The data contains the complete set of data from the enrolment and follow-up at day 7, 14 and 28 of 4179 children included in the cluster randomised controlled study in Ethiopia. It has the child's symptoms from the re-assessment as well as the caregiver characteristics, care seeking behaviour and socio-economic status. It also contains all the translated transcripts from the qualitative interviews with the caregivers (mothers and fathers) and health extension workers. Under World Health Organization's (WHO) integrated Community Case Management (iCCM) strategy, febrile children seen by community health workers (CHW) without a diagnosable illness and without danger signs are advised to return after two days, regardless of symptom resolution. This advice might be unnecessary and place additional time and cost burdens on caregivers and CHWs. However, the safety of not following up children with unclassified fever is unknown. Through a two-arm cluster randomized controlled non-inferiority trial, the safety of conditional follow-up of non-severe unclassified fever, i.e. non-severe illness with fever, no malaria, pneumonia, diarrhea, or danger signs, compared to universal follow-up after two days will be established. The study is conducted in three districts in Southwest Ethiopia. Twenty-five health facilities were randomized to one of two intervention arms; all 144 health posts and the 282 CHWs are included. All enrolled children are followed up on day seven for re-assessment. If still sick on day seven, additional follow-up takes place on days 14 and 28. To demonstrate that there is no significant increase in the percentage of children deteriorating clinically, the sample size needed for a non-inferiority margin of 4%, a power of 80%, an alpha of 5%, and a design effect of 3, is 4284 children with unclassified fever. Main outcome is treatment failure on day seven, defined as death, hospitalization, one or more danger signs, or persistent fever. This study addresses the question whether there is any benefit to recommending universal follow-up among children seen for non-severe unclassified fever, or whether parents can be counselled to return in the event the fever persists, using a cluster randomized controlled trial design embedded in a national program. Outcomes will be relevant for policy makers and are important for the evaluation of current and future WHO guidelines for management of children with fever.

This was a two-arm community cluster randomised non-inferiority trial carried out in 3 woredas (districts) in Southern Nations, Nationalities, and Peoples' Region (SNNPR) in Ethiopia. Clusters, defined by health centre (the referral centre and practical training institution for HEWs where their services are coordinated), were randomised into either the conditional or universal follow-up arm. For more information, see file Methodology_ProjectInformation.