Sepsis in two hospitals in Rwanda: A retrospective cohort study of presentation, management, outcomes, and predictors of mortality

• Published in: PloS one Vol. 16; no. 5; p. e0251321
• Main Authors: Hopkinson, Dennis A., Mvukiyehe, Jean Paul, Jayaraman, Sudha P., Syed, Aamer A., Dworkin, Myles S., Mucyo, Willy, Cyuzuzo, Thierry, Tuyizere, Anne, Mukwesi, Christian, Nyirigira, Gaston, Banguti, Paulin R., Riviello, Elisabeth D.
• Format: Journal Article
• Language: English
• Published: San Francisco Public Library of Science 26.05.2021; Public Library of Science (PLoS)
• Subjects: Anesthesia; Archives & records; Bilirubin; Biology and Life Sciences; Blood; Blood pressure; Care and treatment; Cirrhosis; Cohort analysis; Comorbidity; Critical care; Documentation; Editing; Emergency medical care; Emergency medical services; Encephalopathy; Gynecology; Health care facilities; Health sciences; Hepatic encephalopathy; Infections; Intensive care; Internal medicine; Laboratories; Liver; Liver cirrhosis; Liver diseases; Low income groups; Medical records; Medical schools; Medicine; Medicine and Health Sciences; Methodology; Mortality; Obstetrics; Patient outcomes; Patients; Pediatrics; Respiration; Respiratory rate; Sepsis; Surgery; Teaching hospitals; Vital signs; Rwanda; Haiti; United States > US; Africa; Uganda; Kenya; Virginia
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0251321

Few studies have assessed the presentation, management, and outcomes of sepsis in low-income countries (LICs). We sought to characterize these aspects of sepsis and to assess mortality predictors in sepsis in two referral hospitals in Rwanda. This was a retrospective cohort study in two public academic referral hospitals in Rwanda. Data was abstracted from paper medical records of adult patients who met our criteria for sepsis. Of the 181 subjects who met eligibility criteria, 111 (61.3%) met our criteria for sepsis without shock and 70 (38.7%) met our criteria for septic shock. Thirty-five subjects (19.3%) were known to be HIV positive. The vast majority of septic patients (92.7%) received intravenous fluid therapy (median = 1.0 L within 8 hours), and 94.0% received antimicrobials. Vasopressors were administered to 32.0% of the cohort and 46.4% received mechanical ventilation. In-hospital mortality for all patients with sepsis was 51.4%, and it was 82.9% for those with septic shock. Baseline characteristic mortality predictors were respiratory rate, Glasgow Coma Scale score, and known HIV seropositivity. Septic patients in two public tertiary referral hospitals in Rwanda are young (median age = 40, IQR = 29, 59) and experience high rates of mortality. Predictors of mortality included baseline clinical characteristics and HIV seropositivity status. The majority of subjects were treated with intravenous fluids and antimicrobials. Further work is needed to understand clinical and management factors that may help improve mortality in septic patients in LICs.