Ebola virus screening during pregnancy in West Africa: unintended consequences

Objective: We hypothesized that predictive value of traditional Ebola virus disease (EVD) screening in West African pregnant women is low because febrile and hemorrhagic complications of pregnancy that can mimic EVD are common. Methods: Proportions of various categories of pregnancy loss from a hypothetical cohort of West African gravidas were used to construct a Kaplan-Meier curve. The incidence rate of each category was determined by multiplying its proportion by the overall incidence rate, calculated from the inverse of the area under the curve. Incidence rates of Ebola-like illnesses during pregnancy were obtained by multiplying their percentages by the incidence rates of categories of loss with which they coincide. Results: During pregnancy about 1.5% of suspected EVD cases would prove to have EVD. Most of the remaining 98.5% would have hemorrhagic and febrile complications of pregnancy. Conclusion: Current guidelines consider obstetrical interventions inappropriate in suspected EVD during pregnancy. However, because the overwhelming majority of cases suspected by screening do not have EVD and might benefit from obstetrical intervention, policy makers must consider whether the small risk to properly protected health care workers from the 1.5% with true EVD justifies withholding potentially life-saving care from the 98.5% who ultimately test negative for EVD.