Emergency care of sepsis in sub-Saharan Africa: Mortality and non-physician clinician management of sepsis in rural Uganda from 2010 to 2019

Abstract

Introduction Little data exists from sub-Saharan Africa describing incidence and outcomes of sepsis in emergency units and uncertainty exists surrounding optimal management of sepsis in low-income settings. There exists limited data regarding quality care metrics for non-physician clinicians trained in emergency care. The objective of this study was to describe changes in septic patients over time and evaluate associations between sepsis care and mortality. Methods Secondary analysis of a prospective cohort of all consecutive patients seen from 2010–2019 in a rural Ugandan emergency unit staffed by non-physician clinicians was performed using an electronic database based on paper charts. Sepsis was defined as suspected infection with a quick Sequential Organ Failure Assessment score (qSOFA)≥1. Multi-variable logistic regression was used to analyze three-day mortality. Results Overall, 48,653 patient visits from 2010–2019 yielded 17,490 encounters for patients age≥18 who had suspected infection, including 10,437 with sepsis. The annual proportion of patients with sepsis decreased from 45.0%% to 21.3% and the proportion with malarial sepsis decreased from 17.7% to 2.1% during the study period. Rates of septic patients receiving quality care (“both fluids and anti-infectives”) increased over time (21.2% in 2012 to 32.0% in 2019, pConclusion The largest study of sepsis management and outcomes ever published in both Uganda and sub-Saharan Africa showed sepsis and malarial sepsis decreasing from 2010 to 2019. The increasing quality of non-physician clinician care did not significantly reduce mortality and treatment with “both fluids and antibiotics” increased mortality. With causal associations between antibiotics and mortality deemed implausible, associations between sepsis mortality and interventions likely represent confounding by indication. Defining optimal sepsis care regionally will likely require randomized controlled trials.