Discussion

With a population increasingly traveling abroad (there were ∼23 million travelers from the United States in 2002 [8]) and many of these travelers visiting regions where malaria is endemic, malaria among returning travelers is likely to become increasingly frequent. In addition, despite CDC malaria chemoprophylaxis recommendations, serious malaria-associated morbidity and mortality continue to occur among travelers. In 2001, there were 11 reported deaths attributable to malaria; all involved US civilians who had recently traveled abroad to regions where malaria is endemic [4]. An earlier review of deaths attributable to malaria in the United States identified certain risk factors for fatal malaria, including failure to take the recommended antimalarial chemoprophylaxis, delay in seeking medical care, and misdiagnosis [9]. Among US civilians with cases reported to the CDC in 2001, ∼78% were not taking any chemoprophylaxis or were taking a nonrecommended drug for prophylaxis [4].

The CDC is responsible for providing malaria chemoprophylaxis recommendations for US travelers; thus, reports of travelers acquiring malaria despite taking recommended antimalarials are of particular concern and warrant further investigation. The investigation described in this paper was initiated in response to such a report. In this instance, the report was particularly alarming, because the suspected cases were among persons with reported appropriate use of atovaquone-proguanil, a relatively new chemoprophylactic drug with good efficacy [10, 11]. No previous reports suggesting such chemoprophylaxis failure have been confirmed, although there have been well-documented treatment failures [12, 13]. The ensuing investigation not only demonstrated the importance of confirmation of the diagnosis, but it also highlighted some additional potential risks associated with misdiagnosis of malaria while abroad.

When chemoprophylaxis failure is suspected, it is first necessary to confirm the malaria infection. Serological examination can be a useful tool for identifying individuals with recent malaria infection when patent parasitemia, as detected by blood film examination or PCR, is no longer detectable. The IFA test is sensitive and specific (sensitivity, 95%; specificity, 99%), and infection is inferred when an individual has elevated levels of antibodies to any one or more of the 4 species of malaria infecting humans [6]. For an individual who has experienced symptoms consistent with malaria and who has traveled recently to a region where malaria is endemic, a negative test result is a reliable way to exclude infection with Plasmodium species. It is possible that serological test results may be negative if a patient did not mount an adequate immune response because of early receipt of treatment or to immunodeficiency syndromes. The sensitivity of serological testing does decrease with time after chemotherapeutic cure [7]. However, given that all 5 participants who received a diagnosis of malaria were tested for the presence of malarial antibodies within 6 months after symptom onset and experienced symptoms for a number of days before being diagnosed and treated, we believe that the results of serologic tests confirm that they did not acquire malaria while in Ghana.

Because the blood smears from Ghana were unavailable for review, serological testing was vital to establish whether the initial diagnoses of malaria were correct among the participants who had reported a diagnosis of malaria. On the basis of the negative serological test results, concern regarding chemoprophylaxis failure seemed unfounded.

Misdiagnosis of malaria may lead to a number of inappropriate actions or behaviors, putting the traveler at additional and unnecessary risk. First, misdiagnosis may lead to prescription and administration of inappropriate treatments. Such inappropriate treatments have been reported elsewhere to have had fatal consequences [14]. There were no such adverse outcomes among the group of persons who traveled to Ghana. However, this group did report the administration of chloroquine (including injectable preparations) for the treatment of malaria. This antimalarial is no longer recommended by the CDC, given the high levels of chloroquine resistance reported in West Africa [5]. Not only are patients treated with chloroquine likely to have treatment failure and to experience serious complications of the disease, but injectable preparations of any type potentially expose the patient to a number of blood-borne diseases (such as HIV infection and hepatitis) through the use of inadequately sterilized equipment.

Misdiagnosis of malaria may also lead travelers to change or cease chemoprophylaxis on the basis of the belief that the chemoprophylaxis did not offer the expected protection. This practice puts the traveler at increased risk of truly acquiring malaria for the remainder of their travels and propagates rumors of chemoprophylaxis failure. Anecdotal reports of this phenomenon were evident during administration of the questionnaire among the participants of the Ghana trip. This extended beyond the immediate group with a diagnosis of malaria and impacted behaviors of the entire group. This phenomenon is highlighted by the incomplete adherence to treatment reported in the questionnaire: almost one-half the group was nonadherent to the prescribed antimalarial chemoprophylaxis.

The use of chemoprophylaxis and other strategies in the prevention of serious and potentially fatal malaria outcomes is critical. The results of this investigation also suggest that there is a need to improve prevention behaviors: 44% were not adherent to chemoprophylaxis, 35% did not seek medical attention for symptoms consistent with malaria, few slept under bed nets; insect repellent was used daily by most participants. Incomplete adherence was observed among all chemoprophylactic regimens, including both daily (atovaquone-proguanil and doxycycline) and weekly (mefloquine) regimens. Because numbers were small and reasons for incomplete adherence were not evaluated, it is difficult to draw conclusions regarding whether one chemoprophylactic regimen was superior to another among this group. The absence of reports of adverse reactions suggests this was unlikely to be a significant reason contributing to incomplete adherence.

Alternative diagnostic methods exist. Rapid malaria antigen tests, also known as “rapid diagnostic tests” (RDTs), are commercially available and are capable of identifying malaria parasitemia. These tests are relatively simple to perform, and the results are relatively simple to interpret, and they may be appropriate for diagnosis of malaria in areas where microscopy is unavailable or unfeasible [15]. However, several factors limit the ability to recommend the use of RDTs by US travelers. First, reports of RDT use by travelers for self-diagnosis is limited. Second, there are no RDTs currently approved for use in the United States. Third, there are data suggesting that these tests do not have adequate sensitivity at the low parasite densities that can cause disease in nonimmune travelers to avoid potentially false-negative results. Fourth, there are concerns regarding the ability of ill travelers to perform the test and to manage their illness appropriately [16]. Therefore, use of RDTs by US travelers to areas of endemicity is not recommended at this time.

In summary, there was no evidence to suggest chemoprophylaxis failure with atovaquone-proguanil or any other chemoprophylactic drug used among this group. As a consequence of misdiagnosis and perceived failure of chemoprophylaxis, several participants received unnecessary treatment, and a number of participants modified their adherence and even ceased taking their chemoprophylaxis altogether. We believe that this investigation highlights the need to provide adequate pretravel information and for travelers to understand the importance of adherence to chemoprophylaxis. If any illness occurs, prompt medical attention should be sought. Of note, although fever and chills are classic symptoms of malaria, gastrointestinal and respiratory symptoms may also be attributable to malaria. Investigation of any illness should be encouraged and should include rapid, accurate diagnosis, followed by safe, effective treatment. It would be preferable for programs, such as the one conducted by this university, that send groups abroad to identify reputable local health care facilities where accurate diagnosis and effective treatment can be sought. Another suggestion is to inform travelers to retain any blood smears prepared overseas for review upon return to the United States, which would be helpful in determining whether malaria was the cause of the illness. It is also essential for public health agencies to critically evaluate reports of chemoprophylaxis failure and to determine whether these could potentially represent drug resistance. CDC personnel and laboratory resources are available to assist in such investigations.