Clinical decision support algorithm to optimize management of respiratory tract infection in children attending primary health facilities in Kilimanjaro Region, Tanzania - Abstract
In low- and middle-income countries (LMICs), respiratory tract infections (RTIs) are a leading cause of
preventable death among young children (< 5 years of age). Severe RTIs, usually involving the lower
respiratory tract, constitute a potentially life-threatening medical problem that requires effective diagnosis
and management, including evaluation for antibacterials. At the same time, the vast majority of RTIs in young
children are non-severe and often caused by viruses. For these exceedingly common, non-severe viral RTI
cases, antibacterials are not appropriate and could cause harm. Yet in LMICs of Africa and Asia, research
studies have shown that antibacterials are prescribed for over 75% of outpatient pediatric RTI visits. RTI
management is thus highly problematic: on the one hand, a common syndrome that is grossly over-treated
with inappropriate antibacterials; on the other hand, clinicians in low-resourced LMIC settings can
understandably be concerned that withholding antibacterials could run the risk of a pediatric RTI progressing
to a severe, life-threatening condition. This K43 application presents a career development program to 1)
develop a clinical prediction rule that uses a parsimonious composite of clinical covariates and novel
biomarkers to accurately differentiate viral from bacterial RTI and to provide prognostic risk stratification of
disease severity in young children presenting to health facilities in Kilimanjaro Region, Tanzania; 2) conduct
formative social science research to understand caregiver and healthcare provider expectations, attitudes
and acceptability thresholds for withholding antibacterials in uncomplicated viral RTI; 3) use human-centered
design methodology to package the prediction rule and the attitudes, expectations and needs of caregivers
and healthcare providers into a user-friendly, effective clinical decision support algorithm that could be tested
in future studies for feasibility, safety, and efficacy. The candidate for this career development award is a
Tanzanian medical doctor with advanced training in clinical research, public health, and epidemiology. He
has conducted clinical research on RTI in Tanzania since 2016. For this mentored research award, the
candidate has assembled an exceptional team of mentors with expertise in clinical-epidemiologic research
of infectious diseases in Tanzania, clinical prediction analysis, human-centered intervention design in
Tanzania and other LMICs, as well as a collaborator with expertise in algorithm development for innovative
approaches to RTI management in LMICs. At the conclusion of this award, the candidate will have developed
unique expertise 1) in clinical prediction for infectious disease management in sub-Saharan Africa and 2)
in human-centered design of clinical decision support algorithms. He will emerge as a global leader in
intervention design for management of infectious diseases—a highly-skilled independent investigator
focused on implementation of strategies that will confront early childhood mortality and the growing threat
of antimicrobial resistance.
