The engineering challenge behind antibiotic resistance

01 Medicine is a constrained control system

Think of an antibiotic program as a feedback system. Clinicians observe symptoms and tests, choose a drug, and then observe whether the patient improves. But the system is noisy: symptoms overlap, diagnostics take time, drug concentrations vary by tissue, and the pathogen population can change during treatment.

Resistance is a control failure when the intervention removes susceptible cells without suppressing the resistant subpopulation—or when the intervention arrives too late. Better engineering does not mean treating every infection with maximum force. It means getting the right signal, target, dose and timing with the least collateral pressure.

02 The drug must reach the right place

Pharmacology is an engineering problem before resistance is even considered. A drug must survive the body, reach the infected tissue, cross bacterial barriers and remain above an effective concentration long enough to matter. Kidney function, age, pregnancy, perfusion and drug interactions change that journey.

At the infection site, bacteria may live in oxygen-poor pockets or biofilms where exposure is uneven. A laboratory susceptibility result is valuable, but it is not a guarantee that the same concentration will be achieved in a patient’s lung, bone, urine or bloodstream.

03 Diagnostics are the missing sensor

When clinicians lack a rapid pathogen identification or susceptibility result, they must act under uncertainty. Broad empiric treatment may be lifesaving, yet it exposes more organisms and can delay a precise switch. A slow test can be accurate and still arrive after the critical decision.

Rapid molecular tests, culture automation, local antibiograms and better sampling can shorten the loop. The design goal is not simply “more technology”: it is a result that changes management soon enough, is interpretable, and works where the patient is being treated.

04 Every layer has a different threat model

At the cell level, resistance mechanisms include target modification, drug destruction, reduced permeability and efflux. At the ward level, the threat is transmission through people and surfaces. At the hospital level, it is prescribing, stock-outs, laboratory capacity and staffing. At the national level, it includes surveillance, procurement and reimbursement.

An intervention that fixes one layer can fail at another. A new antibiotic is not enough if it is unavailable, misdiagnosed, unaffordable or used without infection control. Engineering means specifying the boundary and testing the whole system.

05 Innovation must survive the market

Antibiotics are unusual products: they are most valuable when preserved, often taken for a short course, and ideally replaced by prevention. That makes conventional volume-based sales a poor match for public health. A technically successful molecule can still lack a sustainable business model.

Useful portfolios therefore include diagnostics, vaccines, new antibiotics, anti-virulence approaches, phage research and infection-prevention tools. Financial “push” support lowers development risk; “pull” incentives reward availability and readiness without requiring unnecessary volume.

06 Waste and the environment are part of the plant

Drug residues and resistant organisms can leave hospitals, farms and factories through wastewater. The environment is not a passive drain: it is a mixing space where genes, hosts and selection pressures meet. Treatment plants can reduce risk, but performance and monitoring vary.

A One Health design therefore measures what enters and leaves the system, not just prescriptions. It asks where resistance genes are concentrated, which organisms carry them, and whether controls reduce transmission downstream.

07 Build for graceful failure

No diagnostic is perfect, no drug reaches every compartment, and no surveillance network sees every case. A resilient system assumes failures: it keeps reserve treatments, infection-control procedures, local data, expert review and transparent escalation paths.

WHO describes resistant infections as a growing pressure on costs, hospital stays and second-line therapy. The engineering response is a portfolio: make each component better, then make the interfaces between components reliable.