Skip to content
HEAL-SL

Rapid monitoring of Ebola symptoms within Sierra Leone’s national mortality surveillance platform, June–August 2026

A six-question Ebola symptom screen added to routine HEAL-SL household visits in 56 sampling units across five districts, with symptom levels benchmarked against verbal-autopsy deaths from causes unrelated to Ebola.

OngoingFrom 8 Jun 2026Kono, Koinadugu, Bonthe, Western Area Rural, Kenema
Map of Sierra Leone showing the 56 HEAL-SL sampling units screened for Ebola symptoms, against all 661 units

Key findings

  • 3,887 households screened in 56 sampling units across five districts.
  • Fever reported by 2.0% of households and unresolving bleeding by 0.3%.
  • No household reported travel from the outbreak countries.
  • Two households matched the WHO suspected-case pattern (0.05%).
  • Symptom levels fell below those among injury deaths, and far below confirmed Ebola case series. Screening continues.

About this study

Background

In May 2026, the WHO declared an outbreak of Ebola disease caused by Bundibugyo virus in the Democratic Republic of the Congo and Uganda a public health emergency of international concern. Syndromic surveillance of key symptoms, particularly fever, vomiting and unexplained bleeding, provides a crude but useful profile of whether Ebola cases are occurring. Among confirmed and probable cases with a known outcome in the first nine months of the 2014–16 West Africa epidemic, fever was recorded in 87%, vomiting in 68% and unexplained bleeding in 18% (WHO Ebola Response Team, NEJM 2014).

Methods

A cross-sectional screen was administered once per household during routine HEAL-SL Round 6 visits in the sampling units the survey reached between 8 June and 23 August 2026. The instrument comprised six questions, covering household fever in the past 21 days (with a symptom checklist if yes), unresolving bleeding, international travel in the past 21 days, sudden death after fever, bleeding or vomiting in the past four weeks, and, from mid-June, the respondent's funeral attendance in the past two weeks.

One respondent answered for each household. The form does not record the age or identity of affected persons, and the objective was not case detection: the screening collected data during routine visits and did not include referral or follow-up of households reporting symptoms.

Results

HEAL-SL reached 56 sampling units across five districts. These units are randomly selected villages or small urban blocks, drawn from the enumeration areas of the 2015 Sierra Leone census. In total, 3,887 households were screened, about three-quarters of those listed in the previous round.

Seventy-seven households (2.0%, 95% CI 1.6–2.5) reported at least one member with fever in the past 21 days, and 12 (0.3%) reported unresolving bleeding. Twenty (0.5%) reported a sudden death after fever, bleeding or vomiting in the past four weeks. Ten (0.3%) reported international travel in the past 21 days, all to Liberia or Guinea (five each) and none to the Democratic Republic of the Congo or Uganda.

Fever here means any fever, which is broader than the WHO alert definition. Only two households (0.05%, 95% CI 0.01–0.19) reported fever with three or more listed symptoms, the WHO suspected-case pattern.

All three symptoms were rarer in screened households than among injury deaths in earlier HEAL-SL verbal autopsies (fever 10.2%, bleeding 2.5%, vomiting 2.3%), and far rarer than in confirmed Ebola case series.

What this means

Across 3,887 households in 56 sampling units in five diverse districts of Sierra Leone, reports of the symptoms that would signal Ebola were uncommon. Ebola related symptoms are therefore uncommon in the five districts screened at the time of the first data analysis. HEAL-SL, a national surveillance platform, is able to rapidly gather relevant data on a syndromic question within weeks.

Limitations

For speed, answers were recorded only at household level: the affected person's age and identity were not recorded, so results cannot be age-stratified. The funeral question lacks answers for 20% of households. Collection ran 8-15 June and 8-23 August with 21-day and four-week recall windows, so events from mid-June to mid-July fall outside every window. The results do not represent the whole country, but do cover the likely areas of mobility if Ebola cases were to be imported from Guinea or Liberia.

Outputs

  • Household Ebola symptom prevalence from June–August 2026 within the HEAL-SL mortality-surveillance platform

    In reviewReport · September 2026

    Submitted to the National Public Health Agency of Sierra Leone.

  • Journal Comment on the Ebola symptom screening

    In preparationComment

    A link will be added here once it is published.

Related links

Study team

Partners
  • Njala University
  • National Public Health Agency of Sierra Leone
  • Centre for Global Health Research, University of Toronto
  • University of Oxford
Study leads
  • Prof Rashid Ansumana — Njala University
  • Prof Foday Sahr — National Public Health Agency of Sierra Leone
  • Prof Prabhat Jha — University of Oxford