Fevers in childhood are common and usually self-resolve. In sub-Saharan Africa, when a febrile child presents to a community health worker (CHW), the child is assessed for malaria, pneumonia, and diarrhea, and other danger signs, according to WHO guidelines for integrated Community Case Management (iCCM) of childhood illnesses. In the cases where 1) there are no danger signs present, and 2) malaria, pneumonia, and diarrhea have been ruled out, the World Health Organization (WHO) recommends that all children be reassessed in 3 days. It is hypothesized that health outcomes for these cases will be equivalent if the CHW advises to come back in 3 days only if symptoms have not resolved. In order to assess this hypothesis, a two-arm cluster-randomized, community-based non-inferiority trial in Southern Nations, Nationalities and People's Regional State (SNNPR) in Southwest Ethiopia will be conducted to assess the non-inferiority of CHW-advised systematic follow-up on day 3 compared to conditional follow-up for non-severe febrile illness in children age 2 to 59 months, in which no cause of fever can be identified and where danger signs are absent.

Globally, mortality in children under five years (U5MR) stands at 45.6/1000 live births and it is estimated that 6.3 million children under five (U5) die each year. Despite a seemingly impressive global decline in U5MR, Sub-Saharan Africa is the only region that has registered increasing under-five mortality in several of its countries. Most deaths are caused by pneumonia (15%), diarrhea (9%), and malaria (7%); diseases with symptoms that overlap, making differential diagnosis difficult. In response, many countries in sub-Saharan Africa have scaled up integrated community case management (iCCM) targeting malaria, pneumonia and diarrhea in children U5. The programs typically train community health workers (CHWs) to assess, classify and treat uncomplicated cases of pneumonia, diarrhea and malaria and refer children with danger signs and malnutrition. While mortality impact of iCCM has been difficult to demonstrate, there is clear evidence that it can increase the treatment rate among sick children.

In Ethiopia, U5MR stands at 68/1000 live births and it is estimated that 205,000 children under the age of five die each year. iCCM implementation started in February 2011 as part of Ethiopia's Health Extension Program (HEP), and over 38,000 female CHWs - locally referred to as Health Extension Workers (HEWs) - have been trained and deployed to communities throughout the country to provide preventive and curative health services. There are typically two HEWs assigned to a sub-district with a population of approximately 5,000. The HEWs are supervised by health centers that oversee 5 health posts each (with 2 HEWs in each). As of 2014, 29,900 of the HEWs had been trained in iCCM using the WHO guidelines.

As per the WHO ICCM guidelines, children treated for an illness and those with unclassified fever and without danger signs (for whom treatment should be withheld), should be counseled to have a follow-up visit after three days to assess treatment compliance and illness resolution. This is either done through a return visit to the health post or through a home visit by the CHW through a systematic follow-up visit. However, febrile illness is common in childhood, and is often due to viruses or other self-resolving illnesses. In a large proportion of cases fever resolves rapidly, generally within 48 hours, and almost always within 96 hours. It is suggested from a number of studies that it is safe to withhold medical treatment for children with unclassified fever. In Ethiopia, HEWs follow the integrated management of neonatal and childhood illness (IMNCI) manual which stipulates that caregivers of children seen by HEWs should only return when the child fails to respond to treatment, by way of conditional follow-up to the health post, instead of the WHO recommended iCCM guideline. There is limited evidence on which of the two recommendations is safer for the child and it is unclear whether caregivers and HEWs would comply better with the systematic follow-up advice compared to the conditional follow-up advice, and whether the systematic follow-up visit is even necessary.

The purpose of this research is therefore to assess the non-inferiority of HEW-advised systematic follow-up on day 3 compared to conditional follow-up for non-severe febrile illness in children age 2 to 59 months, in which no cause of fever can be identified and where danger signs are absent.

Methodology

Study site:

The research study (TRAction) will be conducted in three Woredas (districts); Halaba special Woreda and Boloso Sore and Damot Gale (Wolayita zone) in the in the Southern Nations, Nationalities and People's Regional State (SNNPR) in Southwest Ethiopia. The three Woredas were selected based on a) strength of iCCM program (i.e. consistency in HEW supervision and supply), b) HEW use rate among caregivers, and c) concurrent community mobilization activities under other grants.

The SNNPR has an estimated total population of 18.4 million, which makes it the third most populous region in Ethiopia. The estimated population density is 141 people/km2 but there are great variations across the region. The health indicators in the Region are among the lowest in the country and according to the Ethiopia Demographic and Health survey (EDHS) 2011, neonatal, infant and U5MR are 38, 78 and 116 per 1,000 live births, respectively.

The iCCM program is functioning in all districts of SNNPR, It is therefore an ideal environment to implement the research, as iCCM services are stable and the program is implemented by the Regional Health Bureau which provides technical advice to this project. There are 18 Health Centers and 102 Health Posts with 204 HEWs in the three selected Woredas. According to latest Malaria Indicator Surveys, awareness of risks associated with febrile illnesses in children and the benefits of early care seeking is low, with only 46.3% of children U5 with fever taken for early treatment. However, care seeking at health post level shows an upward trend, presumably as a result of the increased awareness of the availability and proximity of child health services.

Study design:

The study will be conducted as a randomized, two-arm trial of either conditional (control) or systematic (intervention) follow-up advice for non-severe febrile children who after assessment by the HEWs, are not classified as having malaria, pneumonia, diarrhea, or other symptoms requiring referral. In brief, caregivers of children who meet the above inclusion criteria will be counseled to follow one of two pathways (based on randomization at the health facility level). The control arm consists of a conditional follow-up visit, i.e. caregivers are counseled to return with their child to the HEW for a re-assessment only in cases when the child deteriorates or fever persists for more than 2 days. In contrast, caregivers in the intervention arm will be advised to have a systematic follow-up visit, i.e. caregivers of all children are counseled to return after three days, regardless of symptom resolution.

All enrolled children will have a Day 7 study visit in their home with an independent evaluator to assess their clinical outcome. Management of illness at any subsequent visit (return to HEW on any day, return to HEW for "intervention" Day 3 visit, or Day 7 assessment) will be managed following established iCCM guidelines. The details of the study procedures are described below in the research work plan.

Sample Size:

The primary outcome on which sample size will be based is the proportion of children with persistent fever, persistent illness, or decline (hospital, danger signs develop, death) at Day 7. We assume that the baseline rate of fever/illness persistence is low (~5%, based on rates of ~3% and 10% in previous studies) and we aim to demonstrate that there is not a significant increase in this risk. The table below shows sample sizes (total eligible patients) needed for a non-inferiority margin of 0.03 - 0.05, assuming a power of 80% and an alpha of 0.05. A design effect of 3 will be used to account for clustering at CHW and health facility levels. The shaded row indicates the most plausible as well as logistically feasible sample size estimation. Enrollment will occur over a one-year period to account for seasonality of various causes of febrile illness.

A minimum of 4284 children 2-59 months with fever but no malaria, pneumonia or diarrhea are required for this study. Based on the assumption that 30% of children with fever would be eligible for the study (no malaria, diarrhea, pneumonia or referral signs), a total of 16,333 children 2-59 months will need to be screened by the HEWs. With 12 months of data collection, this means that on average 1,361 children will need to be screened each month. Each health post see on average 16 children 2-59 months each month; hence 85 health posts (each with 2 HEWs) will need to be included in the study. The unit of randomization (cluster) will be the health centers that oversee 5 health posts each (with 10 HEWs). Therefore, a total of 18 health centers (clusters) will need to be included in the study, half of which will be randomly allocated to the intervention arm and half to the control, using restricted randomization to minimize the difference between the intervention and control clusters. Restricted randomization will be based on key indicators such as average cluster distance to nearest zonal referral hospital and number of unclassified fever in children under 5 seen by HEWs.

Data collection and management:

The study will be conducted as a randomized, two-arm trial of either conditional (control) or systematic (intervention) follow-up assessment visit for non-severe febrile children who after assessment by the HEWs, are not classified as having malaria, pneumonia, diarrhea, or other symptoms requiring referral. Caregivers of children who meet the above inclusion criteria will be counseled to follow one of two pathways (based on randomization at the health facility level). The intervention arm consists of a systematic follow-up visit on Day 3 -caregivers are counseled to return with their child systematically at Day 3 to the HEW for a follow-up assessment. In contrast, the control arm (current practice) consists of a conditional follow-up visit-caregivers are counseled to return at any point to the HEW if symptoms persist or worsen.

All enrolled children will have a Day 7 study visit in their home with an independent evaluator to assess their clinical outcome. Children who fail treatment on day 7 will be re-assessed on day 14, and those who still fail treatment on day 14 will be followed up for re-assessment on Day 28. In addition, all children enrolled will be followed-up via a phone call for vital status on Day 28. Management of illness at any subsequent visit (return to HEW on any day "control" visit, return to HEW for "intervention" Day 3 visit, or Day 7 assessment) will be managed following established iCCM guidelines and children still febrile will be referred to a health facility.

Inclusion Criteria:

• Non-severe febrile illness that is classified by HEW as unspecific fever

Exclusion Criteria:

• Malaria, pneumonia, diarrhoea diagnosis
• Symptoms requiring referral