ISSN: 1080-6059
Fabrice Legros,*†‡§ Olivier Bouchaud,¶ Thierry Ancelle,#** Amandine Arnaud,*†
Sandrine Cojean,#†† Jacques Le Bras,¶#†† Martin Danis,*†§‡‡ Arnaud Fontanet,§§
and Rémy Durand,††¶ for the French National Reference Centers for Imported and
Autochthonous Malaria Epidemiology and Chemosensitivity Network
*Centre National de Référence de l'Epidémiologie du Paludisme
d'Importation et Autochtone, Paris, France; †University Pierre et Marie Curie,
Paris, France; ‡Institut de Recherche pour le Développement, Paris, France; §Institut
National de la Santé et de la Recherche Médicale, U511, Paris, France; ¶Hôpital
Avicenne and University Paris 13, Bobigny, France; #University Paris 5, Paris,
France; **Hôpital Cochin; Paris, France; ††Centre National de Référence pour la
Chimiosensibilité du Paludisme Hôpital Bichat-Claude Bernard, Paris, France; ‡‡Groupe Hospitalier Pitié-Salpêtrière, Paris
France; and §§Unité d'Epidémiologie des Maladies Emergentes, Institut Pasteur,
Paris, France
Suggested citation for this article
Abstract
Plasmodium falciparum malaria is a serious health hazard for
travelers to malaria-endemic areas and is often diagnosed on return to the
country of residence. We conducted a retrospective study of imported falciparum
malaria among travelers returning to France from malaria-endemic areas from 1996
through 2003. Epidemiologic, clinical, and parasitologic data were collected by
a network of 120 laboratories. Factors associated with fatal malaria were
identified by logistic regression analysis. During the study period, 21,888
falciparum malaria cases were reported. There were 96 deaths, for a
case-fatality rate of 4.4 per 1,000 cases of falciparum malaria. In
multivariate analysis, risk factors independently associated with death from
imported malaria were older age, European origin, travel to East Africa, and
absence of chemoprophylaxis. Fatal imported malaria remains rare and
preventable. Pretravel advice and malaria management should take into account
these risk factors, particularly for senior travelers.
Imported malaria is increasingly reported in Europe and North America, with an estimated 30,000 cases yearly (1,2). In 2000, the countries with the highest rates of imported malaria were France (≈8,000 estimated cases), United Kingdom (2,069 cases), United States (1,402 cases), Italy (986 cases), and Germany (732 cases) (3,4). Imported Plasmodium falciparum malaria is a serious health hazard for travelers to malaria-endemic areas, owing to the potentially severe illness and high case-fatality rates (case-fatality rate per 1,000: France 4; Italy 6.5; UK 8.5; USA 13; Germany 30.4) (4,5). Risk factors associated with fatal imported malaria are poorly known. Limited series have suggested that the fatality rate is significantly lower for migrants from malaria-endemic areas than for patients from areas not endemic for the disease (6–8). Antimalarial chemoprophylaxis, even incomplete or inappropriate, may also confer a degree of protection (9–11). Better knowledge of the characteristics and risk factors for fatal imported malaria might help to improve prevention and patient management. We retrospectively analyzed the main features of fatal imported falciparum malaria observed in France during 1996–2003 and compared them with those for nonfatal cases.
Imported malaria is not a mandatory notifiable disease in metropolitan France. The data for this study were collected by a reporting network of 120 selected hospital laboratories and were analyzed by the French National Reference Center for Imported and Autochthonous Malaria Epidemiology (CNREPIA). Participants of the network were asked to report imported malaria cases whenever the laboratory observed asexual forms of P. falciparum in a patient's blood film. Data from the national medical informatics systems and from 2 exhaustive studies (National Quality Control Survey) suggested that these cases represented 50%–55% of the total number of imported falciparum malaria cases in France during the study period (12,13). A standard 57-item questionnaire, completed by clinicians and biologists for each reported case, collected basic demographic, epidemiologic, clinical, and parasitologic information (including prophylaxis and treatment). In addition, a detailed clinical description was obtained for each fatal case.
The study population consisted of all P. falciparum–infected patients reported to CNREPIA during 1996–2003. Deaths occurring during hospitalization for malaria were considered malaria related. The case-fatality rate per 1,000 patients was calculated for all relevant exposure variables. Various exposure categories were created for the analysis: patients were divided into European travelers (persons born and residing in areas not endemic for malaria), European expatriates (residing in malaria-endemic African countries), African travelers (persons born in Africa who reside mostly in France or another country not endemic for malaria), African residents (persons born and residing in Africa), and others. Use of malaria chemoprophylaxis, as reported by the patients, was categorized as follows: no use; use of ineffective drugs (e.g., chloroquine, proguanil, pyrimethamine, sulfadoxine-pyrimethamine); and use of effective drugs (mefloquine, atovaquone-proguanil, doxycycline, and chloroquine-proguanil). Logistic regression was used to identify factors associated with fatal malaria and to estimate odds ratios and 95% confidence intervals (CIs) for the association between exposure variables and death. Dummy variables were used for variables with >2 categories. Variables with p<0.25 were introduced in the multivariate logistic regression model. A manual backward stepwise approach was used to remove nonsignificant variables, and only variables with p<0.05 were retained in the final model. Interactions were sought by introducing interaction terms in the logistic regression model and testing for their significance (p<0.05). Because data were missing for the variables "region of malaria acquisition" (n = 9), "chemoprophylaxis" (n = 2,366), and "time between onset and diagnosis" (n = 3,845), multiple imputation for missing data was performed for the final model by using the multivariate imputation by chained equations (MICE) method described by Van Buuren et al. (14). The MICE method involves imputations of missing values by appropriate regression models and generation of multiple datasets (in our case, 5) to take into account the uncertainty involved in imputing the missing values. Standard complete-data methods are then used on each dataset, and results are combined to produce estimates with CIs and p values. Statistical analysis was performed by using EpiInfo, version 3.3 (Centers for Disease Control and Prevention, Atlanta, GA, USA) and Stata 8 (Stata Corporation, College Station, TX, USA).
During the 1996–2003 period, 27,085 malaria cases were reported to CNREPIA; 21,888 of these patients had P. falciparum malaria, which constituted the study population. Included were 20,436 (93.4%) uncomplicated cases, 825 (3.8%) severe cases, 433 (2%) asymptomatic cases, 161 (0.7%) cases of hyper-reactive malarial splenomegaly, and 33 (0.1%) unspecified cases. Cases attributable to species other than P. falciparum and cases attributable to species combinations that included P. falciparum were not considered in the analysis (no fatal cases of imported malaria due to Plasmodium species other than P. falciparum were reported during the study period). The annual number of malaria cases reported by the network increased until 1999–2000, then stabilized at ≈3,000 cases per year (Table 1). Most patients were male (sex ratio M:F = 1.7), and the median (range) age was 29.0 (0–96) years. Most patients had acquired malaria in Africa: 59.2% in West Africa, 26.2% in central Africa, 11.2% in Madagascar and the Comoros Islands, and 0.9% in East Africa (Table 2 and Appendix Table). Others (2.5%) had returned mainly from French Guiana, Haiti, India, Sri Lanka, Thailand, and Indonesia. African travelers were most numerous (44.6%), followed by European travelers (26.5%), African residents (12.9%), and European expatriates living in Africa (5.4%); "others" represented 10.6%. Few patients (30.4%) reported taking effective chemoprophylaxis, and more than half the patients had not taken any. The median duration of stay was 32 days (interquartile range 21–62). The median time from return to symptom onset was 6 days (interquartile range 1–12), and 10% of patients had their first symptoms before returning to France. The median time from symptom onset to diagnosis was 3 days (interquartile range 1–6). Compared with Europeans, Africans were more likely to seek care at a hospital (73.0% vs. 62.7%, p<0.001). Diagnoses were fairly evenly distributed between spring-summer (55.5%) and fall-winter (45.5%). At diagnosis, 7.9% of patients had high-level parasitemia (>5% of parasitized erythrocytes).
Ninety-six patients died of malaria; 55 were European travelers; 12, European expatriates; 11, African travelers; 10, African residents; and 8, other. One case was diagnosed postmortem. Repatriations for medical reasons occurred in 13 of 96 fatal cases (1 African resident, 5 European travelers, and 7 European expatriates). Among the patients who died, the sex ratio (M:F) was 3.3, and the median (range) age was 47 (2–92) years. Three study participants who died were <15 years of age, and 5 were >70 years of age. Thirty (31.3%) of the patients who died had taken antimalarial chemoprophylaxis, and 2 were reported to have correctly taken prophylaxis appropriate for the region visited.
Clinical data for patients who died are shown in Table 3. Fever was the most common initial symptom. All patients, except the one whose case was diagnosed postmortem, were hospitalized and received antimalarial therapy within 12 hours of diagnosis. Median parasitemia at admission was high (10%), although 10% of patients had parasitemia <1%. One case of black water fever and 1 case of splenic rupture were observed. Forty-three (44.8%) patients required mechanical ventilation, and 24 (25%) required hemofiltration. Three (3.1%) patients underwent exchange transfusion.
The case-fatality rate was 4.4 deaths per 1,000 cases during the study period and did not change over time (calendar years) (standard χ2 for trends test; Table 1). Many factors were associated with an increased risk for death in univariate analysis, including older age, male sex, European origin, travel to East Africa, short stays (<15 days), time to diagnosis, and initial visit to a general practitioner (Appendix Table). However, male sex, short stays, visit to a general practitioner, and diagnosis during the fall-winter season were no longer predictive of death after age and ethnic origin were controlled for in multivariate analysis. Table 4 shows the results of multivariate analysis. Low hemoglobin levels (<8 g/dL), low platelet counts (<50×109/L), high leukocyte counts (>10×109/L), and high-level parasitemia (>5%) were all associated with increased risk for death among patients with measurements for these variables (Appendix Table). The risk factors for death identified in this study were not different between Africans and Europeans (test of interaction not significant).
To our knowledge, this is the largest retrospective study aimed at identifying risk factors for fatal imported malaria. France has large numbers of migrants of African origin. Those populations are particularly at risk of acquiring malaria when visiting friends and relatives (15–17). Migrants of African origin travel for longer periods in highly malaria-endemic areas (globally, 75% of malaria infections and 89% of P. falciparum infections in travelers are acquired in sub-Saharan Africa [18]), are less likely to have pretravel encounters with a healthcare provider, and are therefore unlikely to take antimalarial prophylaxis (18). These factors explain why France, in comparison with many other European countries or the United States, has so many malaria cases and why the country appears to receive disproportionally high numbers of malaria-infected returning travelers from Africa (19), rather than from Asia or South America. As a result, P. falciparum is overrepresented in imported malaria in France in comparison with other industrialized countries. The case-fatality rate (4.4 per 1,000 cases) is among the lowest in the World Health Organization (WHO) 2004 Computerized Information System for Infectious Diseases database, probably because of the large proportion of African migrants in our study population.
The following characteristics were independently associated with death from falciparum malaria, according to multivariate analysis: older age, origin in an area not endemic for malaria, infection in East Africa, and no effective chemoprophylaxis. Increasing age has also been a risk factor for fatal falciparum malaria in smaller studies (10,11,13,20,21). As has been reported elsewhere (21), we found a gradual increase in risk over the entire age spectrum, resulting in particularly high risk among elderly patients. This factor should be taken into account when offering pretravel health advice, particularly as the age of international travelers increases.
Severe malaria and death were particularly frequent among nonimmigrants, as previously reported in smaller series (6–8,13,22). These results are consistent with the hypothesis of persistent acquired immunity, even after several years of nonexposure, which may partially protect African immigrants from fatal malaria, as previously shown for severe forms of malaria (23). Genetic factors, selected at the population level over centuries of exposure to the parasite, may also partly explain the relative protection of African immigrants compared with Europeans (24,25).
Except for the Comoros Islands (0.5 cases per 1,000 cases of falciparum malaria), countries accounting for large numbers of cases in this study (Senegal, Cameroon, Mali, and Côte d'Ivoire) had similar case-fatality rates (3.2 to 5.4 cases per 1,000 cases of falciparum malaria). Most travelers returning from the Comoros Island were migrants; few were European tourists. East African countries such as Djibouti, Kenya, Mozambique, and Tanzania accounted for fewer cases but a disproportionate number of fatalities (34.1 cases per 1,000). Ben-Ami et al. recently reported a high rate of severe malaria (6 of 29 cases, including 1 death) among patients who visited Mombassa, Kenya (26). More generally, Krause et al. reported that falciparum malaria acquired in Africa had a higher case-fatality rate than falciparum malaria acquired elsewhere (11). During the period of our study, an increase in malaria deaths, probably related to higher levels of drug resistance, was seen in residents of East Africa but not in those of West Africa (27). Those observations are not necessarily linked, but particular attention should be paid to travelers returning from these areas. Further studies are needed to confirm and explain these findings.
The risk for death was higher when prophylaxis was absent or ineffective than when appropriate prophylaxis was taken. On the basis of interview data, only 2 authentic failures of prophylaxis were suspected among our patients, but drug and metabolite concentrations were not assayed. These results are consistent with those of Krause et al., who reported that study participants who had taken chemoprophylaxis with chloroquine-proguanil were less likely to die than those who had not taken chemoprophylaxis (11). These results once again underline the importance of recommending antimalarial prophylaxis for travelers to malaria-endemic areas (28).
Bruneel et al. reported that platelet counts were significantly lower in patients who eventually died of P. falciparum–infected patients than in survivors (8) and that leukocyte counts also tended to be higher. In our population, a platelet count <50×109/L was associated with increased risk of dying; this effect was particularly marked at counts <10×109/L. Disseminated intravascular coagulation, which is associated with marked thrombocytopenia, was frequent in patients who later died (Table 3). Leukocyte counts >10×109 cells/L were also associated with increased mortality, and the effect was particularly marked at counts >15×109 cells/L. Hyperleukocytosis in this setting may be related to cytokine or cortisol release or to bacterial (especially nosocomial) infections, which were frequent in our patients who eventually died (Table 3). Bruneel et al. found that hyperparasitemia (>5%), a WHO severity criterion, was weakly linked to death. Although the number of circulating parasitized erythrocytes may not accurately reflect the number of adherent red cells in deep organ capillaries, which are the source of most clinical complications (29), hyperparasitemia appeared to be an important predictor of death in our series. However, parasitemia and some biologic data (hemoglobin, platelet counts, and leukocyte counts) could not be included in our multivariate analysis because an excessive amount of data were missing. In particular, hemoglobin, platelet count, and leukocyte count data were missing for more than half the patients.
A limitation of our study is that our network for collecting data accounted for only 50%–55% of total malaria cases imported to France. Two annual exhaustive studies (National Quality Control Survey [30]; F. Legros, unpub. data) suggest that representativeness of our sites was correct. Thus, risk factors associated with death in imported malaria would not likely differ for cases not seen in our network.
Severe and fatal malaria, even though it is eminently preventable, continues to be seen in areas that are not endemic for malaria (31,32). Fatal cases are rare in patients who take appropriate prophylaxis. With the current increases in intercontinental travel, numbers of elderly travelers, risk for transmission in malaria-endemic areas, and drug-resistant strains of P. falciparum (16,33), the numbers of fatal cases of imported malaria should be carefully monitored in the coming years in France and other industrialized countries. Preventive measures remain necessary for all travelers, including those from Africa, for whom adherence is often poor (34). Posttravel care should also be reinforced to reduce the interval between symptom onset and diagnosis (35,36).
We are deeply indebted to all members of the staff, clinicians and biologists, at each CNREPIA and Centre National de Référence pour la Chimiosensibilité du Paludisme corresponding site.
Financial support was provided by the French Ministry of Health (Direction Générale de la Santé). S.C. is the recipient of a thesis fellowship from the French Research Ministry.
Mr Legros is a senior parasitologist/epidemiologist at the Institut pour la Recherche et le Développement and at the French Malaria Reference Center. His main research interests are malaria epidemiology and arbovirology.
Table 1. Distribution of falciparum malaria cases and deaths by calendar
year, 1996–2003, France
Table 2. Distribution of falciparum malaria cases and deaths by country of
acquisition, 1996–2003
Table 3. Clinical data for 96 patients with fatal malaria, 1996–2003,
France
Table 4. Factors independently associated with deaths among patients
treated for falciparum malaria in French hospitals, 1996–2003 (n = 21,888)
Appendix Table. Factors associated with deaths among patients treated for
falciparum malaria in French hospitals, 1996–2003 (n = 21,888)
Legros F, Bouchaud O, Ancelle T, Arnaud A, Cojean S, Le Bras J, et al. Risk factors for imported fatal Plasmodium falciparum malaria, France, 1996–2003. Emerg Infect Dis [serial on the Internet]. 2007 Jun [date cited]. Available from http://www.cdc.gov/EID/content/13/6/883.htm
Please use the form below to submit correspondence to the authors or contact them at the following address:
Rémy Durand, Laboratoire de Parasitologie-Mycologie, Hôpital Avicenne, 125 rue de Stalingrad, 93009 Bobigny CEDEX, France; email: remy.durand@avc.aphp.fr
Please note: To prevent email errors, please use no web addresses, email addresses, HTML code, or the characters <, >, and @ in the body of your message.
Please contact the EID Editors at eideditor@cdc.gov
The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the U.S. Department of Health and Human Services, the Public Health Service, the Centers for Disease Control and Prevention, or the authors' affiliated institutions. Use of trade names is for identification only and does not imply endorsement by any of the groups named above.
This page posted June 1, 2007
This page last reviewed June 1, 2007
Centers for Disease Control and Prevention, 1600 Clifton Rd, Atlanta, GA 30333, U.S.A
Tel: (404) 639-3311 / Public Inquiries: (404) 639-3534 / (800) 311-3435