Wastewater, early warning and the One Health governance gap.
Somewhere beneath a city, millions of people are continuously contributing to an enormous biological sample.
Every shower, toilet and drain feeds into a wastewater system carrying traces of what is happening within the population above it. Viruses can be shed by infected people before they seek medical attention; bacteria and antimicrobial-resistance genes can enter wastewater from homes, hospitals and other sources; and chemical and biological markers can persist long enough to provide information about the health of a population without identifying any individual within it.
During the COVID-19 pandemic, wastewater surveillance moved from a relatively specialist public-health activity into public consciousness. Testing sewage for SARS-CoV-2 demonstrated that changes in infection across a community could sometimes be detected without waiting for thousands of individuals to book tests, visit doctors or arrive at hospitals. Yet COVID was not the beginning of wastewater surveillance, and it is increasingly clear that it will not be the end.
Poliovirus, influenza, antimicrobial resistance and a growing range of other biological threats can potentially be monitored through wastewater and environmental surveillance. The World Health Organization now treats it as a component of wider multi-pathogen surveillance, while the European Union is incorporating wastewater monitoring more explicitly into its public-health architecture.
This creates an extraordinary new source of biological intelligence. It also creates a governance problem — because detecting something in a sewer and knowing what to do about it are very different things.
The environment becomes a sensor
Most infectious-disease surveillance has traditionally depended, directly or indirectly, upon people. Someone becomes infected; they develop symptoms; they decide whether those symptoms warrant medical attention; a clinician decides whether testing is appropriate; a sample reaches a laboratory; and a result is generated and, depending upon the disease, eventually enters a surveillance system. Every stage introduces a potential delay.
There is also an enormous amount that conventional surveillance never sees. People with mild disease may never contact healthcare services; asymptomatic infections may remain completely invisible; and testing policies, healthcare-seeking behaviour and the reach of surveillance systems all vary between populations.
Wastewater approaches the problem from the opposite direction. Instead of asking people to enter the surveillance system, it samples a biological environment into which large numbers of people are already shedding material. The surveillance pathway becomes something closer to infection → shedding → wastewater → sampling → detection, rather than infection → illness → healthcare → testing → reporting → surveillance. The distinction is potentially transformative: the environment itself can become part of the sensor network.
London and the polio nobody had seen
One of the clearest demonstrations came not from COVID, but from polio. The United Kingdom routinely conducts environmental surveillance for poliovirus because sewage provides a means of detecting circulation that may not otherwise become apparent.
In February 2022, vaccine-like type 2 poliovirus was detected in sewage collected from the London Beckton sewage treatment works — the largest such works in Europe, serving a population of almost four million. It did not disappear. Over the following months, further genetically related isolates were repeatedly detected: between February and July, investigators identified 118 genetically linked poliovirus isolates in 21 of 52 sequential sewage samples, with genetic changes indicating that the virus had continued to circulate. Yet there was no corresponding cluster of paralytic polio cases announcing its presence.
The sewer had seen something that conventional clinical surveillance had not. The discovery prompted an enhanced public-health response, including expanded environmental surveillance and a vaccination campaign for children in London. It is difficult to imagine a clearer demonstration of the value of environmental surveillance — and it also raises a much more interesting question: what happens after the sewer knows?
Detection is only the beginning
A laboratory result is not a public-health decision. An unusual pathogen, or an increase in its concentration, has to be interpreted. How unusual is the signal, and is it increasing? How large is the population represented by the sample? Could rainfall or changes in wastewater flow have affected the measurement? Does the organism indicate active transmission, and is the signal localised or geographically dispersed? Does it correspond with clinical surveillance? Is additional sampling required, and should hospitals or laboratories change their testing? Does veterinary surveillance contain anything relevant? Does somebody need to intervene?
The ability to detect biological material therefore creates another requirement: the ability to decide what the detection means. This distinction can be expressed very simply — detectability is not the same as actionability — and the more sensitive environmental surveillance becomes, the more important it will be. If we develop the technical capability to detect hundreds of pathogens, genetic markers and resistance genes, we will inevitably encounter signals whose significance is uncertain. A surveillance system that generates more signals than institutions can interpret does not necessarily make society safer; it can simply create more noise.
Three different gaps
Wastewater surveillance therefore exposes at least three different gaps within biological surveillance.
The first is the detection gap — the interval between a biological event occurring and the surveillance system recognising it. Environmental surveillance can shorten this dramatically, because it does not necessarily depend upon illness, healthcare attendance or individual diagnostic testing.
The second is the interpretation gap. Once something has been detected, somebody must determine whether it represents background activity, an emerging outbreak, an imported infection, sustained community transmission or an analytical anomaly. This is partly a scientific problem, but it is also a systems problem, because interpretation often requires information from elsewhere.
Then comes the third: the governance gap. Once an environmental signal is sufficiently credible to matter, the information must reach the organisations capable of doing something about it — and that transition is surprisingly important. A sample can cross the laboratory threshold in minutes; information can take considerably longer to cross an institutional boundary.
Who owns a signal from a sewer?
Imagine that routine wastewater surveillance detects something unusual on Monday morning. Who owns that information? The laboratory that generated the result? The organisation operating the wastewater network? The environmental regulator? The local public-health authority, the national public-health agency, or a hospital microbiology network? If the signal involves antimicrobial resistance, should veterinary authorities also be informed? If an organism could have entered the system through food production, agriculture or industrial discharge, when do those sectors become involved?
These questions are not arguments against wastewater surveillance. They are arguments for designing the governance around it at the same time as we design the science.
Governance latency
This brings wastewater surveillance directly into a wider One Health Security problem. We describe governance latency as the interval between relevant evidence existing somewhere within a system and institutions connecting that evidence sufficiently to support action. In environmental surveillance, that clock might look something like environmental signal → confirmation → interpretation → escalation → cross-system recognition → decision → intervention, and the laboratory turnaround time is only one component.
Suppose an unusual resistance gene appears repeatedly in wastewater. The laboratory detects it quickly, but it takes several days to determine who should receive the information, whether clinical laboratories are seeing the same pattern, and whether animal-health or environmental data might provide another part of the picture. The delay has not occurred because the technology failed — the technology worked. The delay occurred in the architecture surrounding it. That is governance latency.
Antimicrobial resistance makes the problem harder
Antimicrobial resistance is where the One Health implications become particularly obvious, because resistance does not belong neatly to human medicine. Antibiotics are used in people and animals; resistant organisms and resistance genes can move through healthcare settings, farms, food systems, wastewater and the wider environment; and wastewater can contain biological material originating from homes, hospitals and businesses before treatment and discharge return some of it to environmental systems. WHO has consequently begun developing wastewater and environmental surveillance as part of the wider toolkit for monitoring antimicrobial resistance.
Now imagine that surveillance detects an unusual resistance mechanism. Human-health authorities may want to know whether corresponding resistant infections are appearing in hospitals; veterinary authorities may want to know whether similar organisms have been identified in animals; environmental authorities may need to consider possible sources or discharges; and laboratories may need to determine whether apparently separate findings are genetically related. No single dataset answers the question — the signal only becomes genuinely useful when information from several systems can be connected. That is One Health in a much more meaningful sense than simply placing human, animal and environmental health in three overlapping circles.
Three sensor networks
One way of understanding the future of One Health surveillance is to think in terms of sensor networks. The human sensor network — patients, clinicians, hospitals, diagnostic laboratories and public-health surveillance — produces information about disease in people. The animal sensor network — farmers, veterinary surgeons, animal keepers, laboratories and animal-health authorities — produces information about disease in domestic animals, livestock and wildlife. The environmental sensor network — wastewater, surface water, environmental sampling and other monitoring systems — provides signals about biological activity across populations and ecosystems. Food systems can form another surveillance layer, with microbiological sampling and supply-chain information connecting production with human exposure.
Each can see things the others cannot. The problem is what happens when one sensor network detects something first. The London polio episode provides one answer: environmental surveillance can expose transmission that has not generated an obvious clinical signal. Bluetongue provides another form of the same problem, where farmers and veterinary surgeons may become the first sensors detecting disease in animals. Foodborne outbreaks provide another, where genomic surveillance can reveal relationships between human cases before investigators understand the supply network connecting them. Different diseases, different sensors — and the same governance question: how quickly can information cross from the system that detects the signal into the systems capable of interpreting and acting upon it?
Europe is beginning to answer the question
The European Union’s revised Urban Wastewater Treatment Directive makes this particularly timely. Wastewater legislation might not seem like an obvious component of health-security policy, but the new Directive explicitly incorporates public-health surveillance into the management of urban wastewater. It requires Member States to establish cooperation and coordination between the authorities responsible for public health and those responsible for urban wastewater treatment — coordination that must address responsibilities, address costs, determine where and how frequently samples should be taken, and, importantly, establish arrangements for the appropriate and timely communication of monitoring results to the public-health authorities responsible for evaluating risks and taking action.
The Directive also provides for wastewater surveillance during public-health emergencies and develops a framework encompassing pathogens including SARS-CoV-2 and poliovirus, while antimicrobial resistance becomes increasingly important within future monitoring. This is more significant than simply telling countries to test sewage: Europe is beginning to govern the sewer as a public-health sensor.
The difficult part comes next
Installing a sensor is easier than designing the decision system around it. A country could create an impressive wastewater programme that collects thousands of samples, performs sophisticated sequencing and produces beautiful dashboards — and none of those things guarantees better health security. The important question is whether the information changes a decision. Did it trigger additional clinical surveillance? Did laboratories begin looking for something they were not previously testing for? Did vaccination strategy change? Did animal-health surveillance investigate a possible connection? Did an environmental investigation identify a source? And did the information arrive early enough to make any of those actions useful?
Surveillance should therefore be evaluated not only by how much it detects, but by what happens after detection.
From environmental signal to One Health incident
A more mature system could establish predefined escalation pathways. Routine environmental signals would remain within routine surveillance; an unusual signal meeting agreed criteria would trigger enhanced analysis; and if the signal persisted, increased or matched information appearing elsewhere, the incident could escalate into a cross-system investigation. At that point, relevant human, animal, food and environmental information could be brought together temporarily. This does not require a giant permanent One Health database — it requires interoperability on demand. The same principle applies to the food-traceability problem: different organisations can retain their own systems during normal operation, and what matters is whether relevant information can be connected rapidly when biology creates a reason to connect it.
An environmental incident might therefore generate a temporary graph containing wastewater sampling locations, laboratory observations, genomic sequences, clinical cases, animal-health observations, food or agricultural sites and relevant geographical information. The objective would not be to let software determine causality, but to let investigators see relationships they might otherwise have to discover manually across several institutions.
Measuring whether the system works
Wastewater surveillance also gives us an opportunity to make governance performance measurable. Instead of recording only sample collected → laboratory result, we could measure first detectable environmental signal → confirmed signal → interpretation → escalation → relevant agencies connected → intervention. That allows us to separate different forms of latency: detection latency tells us how long it took surveillance to see something; interpretation latency tells us how long it took to understand that the signal might matter; and governance latency tells us how long it took the system to connect the evidence and reach a position from which action could be taken.
Those measures could reveal something extremely useful. Perhaps the laboratory is not the bottleneck; perhaps sequencing is not the bottleneck; perhaps the delay occurs after we already know enough to start asking the right questions. That is precisely the kind of weakness conventional preparedness metrics can miss.
Testing the architecture
There is also an opportunity to test this idea rather than simply argue for it. The BioChain programme is exploring biological provenance: how biological objects, samples, observations and transformations can remain connected as they move through complex systems. Environmental surveillance provides an interesting extension of that work. A wastewater sample has a location, collection time and sampling method; laboratory testing produces observations; sequencing may produce a genomic profile; and that profile may subsequently be related to an isolate obtained from a patient, animal, food product or another environmental sample. Those relationships can be represented without claiming that one caused another.
A provenance-based demonstrator could therefore create a synthetic One Health surveillance environment containing environmental samples, human cases, animal-health observations, food-system data and genomic relationships. A simulated outbreak could then test how quickly investigators recognise a meaningful cross-system signal using conventional separate datasets, compared with a provenance-based incident graph. The relevant question would not be whether the software can detect a pathogen — laboratories already do that — but whether connecting provenance reduces the time between detection and understanding. That makes governance latency something that can potentially be measured experimentally.
The environment is no longer the third circle
One Health diagrams have traditionally placed human health, animal health and environmental health alongside one another, but in practice the environmental component has often been the least operationally developed. Wastewater surveillance changes that. The environment is no longer simply a background condition influencing disease emergence; it can actively produce surveillance information, and that fundamentally changes its role. A wastewater treatment plant can become an observation point for population health; environmental sampling can reveal biological activity before conventional surveillance recognises its consequences; and genomic analysis can potentially connect environmental observations with organisms found elsewhere. The environmental circle becomes a sensor — and the challenge is making sure somebody is listening.
The next surveillance problem
The future of wastewater surveillance will undoubtedly involve better sampling, more pathogens, improved sequencing and increasingly sophisticated analytical methods. Those developments matter — but the most important limitation may eventually cease to be what we can detect. It may become what we can meaningfully act upon.
The London polio experience demonstrated that the sewer can reveal transmission clinical surveillance has not yet made visible; AMR demonstrates how environmental information can cross the boundaries of human medicine, veterinary medicine and environmental management; and European legislation is beginning to formalise the relationship between wastewater monitoring and public-health decision-making. Together, they point towards a broader principle for One Health security: biological surveillance is only as useful as the system that receives its signals.
The question is therefore no longer simply whether we should watch the sewers. We should. The harder question is whether, when the sewer tells us something important, our institutions can recognise what it means, connect it with what human and animal surveillance are seeing, and act before the warning becomes yesterday’s information.
The sewer may know first. Our governance systems need to learn how to listen.
Related One Health Security analysis
This analysis is part of a wider series on how biological signals do — and do not — cross institutional boundaries: Salmonella and the Governance Gap, Bluetongue Is Moving Again, When Surveillance Systems Connect but Institutions Do Not and, on the global architecture of pathogen data, Who Owns the Warning?
References and further reading
- World Health Organization (2024). Wastewater and environmental surveillance for one or more pathogens: guidance on prioritization, implementation and integration. WHO, 6 December 2024.
- World Health Organization, WHO Hub for Pandemic and Epidemic Intelligence (2026). Wastewater and environmental surveillance — a global landscape analysis. WHO, 18 June 2026.
- European Parliament and Council (2024). Directive (EU) 2024/3019 of 27 November 2024 concerning urban wastewater treatment (recast). Official Journal of the European Union, 12 December 2024.
- European Centre for Disease Prevention and Control (2025). ECDC framework to guide the integration of wastewater-based surveillance into infectious disease surveillance at the EU/EEA level. ECDC, Stockholm, December 2025.
- World Health Organization (2025). Wastewater and Environmental Surveillance: Summary for Antimicrobial Resistance. Pilot version, 1 December 2025.
- UK Health Security Agency (2022). Poliovirus detected in sewage from North and East London. 22 June 2022.
- UK Health Security Agency (2022). Expansion of polio sewage surveillance to areas outside London. 2 September 2022.
- Klapsa, D., Wilton, T., Zealand, A. et al. (2022). Sustained detection of type 2 poliovirus in London sewage between February and July, 2022, by enhanced environmental surveillance. The Lancet, 400(10362), 1531–1538.
- World Health Organization (2023). Environmental surveillance for SARS-CoV-2 to complement other public health surveillance. WHO, 12 September 2023.
Key Takeaways
- Wastewater is a population-scale biological sample: it can reveal infection — from polio to antimicrobial resistance — before people ever reach a clinic.
- The London 2022 polio detections showed the sewer seeing sustained transmission that produced no clinical cases — but detection is only the beginning.
- "Detectability is not the same as actionability": the real constraint is increasingly not what we can detect, but whether institutions can interpret a signal and connect it across human, animal, food and environmental systems in time.
- The fix is interoperability on demand — connecting the relevant data when biology gives a reason to — plus measuring governance latency so we can see where the delay actually is.