From sewage sample to dot on the map — the science behind wastewater surveillance
When a person is infected with a respiratory or gastrointestinal virus, their body sheds viral RNA — genetic fragments of the pathogen — into their waste. These fragments flow into the sewer system and collect at wastewater treatment plants, where they can be measured using a technique called quantitative PCR (polymerase chain reaction).
This means a single wastewater treatment plant can act as a sentinel for an entire community — detecting disease spread among everyone connected to that sewer system, regardless of whether they see a doctor, get tested, or even feel symptoms yet.
The typical chain of clinical disease detection goes: person gets infected → symptoms appear (3–5 days) → person feels sick enough to seek care → visits doctor or urgent care → gets tested → test result processed → result reported to health department → aggregated into public data. This entire chain can take 1–3 weeks.
Wastewater skips most of that chain. Viral shedding begins before symptoms appear, and treatment plants sample continuously. The CDC has documented that wastewater surveillance consistently detects COVID-19, flu, and RSV surges 4–7 days before clinical case counts rise — giving communities a genuine early warning.
Raw viral concentration varies based on factors that have nothing to do with disease — rainfall dilutes samples, plant flow rates change, populations shift seasonally. To make readings comparable across sites and over time, the CDC uses a normalization method based on PMMoV (Pepper Mild Mottle Virus).
PMMoV is a plant virus found in chili peppers and hot sauce. Humans consume it regularly and excrete it at remarkably stable rates. By dividing the pathogen concentration by the PMMoV concentration in the same sample, the CDC produces a ratio that cancels out dilution and flow-rate effects — giving a cleaner signal of true community disease burden.
Outbreak Radar uses this PMMoV-normalized concentration for all calculations and visualizations.
A raw concentration number is hard to interpret — is 50 copies per mL high or low? It depends entirely on what's normal for that site. A large urban plant naturally processes more volume than a rural one.
To make every site comparable, Outbreak Radar scores each weekly reading as a percentile of that site's own historical active readings (weeks where the pathogen was detected above zero). A site at the 90th percentile is seeing levels higher than 90% of all active weeks in its recorded history — that's genuinely unusual, regardless of raw concentration.
The slider at the bottom of the map lets you scrub through every week of NWSS data going back to July 2022, when PMMoV normalization became consistent across sites. Hit the Play button to watch disease waves roll across the country over time — you can see COVID surges, flu seasons, and RSV fall peaks animate in real time.
The default view shows the most recent week where at least 75% of monitoring sites have reported. The CDC collects data on a rolling basis, so the very latest week often has partial coverage as sites finish submitting.
The CDC publishes new NWSS data weekly, typically on Fridays. Outbreak Radar rebuilds its data file automatically on a weekly schedule, so the map reflects the latest available readings shortly after CDC publication.