Search for: All records

While crew configuration in primary care settings has been studied in terms of its impact on patient outcomes, less is known about how it influences the members' workload experience. This study investigates the workload implications of crew configuration based on members' certification in emergency medical services (EMS). Advanced life support (ALS) ambulance crews are commonly comprised of two paramedics (homogeneous crew) or an emergency medical technician (EMT) and a paramedic (heterogeneous crew). The goals of this study were the following: (1) to investigate differences in workload among members of the same crew, and (2) to use workload assessments to inform crew configuration strategies. We mapped one year of an EMS system's dispatch data to members' workload estimates using the visual, auditory, cognitive, and psychomotor (VACP) approach. We found that lead members (lead paramedics) experience higher workload levels compared to support members (support paramedics or EMTs) in both types of crews. Neither configuration had a consistently lower workload than the other, but differences varied for different shifts and stations. These results informed crew configuration recommendations for stations and shifts in the collaborating system, and in terms of more generalizable variables. A minimum number of staffed crews, half-half shift type (covering both day and night hours), and 30-day frequency of calls with priority P7 most significantly impacted the recommended crew configurations. 

One commonly used workload metric in Emergency Medical Services (EMS) is the Unit Hour Utilization (UHU). The UHU is a productivity measure that, by definition, represents the ratio of patient transport calls to the total hours that ambulances are staffed. It is often misinterpreted as a utilization measure representing the percentage of crews’ available working hours that are spent performing work. This paper investigates a surrogate model to estimate a measure of EMS crew utilization that considers not only call response, but also indirect work tasks, such as documentation and shift start activities. We explored Kriging, KPLS, RBF, and physics-based models based on EMS work dynamics. The true measure of utilization was based on Montecarlo samples of estimated work time patterns associated with a year’s worth of dispatch data augmented with the results of a work measurement study. The best performing model in terms of the root mean square error (RSME), the symmetric mean absolute percent error (sMAPE), and Pearson correlation estimates, was the physics-based model. This model requires time studies to estimate the average time spent in shift start activities and documenting calls, geographic information systems to estimate the average time driving back to the post, and dispatch data analysis to estimate the average time to respond to calls. Sensitivity analysis was used to provide recommendations for when to update these parameters and general recommendations were given to implement this approach in other EMS systems.