Three Days to Name the Bug

In MSF field hospitals in conflict and displacement settings (South Sudan, DRC, Yemen), clinicians treating serious infections must prescribe antibiotics empirically because antimicrobial susceptibility testing (AST) results take 48–72 hours. MSF's Mini-Lab and Antibiogo app have expanded bacteriology capacity to settings that previously had none, but neither achieves same-visit turnaround. During the 2–3 day wait, patients receive broad-spectrum antibiotics — driving the very resistance the diagnostic is designed to combat. In settings with high MDRO prevalence, empirical therapy failure rates exceed 30%.

AMR kills ~1.27 million people annually, disproportionately in LMICs and conflict settings where narrow-spectrum alternatives are unavailable. Humanitarian field hospitals see concentrated wound infections and post-surgical complications with high resistance rates. Faster AST would enable targeted prescribing that preserves antibiotic efficacy. The MSF Mini-Lab has demonstrated that decentralized bacteriology is operationally feasible — the remaining barrier is speed.

MSF's Mini-Lab (deployed at 8+ sites since 2022) is a transportable clinical bacteriology lab operated by non-specialist technicians after one month of training. It uses disk diffusion and broth microdilution (MicroScan panels) — methods that inherently require overnight culture (18–24h growth + 18–24h AST). MSF's Antibiogo smartphone app (CE-marked, 9+ countries) semi-automates disk diffusion reading but doesn't reduce culture time. Molecular rapid AST (BioFire, GenMark) detects resistance genes in hours but costs $50–150/test, requires cold-chain cartridges, and detects genetic markers rather than phenotypic resistance — missing novel resistance mechanisms. No platform achieves same-visit phenotypic AST in field conditions (limited electricity, >40°C, dust, no compressed gases).

Phenotypic AST methods bypassing overnight culture — single-cell morphological analysis detecting growth/no-growth at the individual bacterium level within 2–4 hours, or microfluidic devices confining bacteria in nanoliter volumes to accelerate observable growth responses. Target: device <$5,000, per-test <$10, tolerant of 45°C, no compressed gases, battery-operable, clinically actionable results within 4–6 hours.

A team could prototype a microscopy-based rapid AST system that images individual bacteria in the presence of antibiotics and uses time-lapse analysis to detect growth/no-growth within hours rather than overnight. Alternatively, a team could design a microfluidic chip that concentrates bacteria from a clinical sample and exposes them to an antibiotic panel in nanoliter chambers. Biomedical engineering, microbiology, and optics skills apply.