Development and Evaluation of an Electronic Algorithm Using a Combination of a Two-step Malaria RDT, and Other Rapid Diagnostic Tools for the Management of Febrile Illness in Children Under 5 Attending Outpatient Facilities in Burkina Faso.

Background In Sub-Saharan Africa (SSA), febrile illnesses remains a major public health problem in children. The introduction of malaria rapid diagnostic tests (RDTs) in routine healthcare has greatly improved the management of malaria. However, despite the good attitude of healthcare workers to adhere to malaria RDT test results, persisting hrp2 antigen and low sensitivity of pLDH RDT negatively affect antimalarials and antibiotics prescriptions practices. These limitations of the RDTs lead to poor management of febrile diseases and antimicrobial resistance (AMR). To improve the diagnosis of these febrile diseases and subsequent prescription of antimicrobials, it is hypothesized that the implementation of an algorithm including a two-step malaria RDT PfHRP2/pLDH supported by point-of-care tests (POCTs) for C-reactive protein, white blood cells (WBC) count, urine dipstick, oximetry, and bacterial infection such as Group A Streptococcus, and Salmonella/Shigella, will significantly improve the management of febrile diseases and thereby tackling AMR.

Methodology To assess the value of the proposed algorithm, an open-label randomized controlled trial with three arms, enrolling febrile children from 2 to 59 months is proposed.

• In the control arm, febrile children will be managed according to the IMCI (integrative management of childhood illnesses) and the guideline of diagnostic and treatment (GDT), which are part of the routine system existing in Burkina Faso. Treatment will be done according to the national guideline.
• In RDTs decisional algorithm (RDT-DA) arm (intervention), the clinical examination based on IMIC will be supported by two-step malaria RDT and bacterial infections RDTs. Prescription will be left to the discretion of the healthcare workers.
• In e-algorithm arm (intervention), artificial intelligence integrating multiple layers of clinical information such as clinical examination, signs/symptoms and medical history, and laboratory information such as outcomes of biomarkers (CRP and WBC) and pathogen specific POCT (malaria and bacterial infections) and oximetry will be developed. The e-algorithm will serve to guide the diagnostic and management of febrile infections in children from 2 to 59 months.

In the 3 arms, the supports of data collection will be digitalized. A final follow-up visit (day7) will be scheduled for all participants. Patients will be asked to come back to the health facilities before scheduled visit if the symptoms persist or health condition worsening.

Study Duration This is estimated to be 36 months from the start of the study. The first 12 months will be reserved for the development of the algorithm (artificial intelligence) of e-Algorithm arm and the study CRFs for RDT-DA and control arms.