You are here
Unreported Cases and Asymptomatic Infection in an Ebola "Hotspot"
Eugene T. Richardson1; J. Daniel Kelly2; Mohamed B. Barrie3; Annelies W. Mesman3; Komba Quiwa3; Sahr Karku3; George Rutherford2; James H. Jones1; Megan Murray4; Paul Farmer4
1Stanford Univ, Stanford, CA, USA;2Univ of California San Francisco, San Francisco, CA, USA;3Partners In Hlth, Freetown, Sierra Leone;4Harvard Med School, Boston, MA, USA
Evidence for asymptomatic Ebola infection is limited to the extent that, during the 2013-15 outbreak, it was not considered epidemiologically relevant to published epidemic models or projections of intervention effects. We conducted a cross-sectional IgG serosurvey in an Ebola ‘hotspot’ village in Sierra Leone, eight months after reported transmission ceased in that location. The surveyed village had a population of approximately 800 individuals distributed among 110 households. Throughout the entire outbreak, there were 25 cases (18 deaths and 7 survivors) reported by the District Ebola Response Center.
We sampled a total of 227 individuals in 30 of 31 previously quarantined households in the village. We assessed anti–glycoprotein IgG responses to Zaire Ebola virus by means of a commercial ELISA kit (Alpha Diagnostic International [ADI]), according to the manufacturer's instructions, with plasma diluted at 1:200. Optical density was read at 450 nm (subtracting OD at 630nm to normalize well background) on a ChroMate 4300 microplate reader. We used Welch’s t-test to determine if mean antibody concentrations differed by exposure history. To discriminate between positive and negative IgG responses (cutoff), we used the mean concentration for individuals without direct contact with a confirmed case plus three standard deviations. We then performed a log-binomial regression using this seropositivity cutoff as the dependent variable and gender, age, occupational activity, and schooling as predictor variables.
We identified an antibody concentration cutoff of ≥ 1.7 U/mL (roughly equal to 5.1 micrograms per mL). All 7 documented survivors demonstrated positive responses (range 2.1 - 6.3 U/mL). Plasma IgG was positive in an additional 30 of 227 quarantined individuals not known to have Ebola virus disease (Figure 1), 27 of whom denied being symptomatic during the period of active transmission in the village. Only having a higher level of education was significantly associated with seropositivity.
This is the first systematic exploration of asymptomatic infection in an Ebola ‘hotspot.’ These data support the hypothesis that the actual number of infections in the 2013-15 outbreak in West Africa is significantly higher than the reported cumulative incidence. The phenomenon of asymptomatic infection has implications for the management of future Ebola outbreaks, as well as for the definition—and treatment—of survivors.