Article Text

Download PDFPDF
G28 Ebola virus disease in children in Sierra Leone: A retrospective cohort study
  1. FC Fitzgerald1,2,
  2. A Naveed2,
  3. K Wing3,
  4. M Gbessay2,
  5. JCG Ross2,
  6. F Checchi4,
  7. D Youkee5,
  8. MB Jalloh6,
  9. DE Baion7,
  10. A Mustapha7,
  11. H Jah8,
  12. S Lako9,
  13. S Oza3,
  14. S Boufkhed10,
  15. R Feury11,
  16. J Bielicki12,
  17. DM Gibb12,
  18. N Klein1,
  19. F Sahr6,
  20. S Yeung13
  1. 1Infection, Immunity, Inflammation and Physiological Medicine, University College London, Institute of Child Health, London, UK
  2. 2Kerry Town Ebola Treatment Centre, Save the Children International, Freetown, Sierra Leone
  3. 3Faculty of Epidemiology & Population Health, London School of Hygiene & Tropical Medicine, London, UK
  4. 4Save the Children International, London, UK
  5. 5Kings Sierra Leone Partnership, Freetown, Sierra Leone
  6. 634 Military Hospital, Republic of Sierra Leone Armed Forces, Freetown, Sierra Leone
  7. 7Ola During Children’s Hospital, Sierra Leone Ministry of Health, Freetown, Sierra Leone
  8. 8Cap Anamur (German Emergency Doctors), Ola During Children’s Hospital, Freetown, Sierra Leone
  9. 9Welbodi Partnership, Ola During Children’s Hospital, Freetown, Sierra Leone
  10. 10Department of Global Health and Development, London School of Hygiene & Tropical Medicine, London, UK
  11. 11Live Case Management, Western Area Emergency Response Centre, Freetown, Sierra Leone
  12. 12University College London Medical Research Council Clinical Trials Unit, University College London, London, UK


Background The Ebola virus disease (EVD) outbreak in West Africa has claimed over 11000 lives with >27000 cases. Infected children <5 years have a high mortality rate but little is known about disease manifestations and possible predictors of outcome. The impact of epidemiological, clinical and operational factors on mortality is unknown.

Method A retrospective observational cohort was designed including all children <12 years admitted to 11 Ebola Holding Units (EHUs) in the Western Area of Sierra Leone from 1.8.14–1.4.15. Figure 1 describes patient flow. Primary outcome was death or discharge from care with follow up after transfer to specialist Ebola Treatment Centres (ETCs). Retrospective data were collected from the sites from admission books, case investigation forms and clinical records, cross-referenced with district-wide laboratory results, burial records, child protection records and interviewing staff in the absence of documentation. Follow up telephone calls were made to guardians post-discharge.

Results 309 children aged 2 days to 12 years were admitted to EHUs and tested positive for EVD. Outcomes were available for 282 (91%). Mortality was 57% with 55% occurring at EHUs and 45% at ETCs. Lower age was a strong predictor of mortality (multivariable OR 1.74, 95% CI 1.04–2.93 comparing children aged 0- <5 years with those aged 5 - <12) (Table 1), as was presenting with diarrhoea (multivariable OR 1.91, 95% CI 1.08–3.39). Blood test results show marked derangement of renal function and inflammatory markers with significant hypoglycaemia. Up to 60% of children may have been admitted unaccompanied, and 56% of children were transferred between 5–380km to the nearest available ETC bed.

Abstract G28 Table 1

Descriptive and univariable analysis of all children who attended an Ebola Holding Unit (EHU)

Conclusion This is the most comprehensive paediatric cohort of EVD to date, and the only cohort to incorporate data from both EHUs and ETCs. The importance of data sharing between sites cannot be overstated in order to identify potentially modifiable risk factors in both adults and children and guide future operational responses.

Abstract G28 Figure 1

Flowchart of patients attending health care facilities in Freetown, Sierra Leone

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.