Article Text

Download PDFPDF

Undifferentiated febrile illnesses in military personnel
  1. Daniel S Burns1,2 and
  2. Mark S Bailey1,2
  1. 1Department of Infection & Tropical Medicine, Birmingham Heartlands Hospital, Birmingham, UK
  2. 2Department of Military Medicine, Royal Centre for Defence Medicine, Birmingham, UK
  1. Correspondence to Lt Col Mark S Bailey, RAMC, Department of Infection & Tropical Medicine, Birmingham Heartlands Hospital, Bordesley Green East, Birmingham, England B9 5ST, UK; markbailey{at}


Undifferentiated febrile illnesses (UFIs) present with acute symptoms, objective fever and no specific organ focus on clinical assessment. The term is mostly used in developing and tropical countries where a wide range of infections may be responsible. Laboratory diagnosis often requires specialist microbiology investigations that are not widely available, and serology tests that only become positive during convalescence. Optimal clinical management requires a good travel history, awareness of local endemic diseases, an understanding of the features that may help distinguish different causes and appropriate use of empirical antibiotics. This review describes the most common examples of UFI in military personnel on overseas deployments, and provides a practical approach to their initial management.

  • Fever of Unknown Origin

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.

Key messages

  • Undifferentiated febrile illnesses (UFIs) are a past, present and future threat for military personnel.

  • Clinical management maybe difficult due to the range of diseases involved and limited diagnostic tests.

  • ‘Helmand Fever’ in Afghanistan is caused by sandfly fever, acute Q fever or rickettsial infections.

  • UFIs in British troops should be managed in conjunction with the UK Role 4 Military Infections Services.


Undifferentiated febrile illness (UFI) is a term used to describe an acute illness (usually of less than 2 weeks) with confirmed fever and no specific organ focus at presentation. It is distinguished from the classic description of pyrexia of unknown origin (PUO) because the latter has more prolonged symptoms (greater than 3 weeks) and, hence, a higher proportion of non-infectious causes.1 The term UFI is mostly used in developing and tropical countries where numerous infections may be responsible. Diagnosis can be difficult even in resource-rich settings due to the wide differential and use of serological (antibody-detection) tests that are often negative at presentation and only performed at national reference centres. It becomes even more difficult in deployed military settings, where laboratory testing may be limited to point-of-care malaria diagnostics and, possibly, blood cultures with basic microbiological identification. The key to initial clinical management (including evacuation when necessary) is a good travel history, awareness of local endemic diseases, an understanding of the features that may help distinguish different causes and appropriate use of empirical antibiotics.

Historically, UFIs have been hugely important for military populations (from epidemic typhus during the Napoleonic wars to trench fever in World War 1) and were major causes of morbidity and mortality.2 Outside of Europe, malaria and yellow fever have been repeated scourges of many military campaigns, including the British 14th Army in Burma, where at one stage 90% of admissions to a field hospital were due to malaria.3 UFI continues to be an important cause of military casualties as shown by the description of ‘Helmand Fever’, which is a generic name for the UFIs occurring among British troops in Helmand Province, Afghanistan,4 and also the most common infectious cause of aeromedical evacuation in British troops.5 Most notably there is now a growing cohort of military patients with Q fever acquired in Afghanistan, who require surveillance for complications such as endocarditis or ‘Q fever fatigue syndrome’.

With a return to possible contingency operations in countries other than Iraq and Afghanistan, a greater awareness of the full range of possible UFIs will once again become essential, and this includes knowledge of both their clinical presentations and local epidemiology. In the most unstable countries, public health measures and surveillance systems have usually broken down leading to both an increased threat from infectious diseases and a lack of data on local endemic diseases.

This review describes the most common causes of UFIs in military personnel on overseas deployments and provides a practical approach to their initial management. Prevention of these diseases is also vital in a military setting to maintain ‘fighting strength’ and is more effective than treatment alone. However, these preventative aspects lie outside the scope of this review and so are mentioned only briefly.



Malaria is caused by protozoa of the genus Plasmodium, and is spread from person to person by female Anopheles mosquitoes, which typically feed at night. Major Ronald Ross of the Indian Medical Services (IMS) was awarded the second ever Nobel Prize in Medicine for discovering its transmission, and hence, ways of preventing it.6 Malaria is widely distributed in the tropics across Asia, Africa and South America, and historically it penetrated as far north as the southern states of the USA, southern Europe and even East Anglia in the UK. Malaria has been a perennial problem for military forces and as recently as 2000 there were 93 cases of malaria in a 2-month period amongst a small number of British units deployed to Sierra Leone.7 At present most malaria cases in British troops occur in Commonwealth recruits after visiting friends and relatives in their countries of origin rather than from overseas deployments (MSB, unpublished data).

Malaria is usually divided into ‘falciparum’ (life-threatening) and ‘benign’ forms (due to P. vivax, P. ovale, P. malariae and P. knowlesi), although the latter may still cause significant illness if untreated. Vivax and ovale malaria have a hypnozoite stage, which lies dormant in the liver and can cause relapses many years after initial infection unless eradication treatment is given.

After an incubation period of at least 7 days (and up to 2 years), symptoms include fevers, rigors, chills, headache and myalgia and gastrointestinal symptoms (especially diarrhoea) are also common. Falciparum malaria can rapidly progress to septic shock, acute renal failure, disseminated intravascular coagulopathy, cerebral oedema, coma and death. Suspicion of malaria should be high in anyone with a fever and travel to an endemic zone in the last year, even if malaria prophylaxis has been taken. Malarial antigen testing kits are widely available in deployed settings and are accurate and easy to use (Figure 1). However, thick and thin blood films may be more sensitive in expert hands and give important information about speciation and the level of parasitaemia.

Figure 1

Point-of-care antigen detection test for malaria.

Treatment should follow Clinical Guidelines for Operations (CGOs),8 which are based on UK treatment guidelines,9 and involve the early use of antimalarial drugs, adequate fluid resuscitation and early evacuation if features of severe malaria develop. Empirical treatment of unproven malaria is now very rarely indicated due to the availability of point-of-care antigen detection tests. Military-specific advice on malaria and other protozoal infections is available from the UK Role 4 facility for infectious diseases and tropical medicine at Birmingham Heartlands Hospital (BHH), as shown in the British National Formulary.10


Visceral leishmaniasis (previously known as kala-azar or Dum-dum fever) is caused by Leishmania donvani protozoa, and is named after its co-discoverers Major General William Leishman RAMC, and Captain Charles Donovan IMS.11 It was a significant problem in the garrison towns of India in the 19th century, and is now most commonly found from the Horn of Africa through to Bangladesh. The parasite is spread by the bite of female Phlebotamine sandflies and develops into intracellular parasites (amastigotes) within cells of the mononuclear phagocyte system. After an incubation period of at least 14 days (and up to 2 years), it typically presents with a UFI, but may have a more insidious onset in indigenous patients. In due course, more characteristic features develop, such as (massive) splenomegaly, hepatomegaly, pancytopenia and secondary infections (Figure 2). Clinical diagnosis is difficult in the early stages, but a rapid point-of-care test is now available.12 Empirical treatment is not indicated in military personnel due to the cost and toxicity of the drugs involved, and so, specialist advice and medical evacuation to Role 4 are required. The cutaneous form of leishmaniasis does not cause UFI, but is common in military populations.13 Therefore, personnel who have visited endemic areas and have non-healing skin lesions that do not respond to appropriate antibacterial drugs should be referred to Role 4.14

Figure 2

Visceral leishmaniasis in a British soldier from Iraq.


Enteric fever

The enteric fevers caused by Salmonella typhi and paratyphi have often been associated with the military, most notably during the Boer War, where deaths due to typhoid were greater than from enemy action (despite the availability of an effective vaccine developed by the RAMC).15 These infections are endemic across the tropics, especially in Africa and the Indian sub-continent and are spread via the faecal-oral route including contaminated food and water. Hence, outbreaks are common where public health measures have broken down or do not exist, such as in refugee camps. After a typical incubation period of 7–14 days (and up to 60 days) the patient presents with a gradual onset of high fevers, headache and malaise. Although the main bacteriological focus is the small bowel, abdominal symptoms are not always present, and systemic infection usually occurs leading to a wide range of other symptoms. Relative bradycardia is a non-specific feature, and more characteristic features, such as ‘rose spots’ and small intestinal perforation are uncommon.16 Complications typically occur at the start of the third week and intestinal perforation and bleeding, septic shock, pneumonia and nephritis can all be life threatening. In a deployed setting diagnosis will be by clinical suspicion or culture of blood, faeces, urine, or aspiration of the rose spots. Serology tests are rarely of use in the acute situation because they are usually negative at presentation, and following seroconversion they may remain positive for the long term. Antibiotic resistance is variable and widespread, and so, quinolones are usually avoided with intravenous ceftriaxone being used for severe cases and oral azithromycin for uncomplicated cases.17 Although vaccination is widely used in military population, this is not 100% effective, does not give protection against paratyphi species, and protection decreases after just a few years.18


Brucellosis has been called ‘The Corps Disease’ due to its long association with the RAMC and is named after Major General David Bruce, who showed it to be the cause of ‘Malta fever’ in 1887.19 It is a zoonotic disease of ruminants and is usually transmitted via the milk (or meat) of infected animals. It remains endemic across the Arabian Peninsula, and personnel in contact with animals in a rural setting, especially when those animals have been symptomatic, are at high risk (especially Royal Army Veterinary Corps personnel). Following an incubation period of 7–56 days, it typically causes an ‘undulant’ fever that fluctuates over a period of months, and may also have more focal features such as large joint arthritis, spondylitis (Figure 3) and hepatitis. The diagnosis is usually made by serology, but the Brucella bacteria may also grow from conventional blood cultures. Where clinical suspicion exists, laboratory personnel should be alerted because laboratory-associated infections can occur.20 Treatment requires a combination of antibiotics (eg, doxycycline and rifampicin) for at least 6 weeks and, hence, requires medical evacuation to Role 4.

Figure 3

Right sacro-iliac joint infection due to brucellosis.

Q fever

Q fever is a zoonotic infection that has affected recent British military deployments in the Balkans, Iraq, Afghanistan and Kenya. It has numerous animal hosts including goats, cattle, sheep, cats and dogs (but not pigs) and is transmitted from animal body fluids (especially at parturition) and can survive in spore-like form in the environment leading to windborne spread and inhalation. After an incubation period of 9–40 days, patients present with varying degrees of UFI, pneumonitis and hepatitis. The diagnosis can be made by serology or polymerase chain reaction (PCR) tests, but early empirical treatment is recommended to limit subsequent morbidity.21 Although acute Q fever is usually a self-limiting illness, complications include acute myocarditis, chronic Q fever infection (usually of heart valves or prosthetic vascular grafts), and ‘Q fever fatigue syndrome’ that affects approximately one-quarter of acute cases for a year or more. Doxycycline for 2–3 weeks is the preferred antibiotic for acute Q fever, but chronic Q fever requires a combination of antibiotics (eg, doxycycline and hydroxychloroquine for at least 18 months). No licensed vaccine is available in the UK, but tetracycline prophylaxis is protective if given for long enough after exposure,22 ,23 and so doxycycline prophylaxis (as given for malaria) should also prevent Q fever. A seroepidemiological study has shown that 1% of British troops deployed in Helmand, Afghanistan seroconvert to Q fever every 6 months equalling about 200 cases per year, of whom about half would be symptomatic.24 A cohort of more than 80 cases of acute Q fever from Helmand is now being followed up in the UK Role 4 facility at BHH (MSB, unpublished data).



Leptospirosis is a globally widespread zoonotic disease due to Leptospira spirochaetes, which was a significant problem during British military campaigns in Malaya and Borneo.25 The organism is excreted in the urine of various infected mammals (not only rats) and is usually transmitted to humans from contact with infected water through mucus membranes, skin abrasions or lacerations. Occupational exposure is common in paddy farmers, loggers and plantation workers in Southeast Asia, and infection from water-related leisure activities is well described.26 After an incubation period of 3–21 days, the spectrum of disease varies from asymptomatic infection, to a self-limiting UFI and, ultimately, severe sepsis with multiorgan failure may occur. Conjunctival suffusion is considered characteristic (Figure 4), but has relatively low sensitivity and specificity, and leptospirosis should be suspected in anybody with a UFI and exposure to fresh water, especially patients with renal or hepatic features. Microscopy and culture of the organism from blood and urine is possible, but require specialist facilities, and so the diagnosis is usually made from serology or PCR tests. Supportive therapies and early antibiotic treatment are important and fortunately the organism is sensitive to a wide range of penicillins, tetracyclines, fluroquinolones and cephalosporins, which means that many cases are treated unknowingly. Doxycycline prophylaxis seems to be effective27 and military doctors should be aware of the possibility of leptospirosis both on deployments worldwide and also when freshwater exposure occurs during training. The last confirmed military case seen at the UK Role 4 facility originated from an exercise in North Wales (MSB, unpublished data).

Figure 4

Conjunctival suffusion and icterus due to leptospirosis.

Relapsing fevers

The relapsing fevers are caused by Borrelia spirochaetes, which are transmitted by ticks and lice. Louse-borne relapsing fever causes epidemics during war and natural disasters, and over 50 million cases occurred in the early 20th century with a case fatality rate of 10%.28 It is now mostly confined to the Horn of Africa. Tick-borne relapsing fever is widely distributed with foci in Rwanda, Burundi and Tanzania, but is found almost everywhere outside of Australasia.29 After a 4–18-day incubation period, these infections present with high fever, rigors and myalgia, and jaundice is also common. Untreated infections will often resolve for 5–9 days followed by a series of relapsing and remitting episodes. The diagnosis is made by seeing the spirochaetes on blood films (often stained to look for malaria), and doxycycline is most commonly used for treatment (which may produce a Jarisch-Herxheimer reaction).


Rickettsia are obligate intracellular bacteria, and there are a large numbers of species responsible for a variety of diseases including epidemic (louse-borne) typhus, endemic (murine or flea-borne) typhus, spotted fevers (such as African tick typhus) and scrub (mite-borne) typhus and trench fever. These vary in their geographical distribution, their reservoir hosts and arthropod vectors. Historically, louse-borne typhus was the most important with large epidemics occurring during both World Wars30 and the RAMC was responsible for controlling large outbreaks at Naples and during the liberation of the Bergen-Belsen concentration camp.31 They may present with non-specific features such as fever, severe headache, mylagia, arthralgia, a vasculitic rash, thrombocytopenia, liver transaminitis, encephalitis and possible multiorgan failure. An eschar may be present at the site of a tick bite in certain spotted fevers and scrub typhus (Figure 5), but is much less common in other rickettsial infections. The diagnosis is usually made with serology tests, but these are notoriously difficult to perform and interpret, and so, early empirical treatment with doxycycline is often required due to the risk of mortality. Ectoparasites can still be a significant problem on deployments and must be controlled to prevent rickettsial infections from occurring.32

Figure 5

Eschar in the right axilla due to scrub typhus.

Figure 6

Maculopapular rash (with ‘islands of sparing’) due to dengue.


Arthropod-borne viruses (arboviruses) are ecologically grouped together due to their transmission by arthropod vectors.33 Yellow fever was historically the cause of much morbidity in the West Africa Squadron,34 dengue (Figure 6) was a problem during military deployments in Haiti35 and East Timor,36 and sandfly fever is the leading cause of UFI in British troops in Helmand, Afghanistan.4 ,24 Their epidemiology varies enormously and they can be widespread (eg, dengue), localised (eg, o'nyong-nyong) or emerging (eg, chikungunya, West Nile fever). Transmission may be anthroponotic (eg, dengue) or zoonotic with the involvement of an amplifying host (eg, Japanese encephalitis is spread from a bird reservoir to a pig amplifying host and then onto humans.) The commonest vectors are mosquitoes (including Aedes, Culex and Anopheles species), and it is the behaviour of these that determine the best control measures to use.

Arbovirus infections tend to be characterised as fever-arthralgia-rash (eg, dengue, chikungunya, sandfly fever) (Figure 6), encephalitis (eg, Japanese encephalitis, West Nile fever) or haemorrhagic syndromes (eg, yellow fever, dengue haemorrhagic fever). However, these may overlap (eg, dengue, Rift Valley fever) and most will start as a UFI. Diagnoses are usually made by serology or PCR tests, but point-of-care antigen detection tests are now becoming available for dengue,37 which is the most important arbovirus infection worldwide. Treatment is generally confined to supportive measures, but these can have a major impact on mortality rates. Prevention by vaccination is available for yellow fever and Japanese encephalitis, but bite avoidance and vector control is all that is available for the others.

Other haemorrhagic viruses

Viral haemorrhagic fevers (VHFs) include Lassa fever in West Africa, Ebola in Central Africa and Crimean-Congo haemorrhagic fever (CCHF) throughout Central Asia and Africa. These all present with a UFI in early or non-severe disease and only later develop the characteristic clinical features of bruising, bleeding and coagulopathy. A UFI in a patient from an endemic or outbreak area should alert clinicians to the possibility of VHF if symptoms develop within the incubation period of up to 21 days. Updated guidelines on the management of VHFs are now available and include a useful algorithm for the risk assessment of possible VHF cases.38 Strict isolation, supportive treatment and immaculate universal precautions and infection control practices are essential until a case can be confirmed or refuted. Diagnosis is made by PCR and serology tests, and early treatment with ribavirin seems to be of benefit in Lassa fever and CCHF. Protection of medical staff is of paramount importance to prevent degradation of the medical response to an outbreak.

Other viruses

Hantaviruses are transmitted by rodent faeces, urine or saliva and cause either ‘haemorrhagic fever with renal syndrome’ in Europe, Africa and Asia or ‘hantavirus pulmonary syndrome’ in North, Central and South America. Although the names of these syndromes suggest quite specific clinical features, many cases will initially present with a UFI. Hantavirus, probably the cause of trench nephritis in World War 1,39 was first identified following outbreaks throughout the Korean War40 and occurred in British troops during the Balkans conflict.41 Even when characteristic features are present, it is difficult to distinguish hantavirus infection from leptospirosis. Diagnosis usually relies on serology tests because the short viraemic stage of the illness means that PCR tests are less useful. No specific antiviral treatment is available, but supportive measures including renal replacement therapy and ventilation may be life saving.

Acute Epstein-Barr virus (EBV) causing infectious mononucleosis or ‘glandular fever’ is a common cause of UFI in young adults and is usually accompanied by pharyngitis, lymphadenopathy, atypical lymphocytes, raised liver function tests and possibly splenomegaly. Acute cytomegalovirus (CMV) and HIV can also present in a similar way and clinicians should have a low threshold for HIV testing in all UFIs.42

Other infections

Other infections may also be labelled as UFIs if they have occult presentations or if characteristic clinical features are not identified. For example, endocarditis, liver abscesses,43 biliary infections and urogenital infections may lack their usual localising clinical features. Also the characteristic features of acute schistosomiasis (Katayama fever), eosinophilic pneumonia due to hookworm infection or strongyloidiasis (Löffler's syndrome), disseminated tuberculosis, atypical respiratory infections and influenza may all be missed. Obviously, if a military patient is immunocompromised, then the range of infections presenting with a UFI will be different from those described here.

Non-infectious causes

Several non-infectious diseases can cause fever, although the proportion of these is much greater when the classic PUO definition is used.1 Connective tissue diseases, acute sarcoidosis,43 malignancies (especially haematological), drug fevers and thromboembolic disease are all well described causes of fever. However, most of these will be less common in military personnel than the general population. Heat illness also presents with fever, and clinicians should be alert to this possibility when troops present with fever after a period of intense physical exercise, even in relatively temperate climates. It is also important that personnel are not labelled as having heat illness when in fact they have a UFI due to an infection that cannot be easily diagnosed with the facilities available on deployments. Where doubt exists, cases should be referred to the UK Role 4 facility at BHH for clinical assessment and appropriate serology tests that may remain positive for many months after infection.

‘Helmand fever’

From 2006, British troops in Helmand, Afghanistan suffered from UFIs that defied diagnosis and often required medical evacuation until the ‘Helmand Fever Study’ was performed in 2008.4 Since three different infections were found to be responsible (sandfly fever, Q fever and rickettsial infections) the generic term ‘Helmand Fever’ has persisted for cases that are not yet diagnosed, and this is now defined as ‘an acute undifferentiated febrile illness occurring amongst British troops in Helmand, Afghanistan that has no organ focus on clinical and radiological assessment and no positive results from microbiological investigations including blood cultures and malaria antigen tests.’ Clinical guidelines for the initial management of such cases have been produced,44 and these include the use of empirical doxycycline treatment for 2 weeks to minimise the morbidity associated with acute Q fever and rickettsial infections.21 Follow-up of Q fever patients at the UK Role 4 facility is also important to ensure that chronic Q fever and ‘Q fever fatigue syndrome’ are diagnosed and treated as early as possible. Meanwhile seroepidemiology data suggests that more than 600 British troops seroconvert to these infections every year in Helmand Province, although not all of these will be symptomatic.24

Clinical approach to military patients with a UFI

The approach to a miliary patient with a UFI will depend on the healthcare setting and severity of the illness. A clinician seeing a generally well patient in the UK will have time for a full assessment including a detailed history and a range of sophisticated tests. In a deployed setting with a sick patient, an ‘ABC’ and ‘surviving sepsis’45 approach with simultaneous assessment and treatment including early empirical antibiotics will be needed.

A detailed history must be taken, looking specifically for an organ focus and characteristic combinations of symptoms. The travel history should include details of overseas travel for at least the previous 2 years including exposures, such as unsafe water and food ingestion, arthropod bites, animal contacts, fresh water immersion and a sexual history. The timing of such exposures and consideration of the incubation periods for various infections may help to narrow the differential diagnosis considerably.

A thorough clinical examination is essential, and simple investigations, such as blood tests and urinalysis may also point to an organ focus or specific diagnosis. For some diseases in certain settings, there may be clinical prediction rules that can be used to help identify or exclude certain infections.46 However, these require detailed research to develop and should not be used in settings where they have not been evaluated.

Investigations should include malaria films and antigen tests if there has been any travel to a malarious area in the last 2 years. Microbiology cultures of blood, urine and faeces should be done routinely, and for other body fluids (eg, cerebrospinal fluid or sputum) as appropriate. Imaging studies should include a chest radiograph and also abdominal imaging by ultrasound or computerised tomography scanning if the diagnosis remains uncertain. Whenever possible, blood should be taken for PCR tests (which are increasingly available for the infections causing UFIs) and for serology tests during the acute phase of the illness, and also 2–4 weeks later during convalescence. An HIV test should be performed if the patient meets the UK testing criteria,42 or if the diagnosis remains uncertain.

Treatment may need to follow an empirical approach in deployed settings, and the choice of antimicrobial agent will depend on the clinical and laboratory features as well as any background information that is available for a given location. For example, doxycycline is recommended for British troops in Afghanistan if malaria and bacteraemia are excluded in view of the known local causes of UFI.44

All British military patients with UFIs should be referred to the UK Role 4 facility for infectious diseases and tropical medicine at BHH to ensure that a precise diagnosis is made whenever possible. This will ensure that optimum treatment is given and any long-term sequelae are dealt with as effectively and efficiently as possible. The information obtained from confirmed cases is also notified to the military public health authorities and used for epidemiological and preventative purposes. Finally, experience suggests that military personnel greatly appreciate having a full explanation of their UFI provided by military physicians with experience in this topic.


Undifferentiated febrile illnesses can cause significant morbidity and mortality, and have historically had a large impact on the fighting strength of virtually every army, often much more decisively than enemy action. With new infectious diseases emerging worldwide, and distribution of existing diseases changing with urbanisation and climate change, they will continue to be a challenge to any future fighting force, especially when deployments take troops to remote areas of the tropics. Diagnosis in deployed settings can often be difficult, but a robust assessment approach, knowledge of the locally endemic diseases and simple laboratory tests will allow many UFIs to be safely managed. However, reach-back support is now also widely available due to modern battlefield telecommunications. Advice on clinical management is available from the UK Role 4 facility at Birmingham Heartlands Hospital which has unrivalled experience with these infections. Medical officers who are likely to manage soldiers with UFIs are encouraged to attend the recently reformed Military Infectious Diseases & Tropical Medicine Course.



  • Contributors MSB provided some background materials for DSB to write the initial draft and then MSB produced the final article.

  • Competing interests None.

  • Provenance and peer review Not commissioned; internally peer reviewed.