There are over 300 new cases of imported paediatric malaria in the UK each year and this has been increasing over the last 20 years. Malaria in children is particularly difficult to diagnose because the initial presenting features are subtler than in adults and do not display the classical presenting features. However, they are also more likely to deteriorate rapidly and to develop severe malaria. The ‘gold standard’ for ruling out the diagnosis of malaria if clinically suspected is three negative thin and thick blood films, which require serial phlebotomy and the availability of trained technicians. There are now a range of other tests, including rapid diagnostic tests and PCR, as well as clinical features that make the diagnosis more or less likely. We explore the different tests available and whether these might replace the three negative blood films currently needed. We also look at whether we are able to use clinical features to aid the tests used for a diagnosis of imported malaria.
- General Paediatrics
- Infectious Diseases
- Tropical Inf Dis
Statistics from Altmetric.com
There are over 300 new cases of imported paediatric malaria in the UK each year and cases of imported malaria here have been increasing over the last 20 years.1 Malaria in children is particularly difficult to diagnose because the initial presenting features are subtler than in adults. Children may appear quite well initially with a fever and no focus, but they are at risk of a rapid deterioration and are more likely to develop severe malaria.
The ‘gold standard’ for ruling out the diagnosis of malaria if clinically suspected is three negative thin and thick blood films.2 This approach, however, relies on serial phlebotomy and the availability of adequately trained staff. Furthermore, during out-of-hours periods, the time and resources required are likely to result in delays in obtaining results, especially if trained staff have to come in from home. There are now a range of tests available, including rapid diagnostic tests (RDTs), as well as highly specific and sensitive tests such as PCR. In addition, there are features from the examination and routine blood work that can be used to better determine a child's pretest probability of having malaria.
Around 70–80% of malaria cases imported into the UK each year are falciparum malaria and most are contracted from visiting sub-Saharan Africa (>90%).1 In the UK, falciparum malaria is responsible for the most serious cases. Table 1 outlines the epidemiology of UK malaria by species.
Among UK children, three out of four children infected with malaria are non-immune because they have not been previously exposed to Plasmodium falciparum.1 ,4 They then travel to a malaria endemic region to visit friends and relatives and contract malaria that they import to the UK. These non-immune children have a different pattern of disease to immune children and are more likely to have rapidly progressing, severe malaria.5 There are several different criteria that justify a diagnosis of severe malaria, as detailed in table 1. A diagnosis of severe malaria requires the presence of only one feature listed in column 1 of table 2.
The cases of imported malaria are highest between the months of July and October, although there is another smaller peak seen around the winter holiday period.1 Non-immune children from the UK are typically non-compliant with antimalarial prophylaxis with two-thirds taking some form of prophylaxis but only one-third or less being compliant with treatment.4 ,7
Most children present within 2 weeks of returning to the UK.4 The incubation periods for malaria are 13 days for P. falciparum and Plasmodium vivax malaria, 14 days for Plasmodium ovale malaria and 34 days for Plasmodium malariae malaria in non-immune people not on antimalarial chemoprophylaxis.8 Those who present later are less likely to have falciparum malaria or will have immunity or had partial or partially effective chemoprophylaxis.4 P. falciparum malaria can cause recrudescent infections (=re-emergence of parasitaemia after full suppression to microscopically undetectable levels) 2–10 weeks following antimalarial treatment or chemoprophylaxis has been discontinued. In 4.6% of 1290 patients who developed P. falciparum malaria after return to the USA, the onset of clinical malaria was delayed by more than 2 months after return and occurred then despite effective prophylaxis documented in 33.9% of those cases. In P. vivax, P. ovale and P. malariae infections, late onset more than 2 months after return was observed in this study in 62%, 75% and 41%, respectively, despite effective prophylaxis in the majority (61%–81%), which was due to relapse (re-emergence of parasitaemia from liver stages).9 It has been demonstrated that non-immune children not on chemoprophylaxis develop symptoms earliest and can present by 7–9 days after returning from travelling.1 ,10
The most common presenting feature of imported malaria is fever, which may also be the only symptom. Incidence varies in studies from 87% to 98%, but it has been found to be more common in non-immune children.1 ,7
Thrombocytopaenia defined as having a platelet count of <150×109/L is also a common feature. In the five case series we looked at, thrombocytopaenia was seen in 59% of patients (table 3). Thrombocytopenia <150×109/L is more common in non-immune than immune children (71% vs 37%)7 and is associated with more severe disease.1
Anaemia (haemoglobin <10 g/dL) is also a common presenting feature seen in 72% of children. There is no difference in the levels of anaemia seen between immune and non-immune children.7
In the five case series that we looked at, 59% of patients had hepatomegaly, splenomegaly or both (table 3), but again no significant difference was demonstrated between immune and non-immune children.7
There are two widely available tests in UK hospitals to diagnose malaria. These can be divided into light microscopy (the commonly used ‘thick and thin films’) and RDTs. PCR is also used and starting to play more of a role.
This is considered the gold standard test in many areas. Sensitivity is thought to be no higher than 75% for a single thick blood film and is lower in those with immunity, non-falciparum malaria, partially treated malaria or low-level parasitaemia.12 Experienced technicians may be able to detect down to 5 parasites/µL; however, on average, only 50–100 parasites/µL are detected.13 In the case series papers we looked at, the lowest parasite counts were around 0.1%, which equates to 5000 parasites/µL, a count more than high enough to be detected by light microscopy.1 ,10 ,11 Parasite counts are higher in non-immune individuals, indicating that light microscopy may be more sensitive in non-immune individuals.7
The advantages of light microscopy include species identification, determining the peripheral parasite density and monitoring this with treatment. It is relatively inexpensive and requires little infrastructure.12 A blood film can be read in about 20 min by a skilled technician and costs about £0.13. However, it is less accurate at extremes of parasitaemia and in non-endemic areas the technicians reading the slides may not retain the skills to do so accurately.12
For light microscopy, it is recommended that blood is collected in an EDTA tube, ideally at the time of the patient having a fever or developing a fever (particularly in infections with species other than P. falciparum). However, parasites are found during all stages of the infection, particularly in P. falciparum malaria. Hence, if there is a high index of suspicion, there should be no delay in collecting blood for films.2 Blood sampling is done three times to increase sensitivity of detection.
From this, both thick and thin films will be made from each of the three blood samples. The thick film is unfixed—it is made with a few drops of blood that are allowed to dry before being lysed (usually with water) prior to staining. Thin films are fixed with methanol. Blood should be used as soon as possible to minimise morphological changes in the parasites. Ideally, this should be within 2 h. However, even after prolonged exposure to the anticoagulant, parasites can still be detected so the blood will usually still be examined.2
Thick films are more sensitive for detecting the presence of malaria parasites, but not as useful for detecting the species responsible, which is better seen on the thin films.
Rapid diagnostic tests
RDTss detect parasite-specific antigens or enzymes, which may be genus or species specific (table 4).9 The choice of RDT depends on the malaria species epidemiology in the country of travel. For ease of decision, three geographical zones have been created. RDT’s are now routinely used in many hospitals, but the availability should be confirmed with individual laboratories.
Zone 1: most of sub-Saharan Africa and lowland Papua New Guinea. P. falciparum is predominant and RDTs that detect only P. falciparum are the RDT of choice. Those that detect histidine-rich protein 2 (HRP2) are preferred over those for Plasmodium lactate dehydrogenase (pLDH) as they are more sensitive.14
Zone 2: Asia and the Americas and some areas of Africa (mainly Ethiopian Highlands). P. falciparum and non-falciparum infections both occur as single species infections. Here, combination RDTs that detect all species and can distinguish P. falciparum from non-falciparum are preferred.14
Zone 3: East Asia, Central Asia and South America. Only non-falciparum infections occur. Here, RDTs that detect non-falciparum species are best. Those targeting pLDH specific to non-falciparum species (or common to all species) or aldolase are recommended.14
Most of the assays use a broadly similar principle. The blood sample is placed at one end of an immunochromatographic strip, where it mixes with lysing agents, buffer and labelled antibody. The bound antibody then will bind to indicator antigens which will create a line on the strip, showing a positive result.15 RDTs have an average cost of £0.54 and they cannot currently test for resistance or quantify the extent of the infection.16 As a higher peripheral parasite density has a higher probability of developing into severe disease, particularly in non-immune children, this is an important information that an RDT alone cannot provide. Overall, RDTs are reported to have 80–95% sensitivity and 85% specificity.6
It is worth mentioning that the RDTs to detect P. falciparum HRP2 remain positive 2–3 weeks after disappearance of the parasite, so is unhelpful in someone who has been recently treated for malaria to check for success of treatment or to detect a recrudescence.14 However, WHO evaluated the performance of these assays and found that RDTs perform better at high parasite concentrations and are variable with a low parasitaemia. However, variability among products has also been demonstrated.15 Some recent research has demonstrated a higher than thought variation in the P. falciparum HRP, which is the target of many RDTs. This may explain the variation in test performance. Furthermore, some of the tests become degraded and have a reduced performance at higher temperatures, which may be found in the test settings.16
The most reliable method for detecting parasites is PCR and is considered to have a sensitivity of 97% and specificity of 100%.16 It is especially helpful at low-level infections and has a limit of detection of 0.5–5 parasites/μL. However, this is more expensive, costing up to £2.56 per test. Furthermore, it is prone to contamination and requires delicate preparation and specialist equipment.16 In the UK, this is primarily used in research, but it can be performed for clinical reasons at the Malaria Reference Laboratory and at the Hospital for Tropical Diseases. PCR is not performed routinely and it typically takes at least 10 days for a result.2 ,17
The Malaria Reference Laboratory is also able to test for several molecular markers for drug sensitivity. This is not routine, but may be used in cases of treatment failure. Available tests include antifolate resistance, atovaquone or proguanil resistance and artemisinin-based combination therapy/artemisinin failure to clear.2
Indications and limitations
In children with suspected malaria do we need three negative blood films to exclude a diagnosis of malaria ?
There is only one study exploring the combination of blood films together with RDTs in diagnosing imported malaria and it was in adults. Of the 388 cases, 367 (95%) were diagnosed by the initial blood film. Of the 21 that were not diagnosed on the blood film, 19 had RDTs performed. This diagnosed a further 10 leaving only 9 cases (2.3%) not picked up by a single blood film and RDT. Only one case of P. falciparum infection was missed and this was in a partially immune individual who had already received an unspecified treatment. The remaining eight missed cases were P. vivax and P. ovale.
If we extrapolate from this study, then if a single blood film and RDT are negative, a diagnosis of malaria is extremely unlikely. This is especially true in cases of suspected P. falciparum in a non-immune patient who has not received any treatment.18 The most obvious criticism here is that it is difficult to extrapolate adult data and draw conclusions relating to children. However, the available data comparing parasite counts between children and adults suggest that on average children have a comparable or higher parasite count than adults.19 This would suggest that the results seen for adults would be comparable or even favourable in children.
However, because of the paucity of data overall and lack of paediatric data, it is unknown whether one blood film and one RDT are sufficiently sensitive to completely rule out malaria and three negative blood films continue to be required. This is especially true in cases of non-falciparum malaria or in immune children and partially treated children where the single film and RDT are less reliable. If the single film and RDT are negative, however, an alternate diagnosis should be sought.
In a child with suspected malaria can any of the features of the examination or simple investigations be used to better highlight those children with probable malaria?
In a child presenting with a fever, there are no examination findings or simple investigations that can be used to definitively diagnose or rule out malaria.
Malaria has a variable presentation and shows significant overlap with common paediatric febrile illnesses. No single feature can rule out malaria, but certain features can help to identify those most likely to have malaria or at the highest risk of severe disease. In the case series outlined in table 3, 59% of children with malaria were found to have hepatomegaly/splenomegaly or both.7
In addition, the presence of thrombocytopaenia is associated with more severe disease.1 There are also data exploring the positive predictive value of thrombocytopaenia in the diagnosis of imported paediatric malaria. A UK study looking at 14 822 children presenting to the emergency department over a 12-month period found that of those children with platelets <150×109/L and a history of travel to a malaria endemic region, there was a 91% predictive value of the child having malaria.20 The study excluded children with known malignancy, undergoing chemotherapy, taking anticonvulsants or with an underlying pre-existing haematological disorder.
So, in children presenting with thrombocytopaenia and recent travel to a malaria endemic region, it is important to actively investigate for possible underlying malaria.
Topics for further research
More paediatric studies are needed. The study that looked at ruling out malaria with a single film and a single RDT was done in adults.18 Currently, no similar study looking at this in children exists, which would be useful when thinking about diagnosis in childhood imported malaria. Similarly, a study exploring different combinations of either one, two or three films with RDTs in children is required as currently there is very little evidence to support the use of three blood films as the gold standard.
The other research area would be the development of quick and cost-effective PCR techniques that could potentially be used at the bedside. As blood film and RDT were not as sensitive for non-falciparum species, the use of PCR could be investigated specifically with a certain travel history where non-falciparum species would be more likely. Despite being the most accurate method of diagnosis, it is not widely used in clinical practice yet due to cost and the need for specialist equipment.
Clinical bottom line
The typical UK child seen with imported malaria will be non-immune and non-compliant with prophylaxis, will have visited friends and family in sub-Saharan Africa and returned with imported falciparum malaria.
Non-immune children present earlier (within one to two weeks), have a more severe phenotype and have a higher incidence of fever and thrombocytopaenia.
Malaria should always be considered as a diagnosis in children with unexplained thrombocytopaenia, especially if there is a recent history of travel to a malaria endemic area.
A child with a single negative blood film and RDT is very unlikely to have falciparum malaria and an alternate diagnosis should be sort.
Appendix—search strategy: search strategy and selection criteria
Information was obtained from Medline and PubMed searches for years between 1980 and 2015. Using the search term ‘imported malaria’, original articles in English relating to clinical case series on children with imported malaria were retrieved and their references were searched for relevant clinical studies. Articles that focused only on epidemiology were excluded.
Test your knowledge
Which of the following statements are true?
Plasmodium vivax malaria is the most common type of malaria imported into the UK.
Most children with Plasmodium falciparum malaria were compliant with taking prophylactic antimalarials.
Fever spikes every third day are found in all cases of imported P. falciparum malaria.
A thick film is used in the diagnosis of malaria.
What are the features of severe malaria?
Paroxysms of cough
Regarding diagnostic testing for P. falciparum malaria, the following statements are true:
PCR testing is a rapid diagnostic test.
Rapid diagnostic testing can detect P. falciparum when the microscopy is negative.
Microscopic detection of Plasmodia is operator dependent.
Thin film is more sensitive in detection of Plasmodia than thick film.
Thin film is more specific than thick film in identification of the type of Plasmodia causing malaria.
Which of the following statements are correct?
The majority of cases of imported P. falciparum malaria was acquired in South America.
Children with clinical immunity do not need to take malaria prophylaxis when travelling to malaria endemic countries.
In children with clinical immunity to malaria, the presentation of clinical malaria can be delayed compared with non-immune children.
High-grade parasitaemia is supportive of a diagnosis of falciparum malaria.
Which laboratory features are commonly associated with P. falciparum malaria?
Answers to the quiz are at the end of the references.
Answers to the multiple choice questions
(B); (C); (D)
(B); (C); (E)
(B); (D); (E)
Contributors TW conceived of the article, initiated the idea and initiated the research for it. ED was involved in the main writing of the article and research for the article. ME kindly reviewed the article and refined it and wrote the MCQs.
Competing interests None declared.
Provenance and peer review Commissioned; externally peer reviewed.