Ambulance allocation for maximal survival with heterogeneous outcome measures

• Published in: Omega (Oxford) Vol. 40; no. 6; pp. 918 - 926
• Main Authors: Knight, V.A., Harper, P.R., Smith, L.
• Format: Journal Article
• Language: English
• Published: Kidlington Elsevier Ltd 01.12.2012; Elsevier; Pergamon Press Inc
• Series: Omega
• Subjects: Ambulance allocation; Ambulance allocation Health care modeling Survival function Patient outcome Covering model; Applied sciences; Biological and medical sciences; Covering model; Emergency medical care; Exact sciences and technology; Health and social institutions; Health care modeling; Logistics; Medical sciences; Operational research and scientific management; Operational research. Management science; Optimization algorithms; Patient outcome; Patients; Public health. Hygiene; Public health. Hygiene-occupational medicine; Queuing theory; Queuing theory. Traffic theory; Reliability theory. Replacement problems; Stochastic models; Studies; Survival function; United Kingdom > UK; Wales; Health care modeling; Survival function; Patient outcome; Covering model; Ambulance allocation; Medical transport; Modeling; Location problem; Survival; Emergency department; Response time; Ambulance; Time average; Survival probability; Localization; Queue; Public health
• ISSN: 0305-0483; 1873-5274
• DOI: 10.1016/j.omega.2012.02.003

This paper proposes new models for locating emergency medical services (EMS) by incorporating survival functions for capturing multiple-classes of heterogeneous patients. The Maximal Expected Survival Location Model for Heterogeneous Patients (MESLMHP) aims to maximize the overall expected survival probability of multiple-classes of patients, whereby different classes could be defined according to agreed patient categories based on response time targets, or by capturing differing medical conditions each with a corresponding survival function. Furthermore, we propose and demonstrate an approximation approach to solving the extended stochastic version of MESLMHP, which utilizes queuing theory to permit the modeling of congestion and utilization at each ambulance station, and does not require assumptions to be made on the utilization of ambulances. Both models are demonstrated using data from the ambulance service in Wales. We show that our multiple outcome measures and survival-maximizing approach, rather than one based on average response time targets alone or a single patient class provides more effective EMS ambulance allocations. ► We propose a new model for locating emergency medical services. ► Our model takes in to account survival of multiple patient classes. ► A novel solution methodology is proposed to take in to account stochasticity. ► The model is demonstrated using data from a national ambulance service. ► Comparisons are offered that highlight the importance of our approach.