Subsequent mortality in survivors of Ebola virus disease in Guinea: a nationwide retrospective cohort study.

BACKGROUND: A record number of people survived Ebola virus infection in the 2013-16 outbreak in west Africa, and the number of survivors has increased after subsequent outbreaks. A range of post-Ebola sequelae have been reported in survivors, but little is known about subsequent mortality. We aimed to investigate subsequent mortality among people discharged from Ebola treatment units.

METHODS: From Dec 8, 2015, Surveillance Active en ceinture, the Guinean national survivors' monitoring programme, attempted to contact and follow-up all survivors of Ebola virus disease who were discharged from Ebola treatment units. Survivors were followed up until Sept 30, 2016, and deaths up to this timepoint were recorded. Verbal autopsies were done to gain information about survivors of Ebola virus disease who subsequently died from their closest family members. We calculated the age-standardised mortality ratio compared with the general Guinean population, and assessed risk factors for mortality using survival analysis and a Cox proportional hazards regression model.

FINDINGS: Of the 1270 survivors of Ebola virus disease who were discharged from Ebola treatment units in Guinea, information was retrieved for 1130 (89%). Compared with the general Guinean population, survivors of Ebola virus disease had a more than five-times increased risk of mortality up to Dec 31, 2015 (age-standardised mortality ratio 5·2 [95% CI 4·0-6·8]), a mean of 1 year of follow-up after discharge. Thereafter (ie, from Jan 1-Sept 30, 2016), mortality did not differ between survivors of Ebola virus disease and the general population. (0·6 [95% CI 0·2-1·4]). Overall, 59 deaths were reported, and the cause of death was tentatively attributed to renal failure in 37 cases, mostly on the basis of reported anuria. Longer stays (ie, equal to or longer than the median stay) in Ebola treatment units were associated with an increased risk of late death compared with shorter stays (adjusted hazard ratio 2·62 [95% CI 1·43-4·79]).

INTERPRETATION: Mortality was high in people who recovered from Ebola virus disease and were discharged from Ebola treatment units in Guinea. The finding that survivors who were hospitalised for longer during primary infection had an increased risk of death, could help to guide current and future survivors' programmes and in the prioritisation of funds in resource-constrained settings. The role of renal failure in late deaths after recovery from Ebola virus disease should be investigated.

FUNDING: WHO, International Medical Corps, and the Guinean Red Cross.