Ebola Infection vs. Exposure: What Contact Tracing Actually Tracks
Not every Ebola exposure leads to infection. Contact tracing monitors everyone who was exposed — but only a fraction will become infected. Understanding the difference is essential to understanding how outbreak response works, and where it fails.
Two Numbers That Define Every Outbreak
Every Ebola outbreak produces two parallel sets of numbers that are often conflated in news coverage: exposures and infections. Contact tracing teams track both, but they represent fundamentally different things.
An exposure is any interaction with a confirmed or probable Ebola case that carries a non-negligible risk of viral transmission. An infection is a confirmed case of Ebola virus disease, diagnosed by RT-PCR or antigen testing.
The gap between these two numbers is enormous. In the 2018–2020 DRC Kivu outbreak, response teams monitored approximately 130,000 contacts across the duration of the outbreak. Confirmed cases totalled 3,481. That means roughly 2.7% of tracked contacts developed confirmed infections — and even accounting for contacts who were missed, the infection rate among all people who had meaningful exposure was far below 100%.
This gap is not a failure of the outbreak response. It is the biological and epidemiological reality of how Ebola transmits — and understanding it is essential to understanding why contact tracing works as an outbreak control strategy.
What “Contact” Actually Means
In Ebola response, a contact is defined by WHO as:
A person who has been exposed to an Ebola case within the 21 days before or after the case became ill, through any of the following: sleeping in the same household, direct physical contact (touching), or contact with body fluids.
This definition is intentionally broad. Response teams do not have the luxury of knowing in real-time whether a specific exposure was sufficient to transmit virus. The contact list therefore captures everyone in the risk window and monitors them for 21 days — the maximum incubation period.
Not all contacts carry equal risk. Ebola response frameworks classify contacts by risk tier:
| Risk Tier | Definition | Example |
|---|---|---|
| High risk | Direct contact with blood, body fluids, or unprotected patient care | Unprotected family care; needlestick |
| Medium risk | Prolonged proximity to symptomatic case without direct fluid contact | Shared room; household member |
| Low risk | Brief, non-physical proximity | Same building; brief visit without touch |
| No risk | Casual contact with asymptomatic person; no fluid exposure | Passing in street; social acquaintance |
High-risk contacts are prioritised for intensive daily monitoring and are offered ring vaccination where available. Low-risk contacts may be monitored at lower frequency or with self-reporting protocols.
Why Most Exposures Don’t Lead to Infection
Several biological and situational factors determine whether an exposure results in infection:
1. Viral Load at the Time of Exposure
Ebola viral load in body fluids is not constant throughout the disease course. It is very low early in illness, rises through the first week, and peaks during the haemorrhagic phase — typically days 5–10. A contact who interacted with a case during the early, low-viral-load phase has a much lower infection probability than one who provided care during the peak haemorrhagic period.
Analysis of household transmission data from the 2014–2016 epidemic showed that contacts of cases who died (reflecting the most severe, highest-viral-load disease) had significantly higher secondary attack rates than contacts of survivors.
2. Route and Duration of Exposure
Ebola requires direct contact with infected fluids — it does not spread through the air under natural conditions. A contact who shared a household for a week but only had fleeting physical contact carries a much lower risk than one who washed or bathed the patient.
Quantifying this is difficult, but the secondary attack rate among household contacts in well-studied outbreaks ranges from 5% to 20% depending on the nature and duration of care provided. Not every household member of a confirmed case will become infected.
3. Individual Immunological Variation
Emerging evidence suggests that host genetic factors influence susceptibility to Ebola infection. Studies of household contacts who had documented exposure but never developed disease found that some showed detectable anti-Ebola antibodies — suggesting subclinical or abortive infection that the immune system cleared before full clinical disease established.
A 2015 study published in PLOS Pathogens estimated that up to 25% of household contacts of confirmed cases in the 2014 epidemic showed serological evidence of past Ebola exposure without documented illness. The existence of this “seropositive without disease” population suggests that infection following exposure is not inevitable, and that immune factors can terminate infection before the disease threshold is crossed.
This finding has significant implications for vaccine development and epidemiological modelling — it means the true infection-to-exposure ratio is even lower than confirmed case counts suggest.
Contact Tracing as a Mathematical Intervention
Understanding the exposure-infection gap explains why contact tracing is mathematically effective as an outbreak control strategy, even with an imperfect tracing rate.
Consider a simplified model:
- One confirmed case generates an average of 20 traceable contacts
- Of those 20, 2–4 will become infected (10–20% secondary attack rate)
- If contact tracing identifies and isolates those 2–4 people before they become symptomatic, zero onwards transmission occurs from that chain
The critical condition is identifying contacts before symptom onset — during the incubation period. A contact who is found and isolated on day 5 of their 8-day incubation period cannot infect anyone. A contact who is not found, develops symptoms on day 8, attends a family gathering on days 8–9 before seeking care, and is diagnosed on day 10, may already have created a new exposure cluster.
This is why the speed of contact identification is the single most important operational metric in Ebola response, more so than the breadth of tracing. A fast but incomplete contact list that finds 80% of contacts within 48 hours outperforms a thorough but slow list that finds 95% of contacts in 5 days.
The 21-Day Monitoring Window: Why That Number
The 21-day monitoring period is derived from the maximum observed incubation period of Ebola. Analysis of thousands of cases across multiple outbreaks places the 99th percentile of incubation at approximately 21 days, with a median of 8–10 days.
In practice, this means:
- Most contacts who are going to develop disease will do so within 12–14 days of exposure
- A contact who reaches day 17–18 asymptomatic is very likely to be cleared
- Day 21 is a conservative outer boundary that captures essentially all possible cases
The 21-day rule has practical consequences for communities, healthcare systems, and economies. Healthcare workers who are contacts of confirmed cases may be placed on modified duty restrictions for 21 days. Travellers from affected areas may face monitoring requirements for 21 days. The boundary is conservative by design.
Where the Exposure-Infection Framework Breaks Down
The framework above assumes accurate exposure identification — which is frequently not achievable in real outbreak conditions.
Missed contacts
Contact tracing completeness in the 2014–2016 epidemic is estimated to have been around 60–70% in Guinea and Sierra Leone at peak transmission. This means 30–40% of contacts were never identified. Among that missed population, some fraction became infected and generated new untracked chains — contributing to the epidemic’s sustained spread.
Stigma-driven concealment
Contacts who fear isolation, income loss, or social stigma may actively conceal their exposure when interviewed by tracing teams. This is particularly challenging in communities where previous outbreak responses were perceived as coercive.
In some DRC communities during the 2018–2020 Kivu outbreak, individuals actively evaded contact tracing teams — attending funerals, crossing borders, and travelling to urban centres while under official monitoring. This behaviour was not irrational from an individual perspective: the perceived cost of 21 days’ isolation was severe, the perceived risk of actual infection was low (given that most contacts don’t become infected), and trust in the response system was limited.
Retrospective contact reconstruction
When an index case is identified late — after the person has been symptomatic for several days — reconstructing the full 21-day exposure history becomes increasingly difficult. Memories are imperfect. People may not know the names or locations of casual contacts. Particularly in mobile, cross-border communities, full contact enumeration is rarely achievable.
Current 2026 Bundibugyo Situation: Reading the Contact Numbers
As of 20 May 2026, 534 contacts are under monitoring across DRC and Uganda for the Bundibugyo outbreak. Understanding what this number means requires the framework above:
- 534 is the exposure count, not the infection count
- Of those 534, statistical expectation suggests 50–100 may become infected (assuming 10–20% attack rate in identified contacts)
- Whether that number is closer to 50 or 100 depends heavily on the risk tier distribution — how many of those 534 were high-risk (direct care, fluid contact) versus low-risk (brief proximity)
- The actual infection count will be lower than 534 — but by how much depends on factors that cannot yet be known
A contact list of 534 that is comprehensive and monitored daily is a strong containment signal. A contact list of 534 in a community of tens of thousands, where cross-border movement is common, may be significantly incomplete.
The coming two to three weeks — two incubation cycles — will be the clearest indicator of whether this outbreak’s exposure-to-infection chain is being successfully interrupted.
Sources: WHO Contact Tracing Operational Manual (2021); Faye et al., New England Journal of Medicine (2015); Bower et al., PLOS Pathogens (2015); DRC MOH Contact Tracing Summary Reports (2018–2020); Africa CDC Bundibugyo Situation Report (18 May 2026).