Ebola virus infection is a notifiable disease. The case definition for Ebola virus infection is very broad and includes a long list of possible differential diagnoses.
The initial assessment of a patient with suspected Ebola virus infection hinges on two main factors:
Epidemiological risk (e.g., living or working in, or travel to, endemic area in previous 21 days); and
Presence or history of a fever in the past 24 hours.
Isolation and personal protective equipment (PPE)
Infection control risk should be assessed. Having determined that a patient may be infected, the physician needs to determine how infectious the patient is currently. For example, the absence of vomiting/diarrhoea reduces the risk; however, uncontrolled diarrhoea greatly increases the risk of transmission.Identifying that a symptomatic patient may be at risk of infection mandates precautionary isolation procedures and use of PPE until the infection is either confirmed or excluded. It is extremely important to minimise the risk of transmission while working up the patient.
The World Health Organization (WHO) and Centers for Disease Control and Prevention (CDC) produce detailed guidance on PPE:
CDC: guidance on personal protective equipment (PPE) to be used by healthcare workers during management of patients with confirmed Ebola or persons under investigation (PUIs) for Ebola Opens in new window
The CDC and WHO also produce detailed guidance on infection prevention and control for healthcare workers:
A detailed history helps to clarify the level of risk for Ebola virus infection, as well as assess the possibility of other causes of an acute febrile syndrome.
People living or working in endemic areas (e.g., West Africa, Democratic Republic of the Congo) are at high risk of infection. However, recent arrival from endemic areas is also an important risk factor.
Most patients with suspected infection in developed countries will be returning travellers and healthcare workers who have cared for patients during outbreaks. Therefore, a comprehensive travel history is extremely important. History of recent arrival from an endemic area is significant. Up-to-date knowledge of the geographical locations of active epidemics helps to clarify the patient’s epidemiological risk.
Apart from healthcare workers, other high-risk occupations include those where people work with primates or bats from endemic areas, or high-risk clinical samples.
As malaria is still the most common cause of febrile illness in returning travellers from West Africa, the presence of risk factors for acquiring malaria should be assessed (e.g., living/working in, or travelling to, endemic area; inadequate or absent chemoprophylaxis; not using insecticides or bed nets). However, co-infection with malaria was seen in up to 5% of patients in West Africa during the 2014 outbreak, so the possibility of dual infection should be considered in all patients.
Contacts of infected patients (including healthcare workers and household contacts) are at risk of infection if the person was exposed to body fluids of the infected patient without appropriate protective equipment. The incubation period after infection is 2-21 days. Incubation periods may be shorter in children. Brief interactions, such as walking by a person or moving through a hospital, do not constitute close contact.
Contact is defined by the World Health Organization (WHO) as someone who has:
Slept in the same household as a patient
Had direct physical contact with the patient during the illness or at the funeral
Touched the patient's body fluids or clothes/bed linens during the illness
Been breastfed by the patient (babies).
Case definitions are updated frequently and differ depending on the organisation. Links to the case definitions by the WHO and CDC are below:
Patients are not considered infectious until they develop symptoms. The initial presentation is non-specific, which makes early clinical diagnosis difficult; however, typical symptoms include:
Unexplained bleeding or bruising.
The most common symptoms reported on admission during the 2014 outbreak were: fever (76%), fatigue (71%), anorexia (64%), headache (56%), diarrhoea (51%), vomiting (50%), myalgia/arthralgia (48%), abdominal pain (40%), sore throat (29%), and conjunctivitis (27%). Other less common symptoms included difficulty swallowing (22%), difficulty breathing (18%), hiccups (13%), haemorrhagic signs (11%), confusion (9%), and rash (3%).
Three phases of illness are typically recognised, starting with a few days of nonspecific fever, headache, and myalgia, and followed by a gastrointestinal phase where diarrhoea, vomiting, abdominal symptoms, and dehydration are prominent. In the second week, the patient may either recover, or deteriorate with a third phase of illness, which includes collapse, neurological manifestations, and bleeding. This phase is often fatal.
Data from the 2014 outbreak indicate that children are relatively spared; however, this may be confounded by a high fatality rate before being registered as a case, or the bias of high rates in healthcare workers. Children present with similar symptoms to adults; however, in previous outbreaks, younger children are reported to have more respiratory (e.g., cough, dyspnoea) and gastrointestinal symptoms, but less bleeding and neurological signs compared with adults. Data were sparse for this patient group in the 2014 outbreak. A paediatric cohort study in Sierra Leone described symptoms in 282 patients and found vomiting (60%), abdominal pain (59%), diarrhoea (45%), and conjunctivitis (38%) were common, while hiccups (5%) and bleeding (2%) were rare. Another study in Sierra Leone found that weakness, fever, and distress were each present in more than 63% of children, and loss of appetite, diarrhea, and cough were present in more than 50%. Approximately 25% of these children did not have fever at the time of admission.
Anecdotally, children aged under 4 years initially present with more subtle symptoms before developing a fever, and are often diagnosed later in the course of illness.
A full physical examination should be undertaken with the aim of excluding a focus for sepsis while looking for signs of viral haemorrhagic fever (e.g., conjunctival injection, purpuric rash, or other signs of bleeding).
Vital signs should be taken:
Fever: the presenting symptom in up to 90% of patients, its presence is enough to raise concern for infection in the appropriate epidemiological context. Although fever is a major presenting symptom, a normal temperature at presentation is common. Wide variations in body temperature can be observed during the course of illness, especially in children,with normothermia or hypothermia occurring in the later stages of fatal infection. Some patients may initially have a low-grade fever with no other symptoms, or alternatively the temperature may be near normal at first evaluation. The temperature threshold for fever differs among countries and guidelines, and using a lower temperature threshold (e.g., ≥37.5°C) increases the sensitivity of finding cases.The World Health Organization use a threshold of >38°C.However, in a large cohort in Sierra Leone, <30% had a fever of ≥38°C at presentation, although a history of fever was reported by 89% of patients.
Blood pressure: hypotension is a feature of pre-terminal disease and shock. It is under-documented in field studies, owing to a lack of measuring equipment in endemic areas. However, septic shock with vascular leakage and microcirculatory failure does not appear to be a dominant feature.
Pulse rate: bradycardia may be present in the initial stages of illness; however, tachycardia may be seen in the later stages of fatal infections.
Respiratory rate: tachypnoea, along with tachycardia, correlates with a more severe or advanced infection, and is more likely to be respiratory compensation of a metabolic acidosis rather than respiratory involvement. However, respiratory involvement has been described.
Other findings may include:
Maculopapular rash: develops early in the course of illness. It is frequently described as non-pruritic, erythematous, and maculopapular. It may begin focally, then become diffuse, generalised, and confluent. Some have described it as morbilliform. It may become purpuric or petechial later on in the infection in patients with coagulopathy. May be difficult to discern in dark-skinned patients.
Bleeding: bleeding manifestations (e.g., epistaxis, bleeding gums, haemoptysis, easy bruising, conjunctival bleeding, haematuria, oozing from injection or venipuncture sites) were present in 30% to 36% of infected patients in previous outbreaks; however, they were reported in fewer patients in more recent outbreaks. It is less common in children.
Hepatomegaly: tender hepatomegaly with the edge of the liver palpable below the rib cage has been reported, but is uncommon.
Lymphadenopathy: enlarged lymph nodes have been reported, but are uncommon.
Neurological signs: depressed consciousness, encephalopathy, and seizures are rare but their presence indicates advanced infection. Confusion may be multifactorial in children and is associated with a poor prognosis.
All specimens should be collected according to strict protocols. The WHO and CDC have published guidance on this:
The main confirmatory test for Ebola virus infection is a positive reverse transcriptase-polymerase chain reaction (RT-PCR) for Ebola virus. This test should be ordered in all patients with suspected Ebola infection while the patient is in isolation. It has the advantage of returning a result 24 to 48 hours before ELISA testing. Several different commercial PCR kits are available with varying sensitivity, specificity, and limits of detection. In Western settings, the test may only be available in regional or national laboratories that have category 4 facilities. In epidemic settings and some countries, category 4 laboratories are set up locally and results are available 4 hours after the sample has arrived. Viral RNA can be detected in the patient’s blood by RT-PCR from day 3 up to days 6 to 17 of symptom onset. A positive PCR result implies that the patient is potentially infective, particularly if he or she has active diarrhoea, vomiting, or bleeding. If negative, the test should be repeated within 48 hours since viral load is low and can be undetectable early in the course of the illness. Negative tests should be repeated to rule out a diagnosis if it is strongly suspected (or confirm resolution of infection). Higher viral load correlates with adverse outcome and increased mortality.
The choice of whether to test for Ebola virus infection depends on the patient's history and their risk of infection according to the algorithm below.
Malaria is still the most common cause of fever in people who live/work in, or travellers who have returned from, an endemic area and should be ruled out. Co-infection with malaria was seen in up to 5% of patients in West Africa during the 2014 outbreak, so the possibility of dual infection should be considered in all patients. In the case of a positive rapid diagnostic test result for malaria, the infection should be treated while keeping in mind the patient's risk for Ebola virus infection and the possibility of a dual infection. Ebola virus infection should be considered in a patient who does not respond to antimalarial therapy.
It is recommended that appropriate confirmatory tests for Ebola virus infection are performed before, or in tandem with, differentiating tests for other suspected conditions if Ebola virus infection is suspected.
Traditionally, no other investigations outside of a malaria screen and reverse transcription-polymerase chain reaction (RT-PCR) were recommended due to the fear of putting laboratory workers at risk. However, it is now recognised that other investigations can be done safely according to recommended guidelines, as long as the laboratory is informed of the sample in advance, and the bloods are correctly packaged and retained at the end in case the RT-PCR is positive. Local protocols should be clear about safe transport of samples to the local and referral laboratories, and safe handling on receipt in the local laboratory.
The following investigations add valuable information to the work-up and help guide further management, and should be ordered if possible. If investigations are limited due to the geographical location or facilities available, the most important tests to order are renal function, serum electrolytes, and blood lactate (if available).
Renal function and serum electrolytes:
Elevated serum creatinine or urea and abnormal electrolytes may indicate acute kidney injury. This may be seen at the end of the first week of infection. Hypokalaemia or hyperkalaemia, due to vomiting and diarrhoea or acute kidney injury, was seen in approximately 33% of cases in the 2014 outbreak. Hypocalcaemia has been associated with fatal infection. Haematuria and proteinuria may also be seen in severe disease. Oliguria that does not respond to fluid resuscitation is a poor prognostic sign.
Elevated lactate is a marker of tissue hypoperfusion and is an indicator of shock. It is useful in acutely ill patients with signs of sepsis to identify the degree of systemic hypoperfusion and to guide fluid resuscitation. Elevated lactate was one indicator of gram-negative sepsis at day 15 in a patient treated in Germany.
Arterial blood gas:
Arterial or venous blood pH and bicarbonate are useful in acutely ill patients with signs of sepsis to identify the degree of systemic hypoperfusion and guide fluid resuscitation.
Full blood count:
Decreasing platelet count and marked lymphopenia can be seen in the initial stages of infection; however, this is not diagnostic. This is often followed by neutrophil leukocytosis in the later stages of patients who eventually recover, along with normalisation of thrombocytopenia. Leukocytosis may persist and show immature forms. Patients with severe disease may show a progressive decline in platelet count as a manifestation of disseminated intravascular coagulation (DIC). Decreased haemoglobin levels were reported in 24% of patients in the 2014 outbreak, and have been associated with bleeding in previous outbreaks.
Prolonged prothrombin time (PT) or activated partial thromboplastin time (aPTT) is associated with more severe infection and bleeding manifestations such as DIC. Also, patients with fatal infections have been found to have D-dimer levels four-fold higher on days 6 to 8 of infection compared with patients who survive.
Liver function tests:
Both alanine aminotransferase (ALT) and aspartate aminotransferase (AST) are usually elevated; however, most studies show that AST rises out of proportion to ALT, and this is more suggestive of systemic tissue damage rather than hepatocellular injury. The AST:ALT ratio peaked at 15:1 on days 6-8 of infection in fatal cases when compared with non-fatal cases, which had a peak of 5:1. Bilirubin, gamma glutamyltransferase, and alkaline phosphatase are often slightly elevated. Highly elevated ALT with severe jaundice suggests an alternative diagnosis (e.g., viral hepatitis).
Elevated levels have been reported in several studies and indicate the presence of pancreatitis, an indicator of severe infection.
Serum blood glucose:
Negative blood cultures are helpful as they rule out other non-viral infectious causes (e.g., sepsis, enteric fever). Gram-negative bacteraemia, presumably from gut translocation, has been identified as a complication of the disease course in two patients. However, a study in Sierra Leone where blood cultures were taken from patients on admission to an Ebola treatment centre found that only one of the 22 cultures was positive with a presumed contaminant. Therefore, blood should be collected for culture at baseline and/or at the time of the onset of gastrointestinal symptoms or other clinical deterioration.
Antigen-capture enzyme-linked immunosorbent assay (ELISA) testing:
A useful diagnostic test with high specificity; however, it is not universally available. It is most likely to give a positive result from day 3-6 of infection, and can give widely variable results from days 7-16. Can be used to confirm the diagnosis along with a positive RT-PCR result.
IgM and IgG antibodies:
Useful in later stages of infection. IgM antibodies can appear in serum as early as day 2 post infection, but can give variable results up to day 9. They become negative between 30-168 days after symptom onset. An IgG response develops between day 6-18 and can persist for several years. A positive IgM or a rising IgG titre is strong evidence for recent Ebola virus infection.
Useful in patients with respiratory symptoms. Pulmonary infiltrates are not typical of infection and suggest an alternative (or comorbid) diagnosis. May be difficult to arrange in an isolation unit and should only be ordered judiciously to avoid contamination.
Rapid diagnostic tests
Rapid PCR testing for Ebola virus infection remains a major hurdle for effective, targeted isolation of affected patients. Current tests take an average of 4 hours to perform with a fully equipped level 3 or 4 biosafety laboratory close at hand, but results may take several days to arrive in remote areas. This means that, until they are confirmed negative, patients with febrile illnesses other than Ebola virus infection are confined to isolation and often unwittingly exposed to the virus. Rapid bedside tests can therefore make a very significant contribution to infection control in treatment centres.
Several different technologies are being evaluated by WHO for use in field conditions. These include numerous RT-PCR-based assays that have been made simpler to use with a shorter turnaround time of <1 hour. The WHO has listed ReEBOV™ Antigen Rapid Test Kit for potential use; however, it currently only recommends its use in special situations. The alternatives are ELISA-based antigen-detection assays that could be quicker and simpler with the possible advantage of only needing a drop of blood. Their major disadvantage is a reduced sensitivity, particularly in the initial stages of illness.
Nanopore technology may allow rapid detection and sequencing in the presence of very low levels of virus, and can potentially be deployed using a pocket-sized detection kit. Rapid sequencing of Ebola virus using these new technologies during an outbreak could allow real-time understanding of viral dynamics.
A GeneXpert® diagnostic tool has been developed and trialled in the field. The Xpert® Ebola is an automated cartridge-based system that requires minimal laboratory skill. An inactivated sample is placed into a single-use cartridge, which is then entered into the enclosed machine. Sample preparation, nucleic acid amplification and detection, and production of a result are automated processes minimising staff training requirements, risk of infection, and cross contamination.
Other test kits have also been granted emergency use authorisation by the WHO.
This is an evolving field and different kits are approved according to the country and settings in which they are to be deployed.
Use of this content is subject to our disclaimer