INTENSIVE CARE ADMISSIONS FOR CHILDREN WITH IMPORTED MALARIA IN THE UNITED KINGDOM

Abstract

This study describes 977 children with imported malaria in England and Wales between 2004 and 2008, focusing on 29 (3.0%) patients admitted to intensive care, of whom 10 had cerebral malaria, 4 required inotropes, and 1 had concurrent septicemia. The remaining 14 were admitted for monitoring only. None died, but 1 child developed cerebellar infarction.

The United Kingdom has one of the highest rates of imported malaria, with 1500 to 2000 cases reported annually, and children account for 10% to 15% of cases.^1,2 In a recent, prospective, enhanced national surveillance study, Plasmodium falciparum, the species responsible for most of the severe and fatal malaria cases worldwide, caused 85% of imported childhood malaria cases in the United Kingdom and Ireland.³ In addition, one-third of children with P. falciparum malaria fulfilled the World Health Organization criteria for severe or potentially complicated malaria, although only a quarter of those with severe malaria required intensive care, mainly for neurologic complications.³ Few studies have focused on intensive care admissions for imported childhood malaria cases in nonendemic countries, where presentation and course of severe childhood malaria, the proportion developing cerebral malaria, and the mortality by syndrome vary substantially. This study aimed to describe children with imported malaria admitted to pediatric intensive care units (PICU) in England and Wales during a 5-year period.

METHODS

Imported malaria describes an infection that was acquired in a malaria endemic area by a tourist or a native but was diagnosed in a nonendemic area after development of clinical symptoms. The Health Protection Agency Malaria Reference Laboratory (MRL) identifies cases from routine reports submitted by UK hospital laboratories, positive malaria slides sent to MRL for species verification, and notifications of malarial deaths by the Office for National Statistics.^2,3

The Pediatric Intensive Care Audit Network (PICANet) records core demographic and clinical details of the management of all critically ill children in National Health Service PICU, using standardized data collection forms (available at: www.picanet.org). We requested anonymized data on all children <16 years with malaria admitted to PICU in England and Wales between 1 January, 2004 and 31 December, 2008 (data from Scotland and Northern Ireland were not available for the entire period). Data were analyzed using Stata version 11.0 (available at: www.stata.com). Ages are reported as medians and interquartile ranges (IQR) and compared using the Mann-Whitney U test, and proportions are compared with the χ² test or Fisher exact test as appropriate.

RESULTS

National Data.

There were 977 childhood malaria cases in England and Wales reported to the MRL between 2004 and 2008 inclusive. The median age at diagnosis was 9 years (IQR, 5–13 years), 19.8% (193 cases) were in children aged <5 years, and 60.0% were male. P. falciparum accounted for 79.6% (778 cases, including 12 cases of P. falciparum coinfection with other Plasmodium species), followed by Plasmodium vivax (126 cases, 12.9%), Plasmodium ovale (51 cases, 5.2%), and Plasmodium malariae (22 cases, 2.3%). The number of reported cases declined gradually from 326 in 1999 to 161 in 2008 (Fig. 1). Cases showed a bimodal distribution peaking in January and August/September. Two children died (case fatality ratio, 0.2%). These 2 were Black-African children aged 5 to 14 years who had traveled to Sub-Saharan Africa without appropriate antimalarial prophylaxis and became unwell within 2 weeks of return to the United Kingdom. Both children were pronounced dead soon after being brought to the Emergency Department.

FIGURE 1.: Number of childhood malaria cases in England and Wales: 1999 to 2008.

Pediatric Intensive Care Units Admissions.

Over the same time, 29 children with malaria were admitted to PICU (6˜ cases/year; range, 3–10 cases/year), accounting for 3% of all cases (range, 1.7%–4.8% per year). Of these, 16 (55%) were male and none had comorbidities. Twenty children (69%) were admitted to a London PICU, which was similar to the proportion of malaria cases diagnosed in London (606/977, 62.0%; χ² = 0.58; P = 0.45). PICU admissions for malaria increased during the months with the highest diagnosis of malaria, with greater than half the children (16/29 cases, 55%) admitted to PICU in August, September, and January. The median age of children with malaria admitted to PICU was 4 years (IQR, 2–9 years), which were significantly lower than the median age of all children diagnosed in the same period (P = 0.0019). Similarly, the proportion of children aged <5 years (15/29 cases, 52%) who were admitted to PICU was higher (χ² = 17.6; P < 0.0001)—7.8% (15/193 cases) children.

Ten children, including 2 with respiratory distress and 1 with hypotensive shock, presented with coma requiring mechanical ventilation for a median of 2 days (range, 1–4 days). Four others presented with hypotensive shock requiring inotropic support and mechanical ventilation. One other child presented with respiratory distress and had concurrent enteropathogenic Escherichia coli and Salmonella typhi septicemia, but did not require mechanical ventilation. The remaining 14 cases were admitted for observation only. Metabolic acidosis was uncommon, with only 3 children having a base excess more negative than −5 mmol/L. Children admitted to PICU for monitoring only were more likely to have been admitted from the same hospital (10/14 [71%] vs. 5/15 [33%]; χ² = 4.2; P = 0.040). None received renal replacement therapy (dialysis, filtration, or exchange) or an exchange transfusion. Most children (21/29, 72%) remained in PICU for ≤48 hours and none was admitted for more than 6 days. One child developed cerebellar infarction but none died.

DISCUSSION

We have recently reported the results of a prospective study on children with imported malaria diagnosed in the United Kingdom and the Republic of Ireland over a 13-month period.³ In that study, most cases occurred in Black-African children who developed P. falciparum malaria while visiting family and relatives in West Africa during school holidays without appropriate antimalarial prophylaxis.³ The current study focuses on children with malaria who were admitted to PICU over a longer time-period (2004–2008), and although data collected through national surveillance are limited, we would expect the clinical, epidemiologic, and laboratory features of children with malaria to be similar to those reported in our previous study.³

The findings of this study demonstrate that lessons from endemic countries are relevant in the PICU setting, although presentations are more likely to be neurologic than in African series, possibly because children with severe malaria presenting in non-endemic countries tend to be older. In particular, respiratory distress is a serious sign in children with malaria,⁴ and clinicians should have a high index of suspicion for concomitant bacterial infection. Series suggesting a link between severe malaria and Gram-negative infection in Africa tend to have predominantly younger children but, even in the older age group, bacterial coinfection was an issue.^5,6

In our enhanced national surveillance study, it was concerning that one-third of children with P. falciparum malaria fulfilled World Health Organization criteria for severe or complicated malaria, although only a small proportion were subsequently admitted to PICU.³ We identified only one other study of 15 PICU cases in a nonendemic country, conducted in France between 1990 and 1995.⁷ Therefore, although our study is limited by the amount of information collected through PICANet, it does provide the best available data on this unique cohort of children. Considering the high proportion of children with severe or complicated P. falciparum malaria, the finding of so few cases admitted to PICU is surprising but consistent with our enhanced surveillance. We can, therefore, estimate that although a significant proportion of children with P. falciparum malaria present with World Health Organization-defined severe or complicated malaria, only 1.5% are deemed to require intensive care support, with a similar proportion admitted to PICU for close monitoring. Both this and the French study, where all 15 children had neurologic complications, confirm that most children with life-threatening imported malaria have neurologic involvement (coma, convulsions, or reduced consciousness) rather than respiratory distress or severe anemia.⁷ In the French study, 8 children required ventilation, 4 received volume expansion, and 1 required inotropic support.⁷ The predominance of neurologic complications contrasts with pediatric malaria in African endemic areas,⁴ and is most likely to reflect the older age of children presenting with imported severe malaria. Older age is associated with a higher probability of neurologic presentation in Africa,⁸ while the higher proportion of children presenting with neurologic complications in Asian studies is also most likely because of older age at presentation.^9,10 Presentation with circulatory collapse in malaria endemic countries is rare (0.4% and 2.9% in 2 of the studies) but associated with very high case fatality (71% and 82%, respectively).^4,9

Although it is reassuring that case fatality remains low, 2 children died between 2004 and 2008 and 2 other children died in the preceding 5 years.² Previous studies have identified failure to take recommended antimalarial chemoprophylaxis, refusal or delay in seeking medical care, and misdiagnosis as major risk factors for fatal malaria, particularly among nonimmune travelers.¹ Diagnostic delay is also associated with increased risk of severe malaria and intensive care admission.^1,7

Our study highlights several important messages. First, children traveling to malaria-endemic countries without appropriate preventive measures are not only at risk for malaria but also of other travel-related infections such as typhoid. Clinicians should, therefore, consider bacterial coinfection in children presenting with severe or complicated malaria, take appropriate cultures at presentation, and manage them accordingly, but with a low threshold for empiric antibiotics while culture results are awaited. Second, younger children (<5 years) are not only more likely to present with severe malaria,³ but are also more likely to be admitted to PICU. Third, hypotensive shock is an uncommon but potentially fatal presentation of malaria and may require aggressive fluid resuscitation and inotropic support. Finally, although case fatality remains low, a significant proportion of children with P. falciparum malaria presenting to PICU have coma, which may be associated with long-term neurologic sequelae. Clinicians should follow-up children with severe malaria after hospital discharge to identify and manage any long-term complications. Guidelines for the management of severe imported childhood malaria should emphasize the importance of neurologic complications and cardiovascular shock in children presenting with severe malaria.

ACKNOWLEDGMENTS

The authors thank PICANet for providing anonymized data on intensive care admissions for children with malaria. The authors also thank Valerie Smith and Marie Blaze for providing the MRL data. PLC is supported by the UCL Hospitals Comprehensive Biomedical Research Centre Infection Theme.