Background: The epidemiology of fevers and their management in areas of low malaria transmission in Africa is
not well understood. The characteristics of fever, its treatment and association with infection prevalence from a
national household sample survey in the northern states of Sudan, an area that represents historically low parasite
prevalence, are examined in this study.
Methods: In October-November 2009, a cluster sample cross-sectional household malaria indicator survey was
undertaken in the 15 northern states of the Sudan. Data on household assets and individual level information on
age, sex, whether the individual had a fever in the last 14 days and on the day of survey, actions taken to treat the
fever including diagnostic services and drugs used and their sources were collected. Consenting household
members were asked to provide a finger-prick blood sample and examined for malaria parasitaemia using a rapid
diagnostic test (RDT). All proportions and odds ratios were weighted and adjusted for clustering.
Results: Of 26,471 respondents 19% (n = 5,299) reported a history of fever within the last two weeks prior to the
survey and 8% had fever on the day of the survey. Only 39% (n = 2,035) of individuals with fever in last two weeks
took any action, of which 43% (n = 875) were treated with anti-malarials. About 44% (n = 382) of malaria
treatments were done using the nationally recommended first-line therapy artesunate+sulphadoxinepryrimethamine
(AS+SP) and 13% (n = 122) with non-recommended chloroquine or SP. Importantly 33.9% (n =
296) of all malaria treatments included artemether monotherapy, which is internationally banned for the treatment
of uncomplicated malaria. About 53% of fevers had some form of parasitological diagnosis before treatment. On
the day of survey, 21,988 individuals provided a finger-prick blood sample and only 1.8% were found positive for
Plasmodium falciparum. Infection prevalence was higher among individuals who had fever in the last two weeks
(OR = 3.4; 95%CI = 2.6 - 4.4, p < 0.001) or reported fever on the day of survey (OR = 6.2; 95%CI = 4.4 - 8.7, p <
0.001) compared to those without a history of fever.
Conclusion: Across the northern states of the Sudan, the period prevalence of fever is low. The proportion of
fevers that are likely to be malaria is very low. Consequently, parasitological diagnosis of all fevers before treatment
is an appropriate strategy for malaria case-management. Improved regulation and supervision of health workers is
required to increase the use of diagnostics and remove the practice of prescribing artemisinin monotherapy.