Avoiding misdiagnosis of malaria: a novel automated method allows specific diagnosis, even in the absence of clinical suspicion.

To the Editor: We report three cases of malaria to illustrate a novel method that allows diagnosing the disease, even if clinicians do not suspect it or request malaria smears. Lack of clinical suspicion is a well-known factor for malaria misdiagnosis and may be responsible for almost 40% of deaths from Plasmodium falciparum infections in industrialized countries (1-3). A recent study from Canada showed that in 59% of cases malaria was initially misdiagnosed, and in 16% three or more physician contacts occurred before malaria smears were ordered (4).

Early diagnosis of malaria relies crucially on clinical suspicion. A clinician suspecting the disease has to explicitly request malaria smears. This problem has not been solved with the advent of several methods alternative to microscopy, including recently introduced rapid dipstick tests (5). Performing any of these tests blindly without a specific request is impractical. On the other hand, routinely performed laboratory tests in the work-up of febrile patients, e.g., automated full blood counts, have so far detected only nonspecific changes, such as anemia or thrombocytopenia, which are associated with many other conditions (6). These changes on their own are therefore not specific enough to trigger malaria smears without an explicit request.

New automated full blood counts-analyzers incorporate flow-cytometric principles. The Cell-Dyn 3500 (Abbott, Santa Clara, CA) uses scattered laser-light of leukocytes at four different angles to generate a white-blood-cell differential (7). Monocytes and neutrophils may ingest birefringent depolarizing malaria pigment that can be detected by the instrument. The appearance of monocytes (purple dots) above the separation line, in the eosinophil area (green dots), is a highly specific sign of the presence of ingested malaria pigment and consequently malaria.

A study from South Africa investigating 224 directed samples for malaria diagnosis found a sensitivity of 72% and specificity of 96% (8). In Portugal, we observed 45 positives in 120 directed samples. So far, all cases identified by microscopy showed the typical changes in the full-blood-count plots, suggesting a near 100% sensitivity in imported malaria cases. Several thousand full-blood-count plots from patients with a wide range of underlying pathologic features did not show such changes, making them highly specific for malaria diagnosis. However, the changes may persist for some time despite clinical and parasitologic cure, as pigment-containing monocytes may remain in the circulation for 2 to 3 weeks (9). Consequently, the observed changes may not necessarily indicate acute disease but may persist during convalescence.

We report three cases in which clinical suspicion did not lead to the request of a malaria diagnostic test. The final diagnosis of malaria was made only because of the changes observed in the color monitor of the Cell-Dyn 3500. As part of a preliminary investigation of this new method, we reviewed all full-blood-count plots at 24-hour intervals. During a 2-week period, three full-blood-count granularity/lobularity plots compatible with malaria were identified and the full-blood-count results and clinical notes were reviewed. The principal symptoms in the three cases were fever and aches in bones and muscles, case 1; complications of assault, case 2; and feeling generally unwell (from drug abuse), case 3. In all cases, the full-blood-count results were within normal ranges, except for a thrombocyte count of 23,000 in case 2. In cases 1 and 3, the patients were discharged with a clinical diagnosis of flulike syndrome and drug abuse-related problems, respectively, while in case 2, the patient was to be admitted with a diagnosis of assault-related injuries. As attending clinicians had not requested malaria smears, we performed blood films on the recovered specimens that confirmed a diagnosis of malaria. (Case 1: P. ovale, 10,000 [micro]l; case 2: P. falciparum, 9,000 [micro]l; case 3: P. falciparum, 1,500 [micro]l). In case 2, our findings permitted appropriate treatment in the emergency room; in the other two cases, it allowed patients to be contacted at home. All three patients (two male, one female) were of Black African origin but lived in Portugal. They had returned to Portugal after visiting Africa (Angola and Guinea). None of them had taken malaria prophylaxis during their journey.

Anisotropic malaria pigment has been the basis for several microscopy methods for malaria diagnosis (10). However, sensitivities are similar to that of conventional microscopy, and these methods have to be ordered specifically. In contrast, automated full-blood-count is regarded as routine for febrile patients, and the new automated method has the potential to detect additional, unsuspected cases, in which clinical suspicion did not lead to requests for malaria testing. If further studies validate this technique, the instrument could be modified to specifically flag such results, which would alert laboratory staff to perform blood films on these samples, even in the absence of a clinician's request. Finally, if software algorithms are adjusted to enumerate pigment-containing leukocytes, the usefulness of this indicator as a prognostic marker (11) could be further evaluated. The instrument may greatly facilitate quantification of pigment-containing leukocytes, which have been determined by time-honored but cumbersome microscopy.

References

(1.) World Health Organization. World malaria situation in 1994. Wkly Epidemiol Rec 1997;36:269-274.

(2.) Day KP. Malaria: a global threat. In: Krause RM, editor. Emerging Infections. New York: Academic Press; 1998. p. 463-97.

(3.) Greenberg AE, Lobel HO. Mortality from Plasmodium falciparum malaria in travelers from the United States (1959-1987). Ann Intern Med 1990;113:326-7.

(4.) Kain KC, Harrington MA, Tennyson S, Keystone JS. Imported malaria: prospective analysis of problems in diagnosis and management. Clin Infect Dis 1998;27:142-9.

(5.) Hanscheid T. Diagnosis of malaria: a review of alternatives to conventional microscopy. Clin Lab Haematol 1999;21:235-45.

(6.) Giacomini T, Lusina D, Foubard S, Baledent F, Guibert F, Le Pennec MP. Dangers of hematological automated analysis for malaria diagnosis. Bull Soc Pathol Exot 1991;84:330-3.

(7.) de Grooth BG, Terstappen LW, Puppels GJ, Greve J. Light-scattering polarization measurements as a new parameter in flow cytometry. Cytometry 1987;8:539-44.

(8.) Mendelow BV, Lyons C, Nhlangothi P, Tana M, Munster M, Wypkema E, et al. Automated malaria detection by depolarization of laser light. Br J Haematol 1999;104:499-503.

(9.) Day NPJ, Thi Diep P, Thi Ly P, Xuan Singh D, Phu Loc P, Van Chuong L, et al. Clearance kinetics of parasite and pigment-containing leukocytes in severe malaria. Blood 1996;88:4694-700.

(10.) Lawrence C. Laveran remembered: malaria haemazoin in leucocytes. Lancet 1999;353:1852.

(11.) Phu NH, Day N, Thi Diep P, Ferguson DJP, White NJ. Intraleucocytic malaria pigment and prognosis in severe malaria. Trans R Soc Trop Med Hyg 1995;89:200-4.

Thomas Hanscheid,(*) Bernadino G. Pinto,(*) Isabel Pereira,(*) Jose Melo Cristino,(*) and Emilia Valadas([dagger])

(*) University Hospital Santa Maria, Lisbon, Portugal; and ([dagger]) London School of Hygiene and Tropical Medicine, London, United Kingdom