Risk Assessment and Prevention of Malaria Among Italian Troops in Afghanistan, 2002 to 2011


  • Mario S. Peragallo MD,

    Corresponding author
    1. Preventive Medicine Branch, Centro Studi e Ricerche di Sanità e di Veterinaria dell'Esercito, Rome, Italy
    Search for more papers by this author
  • Giuseppe Sarnicola MD,

    1. Medical Services Organization Office, Department of Health, Comando Logistico dell'Esercito, Rome, Italy
    Search for more papers by this author
  • Daniela Boccolini BSc,

    1. Vector-Borne Diseases and International Health Unit, Department of Infectious, Parasitic, and Immuno-mediated Diseases, Istituto Superiore di Sanità, Rome, Italy
    Search for more papers by this author
  • Roberto Romi BSc,

    1. Vector-Borne Diseases and International Health Unit, Department of Infectious, Parasitic, and Immuno-mediated Diseases, Istituto Superiore di Sanità, Rome, Italy
    Search for more papers by this author
  • Giacomo Mammana MD

    1. Preventive Medicine Branch, Centro Studi e Ricerche di Sanità e di Veterinaria dell'Esercito, Rome, Italy
    Search for more papers by this author



Malaria prevention policy is different among coalition troops in Afghanistan, ranging from the combined use of suppressive and terminal chemoprophylaxis to the absence of any prophylactic regimen. The objective of this study was to assess the compliance with malaria prevention measures and the risk of malaria among Italian troops in Afghanistan.


Target population was the cohort of 32,500 army soldiers deployed in Afghanistan, 2002 to 2011; eligible subjects were the 21,900 soldiers stationed in endemic areas, who were prescribed mefloquine chemoprophylaxis. Adherence to chemoprophylaxis was assessed by a cross-sectional study in a volunteer sample of 5,773 (26.4%) of eligible subjects. The risk of malaria was assessed by detecting malaria cases in the target population.


Mefloquine chemoprophylaxis was administered to 4,123 (71.4%) of the 5,773 enrolled soldiers and 3,575 (86.7%) of these took it regularly; however, compliance dropped from 80.9% (2,592/3,202) in 2002 to 2006 to 59.5% (1,531/2,571) in 2007 to 2011 (p < 0.01). Adverse events were reported by 875 (21.2%) of the 4,123 soldiers taking mefloquine, but caused irregularity or interruption of chemoprophylaxis only in 48 (1.2%) and 113 (2.7%) subjects, respectively. No serious adverse events were reported. No malaria cases occurred in Afghanistan, and one Plasmodium vivax case was reported in Italy, yielding an incidence rate of 3.24 cases per 10,000 person-months of exposure (1/3,091) during the transmission season of 2003.


In spite of the decreasing compliance with chemoprophylaxis, suggesting a low perception of the risk of malaria, this study confirmed the good tolerability of mefloquine in the military. The risk of malaria for Italian troops in Afghanistan was very low, and chemoprophylaxis was suspended in 2012. A similar policy may be adopted by the generality of International Security Assistance Force troops, and any chemoprophylaxis may be restricted to soldiers stationing in areas where the risk of malaria is substantial.

When the International Security Assistance Force (ISAF) was deployed in Afghanistan in 2001, malaria was perceived as a major problem. The disease is hypoendemic in most of Afghanistan, and transmission is seasonal, from May to November, in all areas below 2,000 m above sea level. Plasmodium vivax malaria is predominant and accounted for 80% to 90% of all cases in 2001, whereas the remaining proportion was due to Plasmodium falciparum.[1] Suppressive chemoprophylaxis was implemented by most of the ISAF forces, and mefloquine was the generally preferred agent.[2] This regimen was adopted by Italian troops stationed in endemic areas during the transmission season; no terminal prophylaxis with primaquine was prescribed. Additional recommended mosquito bite protection measures included the use of permethrin-treated bed nets and uniforms, and insect repellent skin lotions.

The purpose of this article was to investigate compliance and tolerability of mefloquine chemoprophylaxis and other malaria prevention measures, and to assess the risk of malaria among Italian troops in Afghanistan.


Data Sources

Adherence to malaria prevention measures and occurrence of adverse events (AEs) were evaluated through the individual post-deployment health questionnaire. This form is routinely but voluntarily completed by soldiers after returning home: they have to answer questions concerning personal health during their duty period abroad and compliance with malaria prevention measures.

Chemoprophylaxis was considered: regular, if mefloquine had been taken for the whole duration of deployment at the recommended doses and intervals (250 mg every 7 days); irregular, if mefloquine had been taken less frequently than recommended; and interrupted or not performed, if mefloquine had been started and subsequently interrupted, or never started.

To detect all serious adverse events (SAEs) possibly related to mefloquine, we examined all notifications sent to the Army Logistic Command, Health Department, concerning medically evacuated personnel from Afghanistan from 2002 to 2011 for any neuropsychiatric reason.

Data concerning malaria cases, the involved Plasmodium species, the period and place of exposure, and the individual compliance with prevention measures can be taken from notification forms.

Study Population and Design

All army members deployed from 2002 to 2011 in malaria-endemic areas of Afghanistan during the transmission season (May to November) were eligible for the study. According to available information (Figure 1), we intended as endemic for malaria all the provinces of deployment: Kabul, Herat, Badghis, Ghor, and Farah.

Figure 1.

Afghanistan provinces can be clustered into three groups with different risk levels of malaria.[3] The risk is, however, far from being homogeneous within each province; transmission occurs mainly in river valleys where rice fields are common.[4] The provinces comprised within the Regional Command West represent the main deployment areas of the Italian forces since 2007.

Adherence to malaria chemoprophylaxis and mosquito bite prevention measures were evaluated through a retrospective study, by computing the proportion of soldiers using these measures, the regularity of chemoprophylaxis, and the occurrence of AEs. Definitions of the International Conference on Harmonisation Guidelines (http://ichgcp.net/1-glossary) were used to identify AEs and SAEs. The association between mefloquine exposure and SAEs was tested by computing the evacuation rates for any neuropsychiatric cause when mefloquine prophylaxis was taken and when it was suspended, using the number of medical evacuations as numerators and the person-months of deployed personnel as denominators.

A malaria case was defined as a microscopically confirmed case occurring either in Afghanistan or in Italy. The risk of malaria was assessed by identifying all malaria cases that occurred in the cohort of deployed soldiers and then computing the incidence rate of malaria.

Data Analysis

Information from questionnaires was stored in a database made in EpiInfo v. 3.5.3 software (Centers for Disease Control and Prevention, January 26, 2011). Statistical analysis was performed using the Mantel–Haenszel chi-square test to compare the difference in compliance with malaria prophylaxis and the occurrence of AEs. Statistical tests for all analyses were two-sided and significance was considered at a p value of <0.01.

The evacuation rates when mefloquine prophylaxis was taken and when it was suspended were compared by computing the conditional maximum likelihood estimate of rate ratio and the mid-p exact 95% confidence intervals (RR, 95% CIs).

Ethical Aspects

Soldiers gave their written consent in the post-deployment questionnaire for the processing of reported data. Informed consent from soldiers evacuated or affected by malaria was not necessary, because reporting of medical evacuations is established by military regulations, and malaria notification is mandatory: all cases must be notified to the national surveillance system, and the Istituto Superiore di Sanità is charged with microscopic diagnosis confirmation of cases. Moreover, the armed forces operate a parallel surveillance system, which covers all reportable infectious diseases occurring among soldiers. All information concerning patients and each individual included in the study was rendered anonymous.


Enrolled Subjects

The Italian army contribution to ISAF was of 32,500 soldiers from 2002 to 2011. Of these, 21,900 (67.4%) were eligible for the study: 6,900 were mainly stationed in Kabul from 2002 to 2006, and 15,000 mainly in the provinces of Herat, Farah, Badghis, and Ghor from 2007 to 2011. A total of 6,938 soldiers filled the questionnaire: 5,773 of these (83.2%) met the eligibility criteria and were included in the study. Their demographic and deployment data are summarized in Table 1.

Table 1. Demographic and deployment characteristics of 5,773 Italian army soldiers stationed in Afghanistan, 2002 to 2011
CharacteristicsYears 2002 to 2006, N (%)Years 2007 to 2011, N (%)Total, N (%)
Male3,132 (97.8)2,488 (96.8)5,620 (97.3)
Female70 (2.2)83 (3.2)153 (2.7)
Total3,202 (100.0)2,571 (100.0)5,773 (100.0)
Age (years)   
20–292,535 (79.2)1,501 (58.4)4,036 (69.9)
30–39462 (14.4)779 (30.3)1,241 (21.5)
≥40205 (6.4)291 (11.3)496 (8.6)
Total3,202 (100.0)2,571 (100.0)5,773 (100.0)
Officers275 (8.6)331 (12.9)606 (10.5)
Non-commissioned officers558 (17.4)558 (21.7)1,116 (19.3)
Enlisted men2,293 (71.6)1,674 (65.1)3,967 (68.7)
Unknown76 (2.4)8 (0.3)84 (1.5)
Total, N (%)3,202 (100.0)2,571 (100.0)5,773 (100.0)
Duration of deployment (days)   
≤90603 (18.8)140 (5.4)743 (12.9)
91–120903 (28.2)126 (4.9)1,029 (17.8)
121–150950 (29.7)435 (16.9)1,385 (24.0)
151–180473 (14.8)886 (34.5)1,359 (23.5)
>180247 (7.7)984 (38.3)1,231 (21.3)
Unknown26 (0.8)026 (0.5)
Total3,202 (100.0)2,571 (100.0)5,773 (100.0)
First prescription of mefloquine prophylaxis   
Yes3,006 (93.9)2,211 (86.0)5,217 (90.4)
No196 (6.1)360 (14.0)556 (9.6)
Total3,202 (100.0)2,571 (100.0)5,773 (100.0)

Chemoprophylaxis Compliance

Over the study period, 72.0% of the troops (4,160/5,773) took malaria chemoprophylaxis and the greatly prevalent regimen was mefloquine (4,123/5,773 = 71.4%); however, the proportion of subjects taking mefloquine dropped from 80.9% (2,592/3,202) in 2002 to 2006 to 59.5% (1,531/2,571) in 2007 to 2011 (p < 0.01); correspondingly, the proportion of subjects who did not use any chemoprophylaxis increased from 18.5% (591/3,202) to 39.8% (1,022/2,571) (p < 0.01). Mefloquine was taken regularly by 86.7% of soldiers (3,575/4,123), and subjects taking irregularly or interrupting chemoprophylaxis for any cause were 4.0% (164/4,123) and 9.3% (384/4,123), respectively; no significant changes concerning these proportions were observed between 2002 to 2006 and 2007 to 2011. Personal protection measures were adopted by a minority of soldiers: the use of permethrin-treated bed nets and skin repellents was below 20%, whereas the use of permethrin-treated uniforms significantly dropped from 24.9% (798/3,202) in 2002 to 2006 to 14.3% (367/2,571) in 2007 to 2011 (Table 2).

Table 2. Adherence to preventive measures of malaria among 5,773 Italian army soldiers deployed in Afghanistan, 2002 to 2011
Preventive measures of malariaYears 2002 to 2006, N (%)Years 2007 to 2011, N (%)Total, N (%)p Valuea
  1. a

    Analysis by corrected Mantel–Haenszel (two-sided) chi-square test, adjusted by age, duration of deployment, gender, rank, and first deployment with prescription of mefloquine prophylaxis.

  2. b

    Chloroquine/proguanil (N = 22), doxycycline (N = 12), and atovaquone/proguanil (N = 3).

  3. c

    Percentages do not total 100 because of rounding of decimals.

  4. d

    Regular use (at least once a day).

Malaria chemoprophylaxis during deployment    
Mefloquine2,592 (80.9)1,531 (59.5)4,123 (71.4)<0.01
Other regimensb19 (0.6)18 (0.7)37 (0.6)0.06
None591 (18.5)1,022 (39.8)1,613 (27.9)<0.01
Total3,202 (100.0)2,571 (100.0)5,773 (99.9)c 
Mefloquine chemoprophylaxis    
Regular2,318 (89.4)1,257 (82.1)3,575 (86.7)0.11
Irregular72 (2.8)92 (6.0)164 (4.0)0.32
Interrupted202 (7.8)182 (11.9)384 (9.3)0.02
Total2,592 (100.0)1,531 (100.0)4,123 (100.0) 
Mosquito bite prevention measures    
Permethrin-treated bed nets    
Yesd624 (19.5)464 (18.0)1,088 (18.8)0.13
No2,578 (80.5)2,107 (82.0)4,685 (81.2) 
Total3,202 (100.0)2,571 (100.0)5,773 (100.0) 
Skin repellents    
Yesd539 (16.8)469 (18.2)1,008 (17.5)<0.01
No2,663 (83.2)2,102 (81.8)4,765 (82.5) 
Total3,202 (100.0)2,571 (100.0)5,773 (100.0) 
Permethrin-treated uniforms    
Yesd798 (24.9)367 (14.3)1,165 (20.2)<0.01
No2,404 (75.1)2,204 (85.7)4,608 (79.8) 
Total3,202 (100.0)2,571 (100.0)5,773 (100.0) 

The main causes of irregular and interrupted chemoprophylaxis were, respectively, forgetfulness of the weekly doses and occurrence of AEs. Although it was not explicitly requested, 7.6% of subjects interrupting prophylaxis (29/384) and 2.9% of those who never started it (47/1,613) recorded in the questionnaire that prophylaxis had been discarded because of the “lack of mosquitoes” or the “absence of malaria cases in the deployment area”; among the 1,613 soldiers who never started chemoprophylaxis, 1,514 (93.9%) did not give any information about this choice (Table 3).

Table 3. Motivations for irregular, interrupted, or missed mefloquine chemoprophylaxis among 5,773 Italian army soldiers deployed in Afghanistan, 2002 to 2011
MotivationsYears 2002 to 2006, N (%)Years 2007 to 2011, N (%)Total, N (%)p Valuea
  1. a

    Analysis by corrected Mantel–Haenszel (two-sided) chi-square test, adjusted by age, duration of deployment, gender, rank, and first deployment with prescription of mefloquine prophylaxis.

  2. b

    Percentages do not total 100 because of rounding of decimals.

For irregular prophylaxis    
Forgetfulness30 (41.7)47 (51.1)77 (47.0)0.85
Adverse events18 (25.0)11 (12.0)29 (17.7)0.13
Other causes9 (12.5)4 (4.3)13 (7.9)0.57
Unknown15 (20.8)30 (32.6)45 (27.4)<0.01
Total72 (100.0)92 (100.0)164 (100.0) 
For interrupted prophylaxis    
Forgetfulness32 (15.8)34 (18.7)66 (17.2)0.68
Adverse events45 (22.3)52 (28.6)97 (25.3)0.03
Considered unnecessary12 (5.9)17 (9.3)29 (7.6)0.43
Other causes48 (23.8)16 (8.8)64 (16.7)<0.01
Unknown65 (32.2)63 (34.6)128 (33.3)0.98
Total202 (100.0)182 (100.0)384 (100.1)b 
For missed prophylaxis    
Forgetfulness1 (0.2)34 (3.3)35 (2.2)<0.01
Considered unnecessary15 (2.5)32 (3.1)47 (2.9)0.64
Other causes2 (0.3)15 (1.5)17 (1.1)<0.01
Unknown573 (97.0)941 (92.1)1,514 (93.9)0.02
Total591 (100.0)1,022 (100.0)1,613 (100.1)b 

Chemoprophylaxis Tolerability

Among the 4,123 subjects who started mefloquine, 875 (21.2%) reported untoward symptoms attributed to chemoprophylaxis: AEs were gastrointestinal in 515 (58.9%), neuropsychiatric in 389 (44.5%), and others in 99 (11.3%). Every subject could report more than one symptom, and percents do not equal 100. Gastrointestinal AEs were represented by diarrhea in 336 of 875 subjects (38.4%), nausea in 246 (28.1%), abdominal pain in 177 (20.2%), and vomiting in 57 (6.1%); neuropsychiatric AEs were asthenia in 246 (28.1%), insomnia in 224 (25.6%), strange dreams in 65 (7.4%), and vertigo in 56 (6.4%); and additional AEs were pruritus and other minor symptoms in 35 (4.0%). AEs were reported in the first 2 weeks of chemoprophylaxis by 435 soldiers (49.7%) and in the third week by 153 (17.5%).

Any AEs were reported by 714 (20%) of the 3,575 soldiers regularly taking mefloquine, but their occurrence did not alter the regularity of chemoprophylaxis. AEs causing irregularity or discontinuation of chemoprophylaxis were reported, respectively, in 48 (1.2%) and 113 (2.7%) of the 4,123 soldiers who started mefloquine. The occurrence of AEs was not significantly influenced by gender and age, but their incidence significantly decreased from 2002–2006 to 2007–2011 and, among subjects taking regular prophylaxis, in those with longer duration of chemoprophylaxis and in first-time users of mefloquine (Table 4).

Table 4. Reported adverse events among 5,773 Italian army soldiers deployed in Afghanistan, 2002 to 2011, with regular, irregular, and interrupted mefloquine chemoprophylaxis
 Soldiers with adverse events to mefloquine, N (%)P valuea
Period of deployment2002 to 20062007 to 2011Total
  1. a

    Analysis by corrected Mantel–Haenszel (two-sided) chi-square test, adjusted by age, duration of deployment, gender, rank, and first deployment with prescription of mefloquine prophylaxis.

Regular prophylaxis591/2,318 (25.5)123/1,257 (9.8)714/3,575 (20.0)<0.01
Irregular prophylaxis31/72 (43.1)17/92 (18.5)48/164 (29.3)<0.01
Interrupted prophylaxis58/202 (28.7)55/182 (30.2)113/384 (29.4)0.47
Total680/2,592 (26.2)195/1,531 (12.7)875/4,123 (21.2)<0.01
Regular prophylaxis707/3,515 (20.1)7/60 (11.7)714/3,575 (20.0)0.31
Irregular prophylaxis46/161 (28.6)2/3 (66.7)48/164 (29.3)0.15
Interrupted prophylaxis106/364 (29.1)7/20 (35.0)113/384 (29.4)0.74
Total859/4,040 (21.3)16/83 (19.3)875/4,123 (21.2)0.51
Age (years)20–29 years≥30 yearsTotal 
Regular prophylaxis507/2,627 (19.3)207/948 (21.8)714/3,575 (20.0)0.20
Irregular prophylaxis33/114 (28.9)15/50 (30.0)48/164 (29.3)0.27
Interrupted prophylaxis69/261 (26.4)44/123 (35.8)113/384 (29.4)0.22
Total609/3,002 (20.3)266/1,121 (23.7)875/4,123 (21.2)0.10
Duration of deployment (days)<150≥150Total 
Regular prophylaxis551/2,255 (24.4)163/1,320 (12.3)714/3,575 (20.0)<0.01
Irregular prophylaxis19/52 (36.5)29/112 (25.9)48/164 (29.3)0.45
Interrupted prophylaxis39/130 (30.0)74/254 (29.1)113/384 (29.4)0.97
Total609/2,437 (25.0)266/1,686 (15.8)875/4,123 (21.2)0.14
First-time use of mefloquineYesNoTotal 
Regular prophylaxis681/3,247 (21.0)33/328 (10.1)714/3,575 (20.0)<0.01
Irregular prophylaxis42/141 (29.8)6/23 (26.1)48/164 (29.3)0.54
Interrupted prophylaxis108/342 (31.6)5/42 (11.9)113/384 (29.4)0.01
Total831/3,730 (22.3)44/393 (11.2)875/4,123 (21.2)<0.01

Reported symptoms among the 113 soldiers who interrupted mefloquine prophylaxis because of AEs were gastrointestinal in 81 (71.7%), neurological in 61 (54.0%), and others in 12 (10.6%).

Among the 32,500 soldiers deployed from 2002 to 2011 in Afghanistan, 63 were medically evacuated for any neuropsychiatric cause, mostly (N = 44, 70%) for alleged posttraumatic stress disorder. Mefloquine was never mentioned as a possible causal or concausal factor, and no SAEs referable to chemoprophylaxis (seizures and psychoses) were reported. Medical evacuation rate due to any neuropsychiatric cause did not change significantly after the suspension of chemoprophylaxis in Kabul since 2006: for this area the rate was 1.77 per 10,000 person-months (4/22,613) in 2002 to 2005 and 3.16 (13/41,202) in 2006 to 2011 (RR 0.56, 0.16–1.65); the evacuation rate from the Regional Command West, where mefloquine was prescribed until 2011, was 4.07 (46/113,004) in 2006 to 2011, not significantly different from that of the Kabul area (RR 1.29, 0.71–2.48) in the same period.

A suicide of an Italian soldier was reported in 2010: this subject was durably stationed in Kabul and did not take any chemoprophylaxis. This event cannot therefore be related to mefloquine use.

The Risk of Malaria

No cases of malaria occurred among Italian troops in Afghanistan during the years 2002 to 2011. In the same period, 21 imported cases of malaria, which occurred among Italian and foreign citizens coming from Afghanistan, were reported to the Istituto Superiore di Sanità: 20 cases were due to P vivax and 1 P falciparum. Of these, only one case of P vivax malaria occurred, in April 2004, 6 months after his return to Italy,[5] in an Italian soldier deployed in Afghanistan in July to October 2003. He had been mainly based in Kabul, never started chemoprophylaxis, and no other exposure to malaria-endemic countries was reported in his medical records. An incidence rate of 3.24 cases per 10,000 person-months of exposure during the transmission season (1/3,091) in 2003 can therefore be estimated.


Chemoprophylaxis Compliance and Tolerability

These results differ from previous experiences of Italian troops in malaria prevention; indeed, compliance with mefloquine prophylaxis was very high (90%–95%) in Mozambique,[6] East Timor,[7] and Sudan,[5] whereas in Afghanistan it was about 80% in 2002 to 2006 and below 60% in 2007 to 2011. However, no significant changes concerning the regularity of chemoprophylaxis and the occurrence of AEs were registered during the study period.

Concerns over the safety profile of mefloquine arose soon after the introduction of drugs, and its tolerability in malaria prophylaxis is still a controversial issue. If mefloquine is, overall, less tolerated than other regimens,[8] serious neuropsychiatric outcomes occur, however, at rates not higher than those observed with other prophylactic drugs.[9] Mefloquine use in malaria prevention has been associated with an incidence of disabling neuropsychiatric AEs ranging from 0.008% to 0.1% among users[10] or, more generally, of SAEs (requiring hospitalization) from 0.005% to 0.2%.[11] Risk factors for mefloquine-associated neuropsychiatric AEs are female gender, low body mass index, and first-time use.[12]

The absence of SAEs over 10 years among more than 20,000 soldiers, the fact that mefloquine suspension did not change significantly the evacuation rate from Afghanistan for any neuropsychiatric cause, the high proportion of subjects taking mefloquine regularly, the very low proportion of subjects taking irregularly or interrupting mefloquine because of AEs, and the fact that most AEs did not alter the regularity of chemoprophylaxis confirm the good tolerability of mefloquine for malaria prevention. These findings are consistent with previous experiences of Italian troops and, more recently, of other ISAF contingents.[13] The decline in chemoprophylaxis compliance, the relevant proportion of soldiers who forgot some doses of mefloquine, and the poor adherence to personal protection measures suggest a low perception of malaria risk. Similarly, forgetfulness and low risk perception are important causes of noncompliance with chemoprophylaxis (mainly doxycycline) among US troops in Afghanistan.[14]

This fact may represent a potential problem for future deployments in highly endemic areas, where it should be contrasted by frequently and regularly briefing troops about malaria risk, not only before entering the endemic area but also during deployment, and by a close monitoring of soldiers, to ensure high compliance with prevention measures. A similar procedure was followed with good results in past experiences.[5-7]

The Risk of Malaria in Afghanistan

Malaria is one of the main public health problems in Afghanistan. Although it was almost effectively controlled in the 1970s, the burden of disease increased gradually during the 1980s and the 1990s. Following the implementation of effective control measures after 2001,[15] an important decline of malaria was documented, and the number of reported P falciparum and P vivax cases decreased by 95 and 77%, respectively. In particular, the proportion of P falciparum cases, which represented 20% of the reported cases in the 1990s, was reduced to 1% in 2008.[4] Most malaria cases were reported in 2008 from the provinces of Helmand, Kunar, Khost, and Nangarhar. The contribution of the provinces where Italian troops were deployed since 2007 (Herat, Farah, Badghis, and Ghor) was <1% for P vivax malaria and virtually 0 for P falciparum cases.[3]

As for malaria occurrence among coalition troops, an attack rate of 52.4 cases per 1,000 soldiers was reported in a US ranger task force deployed in 2002 in eastern Afghanistan; all cases were delayed clinical presentations of P vivax infections.[16] However, this rate was referred to areas where the risk of malaria transmission is particularly high. Additional reports document malaria occurrence among ISAF troops, but represent only case reports among German,[17] British,[17, 18] and US troops,[19, 20] and are not useful to outline the risk in the different areas of Afghanistan. Because of their large contingent of troops, the relevant proportion of malaria cases diagnosed after 2002 among the US military has been exposed to endemic areas of Afghanistan.[21] Malaria risk for ISAF troops can therefore be assessed using US troops surveillance reports; overall, its incidence rate was lastly 0.1 cases per 100 soldiers,[22] well below the minimal threshold incidence of an accumulated risk of 1.13 cases per 100 at which chemoprophylaxis has been considered as a cost-effective preventive measure.[23]

Although Kabul is generally acknowledged as malaria-free, several cases of malaria in military members stationed in Kabul have been reported; apart from the only one case in the Italian troops described in this article, at least four cases occurred in 2002 to 2003 among UK troops and two cases in the German contingent.[17] Therefore, it is possible that the risk of malaria in Kabul was very low, but not absent, during the first years of ISAF deployment.

Besides mosquito bite prevention measures, chemoprophylaxis is recommended for travelers to endemic areas of Afghanistan during the transmission season; the use of chloroquine/proguanil is recommended in the UK,[24] and atovaquone/proguanil or doxycycline or mefloquine by the World Health Organization,[25] United States,[26] Canada,[27] and France.[28] Otherwise, Germany and Switzerland recommend only personal protection measures against mosquito bites and stand-by emergency treatment for countries where the risk is low and mainly represented by P vivax malaria, as the Indian subcontinent.[29] Particularly, Swiss recommendations currently include Afghanistan in this last group of countries.[30]

These considerations may be only partly applicable to soldiers whose risk of malaria is generally higher and exposure to endemic environment longer than among travelers[31]; however, following the decrease of malaria incidence in Afghanistan in the last years, it should be evaluated if malaria chemoprophylaxis among ISAF troops is still justified.

Concerning the Italian contingent, in spite of the good tolerability of mefloquine prophylaxis, the risk of rare SAEs might not be balanced by the risk of contracting malaria. As a matter of fact, although contraindications to mefloquine use may be less frequently found in the military than among the general population, they are anyway present.[32] Current screening procedures include anamnestic investigation about previous use of any prophylactic regimens, their compliance and tolerability, as well as the review of medical records and counseling before prescribing mefloquine or other regimens. However, chemoprophylaxis might anyway be prescribed to soldiers carrying contraindications to the chosen regimen, if they are not accurately screened.[33]

Malaria chemoprophylaxis was discontinued among Italian troops in Kabul in 2006, and subsequently also for all other deployment areas of Afghanistan in 2012. About this matter, a predictive modeling suggested that the risk of malaria among Canadian forces in Afghanistan was similar or even lower than the expected incidence of SAEs for some chemoprophylactic regimens, particularly mefloquine; hence the discontinuation.[34]

Another important issue is the use of primaquine terminal prophylaxis. This procedure seems to be prescribed only within several coalition contingents.[35] Terminal prophylaxis is indicated for subjects with intense and prolonged exposure to P vivax malaria, but this does not seem to be the case in most areas of Afghanistan; this procedure is probably not generally warranted for the current situation, given the low risk of P vivax malaria and the potential hazards of the extensive use of primaquine,[36] whose compliance seems even lower than the already low adherence to suppressive chemoprophylactic regimens.[16]

Primaquine treatment can trigger severe hemolytic anemia in glucose-6-phosphate (G6PD)-deficient individuals. This genetic disorder is present all over Italy; its prevalence is 0.5% to 2.9%,[37] and it is higher in the southern regions and the islands, where it can reach values of 15% or more in some areas of Sardinia.[38] The Mediterranean variant, associated with severe clinical manifestations, is the most commonly found G6PD deficiency in Italy.[39] Therefore, the use of primaquine for mass terminal prophylaxis in Italian troops may be particularly hazardous, even because most of the Armed Forces members come from the southern regions.

Limits and Strengths of the Study

This study presents some limitations. Chemoprophylaxis compliance was self-reported and therefore potentially overstated to evade any possible disciplinary action if nonadherence is admitted. However, filling the questionnaire is not mandatory and, according to completing instructions, all given information is processed for statistical purposes only. This is probably the main reason why only one fourth of deployed troops filled the questionnaire, and most of subjects who reported noncompliance did not indicate any cause; moreover, a non-negligible proportion of these subjects explained that chemoprophylaxis was not necessary. Finally, these results are consistent with those of other studies, both for withdrawal rate and occurrence of adverse effects. Therefore, we believe that most soldiers reported honestly their behavior about chemoprophylaxis.

Subjects with SAEs to mefloquine may not have been included in the study, but the analysis of all medical evacuations from Afghanistan due to any neuropsychiatric cause should intercept most of possible SAEs to mefloquine.

We found no significant differences between male and female genders concerning AEs, although mefloquine is reportedly less well tolerated in women. These results, however, may be due to the very small proportion of deployed women soldiers (2.7%), and may render these observations less generalizable, at least for women.

Some malaria cases could be missed by the surveillance. It is improbable, however, that malaria cases may go unnoticed in Afghanistan, where soldiers rely only on military health facilities. After returning to Italy, military personnel can rely also on the National Health Service, which should intercept part of the imported malaria cases possibly missed by the military surveillance; the proportion of missed cases should not therefore be relevant. Finally, although soldiers enrolled in the study represent a consistent proportion of deployed troops, they were not randomly selected, and may not therefore be fully representative of the eligible population.


We believe that these results are adequate to support our decision to discontinue suppressive chemoprophylaxis. This policy may be adopted by the generality of the coalition troops; mosquito bite prevention may be combined with any chemoprophylaxis only for soldiers deployed in areas where the risk of malaria is still substantial during the transmission season. In addition to surveillance data, analysis of some meteorological and environmental factors directly affecting malaria transmission, which can remotely be sensed by satellites, can be useful to identify these areas.[40, 41] The same considerations could be done for terminal prophylaxis; the use of primaquine should be implemented only after an adequate assessment with regard to costs and benefits of this regimen.


The study design and conduct, data collection and management, data analysis and interpretation, and manuscript preparation and review were conducted solely by the authors. The submission of this manuscript for publication was authorized by the Italian Army General Staff, but the ideas expressed in this article are those of the authors, and do not necessarily reflect the official views of the Italian Ministry of Defence or the Italian army, or the Istituto Superiore di Sanità.

Declaration of Interests

The authors state that they have no conflicts of interest.