Conference abstract
Unclassified fever at the community level, how should it be handled? universal versus conditional three-day follow-up visit: a cluster randomized trial in Ethiopia
Pan African Medical Journal - Conference Proceedings. 2018:9(34).27
Aug 2018.
doi: 10.11604/pamj-cp.2018.9.34.761
Archived on: 27 Aug 2018
Contact the corresponding author
Keywords: Unclassified fever, children, Ethiopia, community health workers, cluster randomized trial
Oral presentation
Unclassified fever at the community level, how should it be handled? universal versus conditional three-day follow-up visit: a cluster randomized trial in Ethiopia
Tobias Alfvén1,2,&, Laura Steinhardt3, Julie Gutman3, Tjede Funk2, Ayalkibet Abebe4, Abreham Hailemariam4, Dawit Getachew4, Max Petzold5,6, Karin Källander1,7
1Department of Public Health Sciences, Karolinska Institutet, Stockholm, Sweden, 2Sach’s Children and Youth Hospital, Stockholm, Sweden Malaria Branch, US Centers for Disease Control and Prevention, Atlanta, USA, 3Malaria Consortium, Addis Ababa, Ethiopia Health Metrics, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden, 4School of Public Health, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa, 5Malaria Consortium, London, United Kingdom
&Corresponding author
Tobias Alfvén, Department of Public Health Sciences, Karolinska Institutet, Sach’s Children and Youth Hospital, Stockholm, Sweden
Introduction: with declining malaria prevalence and improved use of malaria diagnostic tests, an increasing proportion of children seen by frontline health workers in low-income settings in sub-Saharan Africa have fever of unknown origin. When seen by community health workers (CHWs) Children with non-severe unclassified fever are according to current guidelines, advised to return to CHWs after two days for re-assessment. We compared the safety of conditional reassessment only in cases where symptoms do not resolve with universal follow-up by CHWs at day three.
Methods: hypothesising that treatment failure is not more common with conditional than universal follow-up, we undertook a two-arm cluster-randomised controlled non-inferiority trial in Southwest Ethiopia. All 282 CHWs within 25 health facility clusters enrolled children aged from 2 - 59 months with fever and without malaria, pneumonia, diarrhoea or danger signs. Caregivers received advice to bring children on day three (universal arm), or to come back only if symptoms persisted (conditional arm). We conducted a per-protocol analysis using generalised linear models with a non-inferiority margin of 4% for treatment failure by day seven.
Results: from 1st December, 2015 to 30th November, 2016, 4,179 children were enrolled and had primary outcome data; the per-protocol populations included 1,953 (95.0%) in the universal follow-up arm and 1,993 (93.8%) in the conditional follow-up arm. Overall, 2.7% had treatment failure by day seven: 0.8% in the conditional follow-up arm and 4.6% in the universal follow-up arm. With a difference of -3.8% and an upper 95% CI limit of -0.65%, which is lower than the non-inferiority limit of +4%. No deaths were recorded by day 28.
Conclusion: conditional follow-up of children with non-severe unclassified fever in a low-malaria setting in Ethiopia is non-inferior to universal follow-up advice for outcomes measured through day seven, and no difference in mortality through day 28. Allowing CHWs to advise caregivers to bring children back only in case of continued symptoms might be more efficient in some settings, including settings where MSF works.
