlighted to PK3D the importance of the motorbikes to act as a rapid response (and of course can travel more quickly through a heavily congested city than a minivan) and we have quantified where further investment would be beneficial to increase capacity and where to optimally locate this. Ambulans 118 have also kindly named a new ambulance after the Cardiff School of Mathematics in honour of our collaboration (see photo below)!
We wish to thank the The Engineering and Physical Sciences Research Council (EPSRC) who initially funded our research (EP/T003197/1) and for follow-on funding from HEFCW ODA funds. We would like to express our sincere gratitude to staff at Ambulans 118 and PK3D for their support, insights and provision of data. In particular we wish to thank Professor Aryono Djuned Pusponegoro and Ms Asti Puspita Rini, Founder and Director of 118 Ambulance Service Foundation, and Dr Winarto, Director General of PK3D. We continue our collaboration with 118 and PK3D to monitor the uptake of the new 119 number and changes in survival rates, to reassess periodically the resource needs and location-allocations of EMS vehicles across Jakarta, to continue to provide training in OR methods, and we hope to gather data from other cities to assist EMS planners across Indonesia.
The following staff in the OR Group at Cardiff University have all contributed to the research: Dr Sarie Brice, Prof. Daniel Gartner, Prof. Paul Harper, Prof. Vince Knight, Dr Geraint Palmer and Dr Mark Tuson.
Prof. Paul Harper (Cardiff University) at the ambulance launch with Prof. Pusponegoro (Founder and Director of 118) and the Director General of Health Workers from the Ministry of Health.
References
[1] Brice, S. et al. 2022. Emergency services utilization in Jakarta (Indonesia): A cross-sectional study of patients attending hospital emergency departments. BMC Health Services Research 22, article number: 639. (10.1186/s12913-022-08061-8) https://bmchealthservres.biomedcentral.com/articles/10.1186/s12913-022-08061-8
[2] Knight, V. A., Harper, P. R. and Smith, L. 2012. Ambulance allocation for maximal survival with heterogeneous outcome measures. OMEGA -The International Journal of Management Science. 40(6), pp. 918-926. (10.1016/j.omega.2012.02.003) https://www.sciencedirect.com/science/article/abs/pii/S0305048312000436?via%3Dihub
[3] Palmer, G. I., Knight, V. A., Harper, P. R. and Hawa, A. L. 2019. Ciw: An open-source discrete event simulation library. Journal of Simulation 13(1), pp. 68-82. (10.1080/17477778.2018.1473909) https://www.tandfonline.com/doi/full/10.1080/17477778.2018.1473909
[4] Vile, J. L., Gillard, J. W., Harper, P. R. and Knight, V. A. 2012. Predicting ambulance demand using singular spectrum analysis. Journal of the Operational Research Society 63, pp. 1556-1565. (10.1057/jors.2011.160) https://www.tandfonline.com/doi/full/10.1057/jors.2011.160
[5] Knight, V. A. and Harper, P. R. 2012. Modelling emergency medical services with phase-type distributions. Health Systems 1(1), pp. 58-68. (10.1057/hs.2012.1) https://www.tandfonline.com/doi/full/10.1057/hs.2012.1