Malaria in Sudan is a leading cause of morbidity and mortality with an annual estimate of 7.5 million cases and 35,000 deaths. This represents 50% and 70% of the WHO/Eastern Mediterranean Region cases and deaths respectively 1 . Malaria is the main cause of fever in children less than 5 years of age. Early diagnosis and appropriate treatment are essential to reduce morbidity and mortality related to malaria among this group 2 3 . This is always influenced by several factors especially in rural areas. Treatment-seeking behaviour is related to cultural beliefs about the cause and cure of illness 4 . The choice of treatment source was found to be influenced by accessibility, disease type and severity, patient's gender and parents' educational level 5 6 7 . Attitude towards providers was also an important factor 8 . Patients are more likely to start with self-treatment at home as this allows them to minimize expenditure 9 10 11 . In Sudan, two third of the patients in rural areas administer anti-malarial drugs without laboratory confirmation with preference of injectable forms 12 .

However, with the shift to artemisinin-based combination therapy (ACTs) in Sudan, the issue of treatment-seeking was considered as the cost of drugs is clearly higher compared to the cost of chloroquine. This study aimed to identify the basis on which fever was recognized and classify and explore factors involved in selection of different treatment options.

Effective management of malaria requires the consumers and the care- givers, seek, obtain, and use drugs appropriately 14 . This is linked to timely decision, accessibility, correct use of the drugs and follow-up after prescription.

Malaria in children under 5 years requires caregiver's early recognition and classification of fever. In the present study, fever and malaria were defined correctly by the majority of caregivers and malaria was identified as a main cause of fever. These findings have been shown to be the key to intervention in rural Ghana 15 . The study results reflected mother's good knowledge about malaria, its transmission and prevention; as in other parts of Africa 16 . Furthermore, they identified that high fever, inability to stand or walk, refusal to feed, loss of consciousness, yellowish sclera, severe diarrhoea and repeated vomiting were the features if malaria episode evolved into a more serious situation (severe malaria) and that requires urgent treatment at health facilities. These findings were consistent with similar studies carried out in Sri Lanka 17 and in other parts of Africa 9 18 .

Four treatment options were available for a febrile child in the area. Two included the use of drugs (consulting health workers and self treatment). The other two (traditional medicines and herbs) were deeply rooted. Accessing care from a variety of sources is a common practice in malaria endemic areas. A study in Philippine 19 showed availability of six treatment choices for families ranging from 'not doing any thing for the patient' to 'treatment with drugs based on formal prescription'. Sources of health care identified in Uganda, included public health institutions, private practitioners, traditional healers and self-treatment 20 .

Treatment-seeking behaviour was comparable between villages in the study area. Commonly people start care for a febrile child at home with what available (herbs, remaining drugs, drugs from shops, tepid sponging), when there is no response or if the condition deteriorates then they seek advice from health personnel. Medication at home before moving to health facilities was reported also in Tanzania 21 . A study in Kenya 9 showed that moving from different options determined by duration of sickness, its intensity and the expected cost. As stated by others, the delay in seeking care at health facilities level was related to existence, accessibility, satisfaction 20 and cost 15 of service, as well as satisfaction with traditional medicine and herbs. However, two factors affecting early consultation were actually leading to contradicting results: firstly, severely ill child need urgent consultation and hence short duration and secondly, appearance of illness at night deter the child from health facilities care waiting for the sun to rise and hence prolong the duration. These and other barriers were recognized by other researchers. For instance, Nuwaha F. in Uganda 20 added long waiting time, health workers abusing patients and being given tablets instead of injections as important barriers. Hill Z. et al in Ghana 15 considered financial access as a major barrier to care seeking.

Self-treatment and traditional medicine are habitual among the population of the study area. Similar finding had been reported in other parts of Sudan 12 and in Tanzania 16 . In the present study, seeking health care at health facilities is predominantly decided by mothers, this in contrast to what mentioned by Nsungwa-Sabiiti et al in Uganda 18 where mothers decide only when treatment is uncharged.

In Sudan, more emphasis for delivery of antimalarials including ACTs is given to health facilities. However, as shown by this study, in situations where coverage with health facilities is low, promoting adequate case management practices at the community level appears necessary. It has been documented 22 23 , through training of mothers and availing adequately packaged drugs, mothers could recognize malaria and as well give appropriate treatment at home and, by doing so, reduce the incidence of severe disease and thus mortality.

This study explores the requirements of a successful home management strategy 24 25 . The implementation of such strategy should take into account the ongoing treatment options in the area including private drug dispensers. A volunteer nominated from each village (called malaria control assistant -MCA) can be trained, equipped and linked with health facilities in each area. Pharmacies at health centres or hospitals can be used to store drugs for the whole area over the malaria season to ensure the quality and to distribute regularly to MCAs. Severe malaria cases identified by MCAs can then be referred to the health facilities after receiving one dose of artesunate suppository.

EMM carried out the field work, analysed the data and drafted the manuscript.

KH participated in the design of the manuscript and revised the manuscript and given the final approval to be published. All authors read and approved the final manuscript.