Usefulness of serial testing for the diagnosis of malaria in cases of fever upon return from travel.

Abstract

When malaria is suspected in case of fever after travel in endemic areas, the current recommendation is to repeat the malaria test at 24-hour intervals, with up to two additional tests, as long as the test result is negative. A retrospective analysis was conducted to investigate the appropriateness of this recommendation by determining the proportion of tests with negative result at first and subsequently with a positive one at second or third attempt. A retrospective study was conducted at the Centre for Primary Care and Public Health, Lausanne, covering a period of 15years. All patients tested once for malaria were included. Testing included microscopy thick and thin films as well as malaria rapid diagnostic test used in combination. The main outcome measure was the proportion of patients with a first negative test result, subsequently positive on second or third test over the total patients with suspected malaria assessed. Demographic, travel, clinical, and laboratory variables were collected from patients' records to identify potential predictors of an initially negative and then positive test result. Four thousand nine hundred seventy-two patients were included. Of those, 4557 (91.7%) had definitive negative test results, and 415 (8.3%) had a positive result on the first test [332/415 (80%) Plasmodium falciparum, 40/415 (9.6%) P. vivax, 21/415 (5.1%) P. ovale, 12/415 (2.9%) P. vivax/ovale, 9/415 (2.2%) P. malariae and 1/415 (0.2%) P. knowlesi], and 3/4972 (0.06%) had a positive result on the second test after a first negative result, 1/4972(0.02%) had a positive test result after 2 negative results, all with P. falciparum. One of the four patients that were positive after their initial negative test was pregnant. The very small number of patients with an initially negative test result and secondarily positive did not allow for risk factor analysis. The current recommendation of serial malaria testing is not supported by the present study, a fortiori for those who do not present with a strong clinical or laboratory predictor of malaria.