Predicting the effects of introducing an emergency transport system in low-income and middle-income countries: a spatial-epidemiological modelling study

Katie Scandrett*, Richard Lilford, Dmitri Nepogodiev, Srinivasa vittal Katikireddi, Justine Davies, Stephen Tabiri, Samuel i Watson

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

8 Downloads (Pure)

Abstract

Introduction Many low-income and middle-income countries lack an organised emergency transportation system, leaving people to arrange informal transport to hospital in the case of a medical emergency. Estimating the effect of implementing an emergency transport system is impractical and expensive, so there is a lack of evidence to support policy and investment decisions. Alternative modelling strategies may be able to fill this gap.

Methods We have developed a spatial-epidemiological model of emergency transport for life-threatening conditions. The model incorporates components to both predict travel times across an area of interest under different scenarios and predict survival for emergency conditions as a function of time to receive care. We review potentially relevant data sources for different model parameters. We apply the model to the illustrative case study of providing emergency transport for postpartum haemorrhage in Northern Ghana.

Results The model predicts that the effects of an ambulance service are likely to be ephemeral, varying according to local circumstances such as population density and road networks. In our applied example, the introduction of the ambulance service may save 40 lives (95% credible interval 5 to 111), or up to 107 lives (95% credible interval −293 to –13) may be lost across the region in a year, dependent on various model assumptions and parameter specifications. Maps showing the probability of reduced transfer time with the ambulance service may be particularly useful and allow for resource allocation planning.

Conclusions Although there is scope for improvement in our model and in the data available to populate the model and inform parameter choices, we believe this work provides a foundation for pioneering methodology to predict the effect of introducing an ambulance system. Our spatial-epidemiological model includes much oppurtunity for flexibility and can be updated as required to best represent a chosen case study.
Original languageEnglish
Article numbere000321
JournalBMJ Public Health
Volume2
Issue number1
DOIs
Publication statusPublished - 21 Feb 2024

Bibliographical note

Funding: JD was funded by National Institute of Health and Care Research (NIHR) Global Health Group on Equitable Access to Quality Health Care for Injured People in Four Low or Middle Income Countries: Equi-injury; NIHR133135. RL and JD were funded by NIHR Research and Innovation for Global Health Transformation (RIGHT) grant—Rwanda912: Use of an innovative electronic communications platform to improve prehospital transport of injured people in Rwanda; NIHR203062. RL and SIW were funded by NIHR Applied Research Collaboration West Midlands; NIHR200165. RL, DN and ST funded by NIHR Global Health Research Unit Global Surgery Unit; NIHR133364. RL and SIW were funded by NIHR Global Health Research Unit on Improving Health in Slums; 16/136/87. SVK was funded by Medical Research Council (MC_UU_00022/2), Scottish Government Chief Scientist Office (SPHSU17) and European Research Council (949582). RL, KS and SVK were funded by NIHR Global Health Research Unit on Social and Environmental Determinants of Health Inequalities; NIHR134801. RL and SIW were funded by NIHR Transforming the Treatment and Prevention of Leprosy and Buruli ulcers in Low and Middle-Income Countries (LMICs); NIHR200132. KS was funded by the NIHR Birmingham Biomedical Research Centre; S-BRC-1215-20009.

Fingerprint

Dive into the research topics of 'Predicting the effects of introducing an emergency transport system in low-income and middle-income countries: a spatial-epidemiological modelling study'. Together they form a unique fingerprint.

Cite this