Using a simulation-based approach to evaluate a contextually appropriate, non-internet dependent mobile navigation tool for emergency medical dispatch (EMD) of lay first responders (LFRs) in Sierra Leone: A multi-cohort feasibility trial

Abstract

Introduction

Despite disproportionately bearing the global injury burden, low- and middle-income countries often lack emergency medical services(EMS). Equipping lay first responders(LFRs) with emergency medical dispatch(EMD) is a critical next step for formal EMS development. However, few context-appropriate mobile dispatch solutions are available for LFRs, and implementation feasibility and impact on response intervals are not well understood

Materials and Methods

A simulation-based feasibility trial assessed a novel EMD tool, previously used for shipping in resource-limited settings without formal addresses. Two cohorts of 10 non-EMD enabled LFRs trained in 2019 in Sierra Leone were recruited. 100 total simulations were launched in randomized order over 6 months(Cohort 1 distributed along 10 kms of highway(n = 50), Cohort 2 distributed across 24 square-kilometers of an urban setting(n = 50)). On-scene first aid skill performance was assessed under direct observation with a standardized patient actor using checklists. Participants were blinded to randomized dispatch timing/scenario to assess response intervals, replicating real-world conditions, and compared with two-sample t-tests. At six-month follow-up, participants were surveyed on tool ease-of-use and appropriateness, confidence, and ranked dispatch variable relative importance.

Results

Median total response interval (initial notification to LFR arrival on scene) for Cohort 1 for linearly-plotted highway simulations was 6 mins 33 ss(IQR: 2m27 s; 10m48 s), while Cohort 2 for dispersed urban simulations was 6m41s(IQR:3m59 s;14m47 s) (p = 0.720). Median distance between simulated emergency and LFR at the time of notification acceptance=1.675 km(IQR:1.13 km;2.47 km) and 1.73 km(IQR:0.82 km;2.28 km). Mean completion percentage of all discrete first aid steps across all 10 simulation scenario types for Cohort 1 = 89.8 %(IQR: 80 %;100 %) and Cohort 2 = 94.9 %(IQR: 88.89 %;100 %) (p = 0.017). Mean confidence was 9.4/10(median=10) and 9.5/10(median=10)(p = 0.889). 75 % of LFRs (15/20) used the compass feature for navigation while 25 % used turn-by-turn directions (5/20). 70 % LFRs (14/20) reported no unexpected data costs. Emergency location was considered the most important dispatch variable factor, followed by nature/severity of injury.

Discussion

A novel mobile navigation tool for EMD accurately dispatches LFRs to simulated emergency incidents across linear/dispersed settings without significant difference in response interval. Equipping LFRs with EMD tools may facilitate efficient dispatch in resource-limited settings to trauma while expanding emergency care access, meriting further study.