Prognostic Indicators for Ebola Patient Survival

The epidemic of Ebola virus (Zaire ebolavirus) disease (EVD) in West Africa began in eastern Guinea in December 2013 (1) and quickly spread into Liberia and Sierra Leone, eventually overwhelming the fragile healthcare infrastructures in these countries (2). During the peak of the epidemic, many healthcare facilities were quickly filled beyond capacity, which often forced clinicians to make difficult decisions about how to triage patients and how to manage patient and family expectations regarding probable outcomes. Reliable prognostic indicators available at the time of patient admission could help clinicians make these decisions. We therefore assessed the reliability of 2 potential prognostic indicators: 1) the total elapsed time from reported symptom onset to healthcare facility admission and 2) cycle threshold (Ct), which can serve as an approximation of viral load, at the time of EVD diagnosis. Early treatment, which is made possible by early admission, is thought to improve chances of survival (3–5), but there is little supporting empirical evidence. Analyses of EVD patients in Ebola treatment units (ETUs) have shown that Ct values predict outcomes (6–8), but these studies do not account for those who died before ETU admission. By using onset-to-outcome data for all identified EVD patients during a 4-month period in Bo District, Sierra Leone, we explored the extent to which these indicators predicted outcome.