How sewage testing helps contain COVID-19
Scientists now know that the genetic signature of SARS-CoV-2 – the virus that causes COVID-19 – can be detected in sewage samples.
In fact, sewage analysis can detect the virus’ presence in a population days before positive cases or clusters show up in results of public screening programs.
Sewage sampling can detect virus levels within a wastewater catchment, the rate of increase over time, whether the virus is mutating out from epicentres, or if new strains are being introduced from outside the catchment.
This timeline provides a guide to the COVID-19 sewage testing process.
Dr Paul Bertsch is a member of a joint CSIRO–University of Queensland team that’s been at the forefront of research into the use of wastewater surveillance for detecting SARS-CoV-2 in the population.
Here, he explains how the method works, and why it’s becoming a critical part of our public health response toolkit.
ECOS: Why sample raw sewage? And how do you detect coronavirus?
Paul Bertsch: We now know that people start shedding the virus in their faeces about two to three days after first being infected with SARS-CoV-2 – well before they show symptoms of COVID-19, if they notice any symptoms at all.
After it’s flushed into the sewerage system, the virus gradually disintegrates, leaving behind fragments of its unique RNA signature.
Our approach is like the one used in TV shows, where forensics teams use genetic ‘fingerprinting’ to solve crimes. The RNA encoded gene fragments we recover from sewage are the unique fingerprint of the SARS-CoV2 virus.
We use filtration techniques to separate out the nano-scale viral and gene fragments from untreated wastewater, then ‘amplify’ the fragments, enabling us to isolate and confirm the virus’s fingerprint over any other genetic material present in the sample.
The process is called reverse transcription-quantitative polymerase chain reaction (RT-qPCR).
PCR is the molecular amplification or ‘molecular photocopying’ part – you make millions of copies of the gene fragments. The RT refers to reverse transcription, where you encode the virus’s single-strand RNA into complementary double-stranded DNA to produce the millions of copies of SARS-CoV-2 gene fragments. The amplified product is then sequenced to confirm it is the SARS-CoV-2 gene fragments.
RT-PCR is the standard way of testing for SARS-CoV-2, whether it involves samples of saliva, nasal mucus, or wastewater. So we’re simply applying a similar diagnostic testing approach used in individual testing to the collective SARS-CoV-2 ‘signature’ of entire communities.
Lead researcher Dr Warish Ahmed in the CSIRO laboratory where testing occurs.
Why has CSIRO’s Queensland study attracted attention?
The joint University of Queensland and CSIRO team produced the world’s first peer reviewed journal article documenting a proof-of-concept study on the use of wastewater analysis for detecting SARS-CoV-2 in the population.
The team sampled two wastewater treatment plants in the Brisbane region early in 2020, when Australia experienced its first wave of COVID-19. From the sampling, the team found the intensity of the SARS-CoV-2 signal in Brisbane’s wastewater was tracking clinical screening results in the community.
So the clinical testing results confirmed the wastewater methodology.
Is sampling wastewater for coronavirus safe?
Coronavirus is fragile and quickly rendered unviable, that is, not contagious, by detergents and the other surfactants in wastewater, and by typical water-treatment processes.
The UQ and CSIRO team members are trained to collect and analyse wastewater samples safely, using correct handling procedures and PPE equipment.
It’s also important to note that research to date suggests that drinking water is safe.
We constantly see news images of people being swabbed at testing points – isn’t clinical testing enough, in terms of tracing the virus’s spread through the community?
I should point out that wastewater sampling can’t replace individual sampling – it should be seen as complementary, providing unique information.
It’s not feasible for everyone in a community to be tested individually. Wastewater sampling is like an early warning system that covers an entire community. Especially in the early stages of COVID-19, many people will be asymptomatic. Some may never show symptoms and others may not want to be tested.
Researchers in the US recently did some modelling showing that up to 2.5 billion people globally could be monitored using wastewater surveillance, through regular sampling of 105,000 wastewater facilities.
The US study also suggested billions of dollars could be saved globally through reducing the need for individual testing and reducing the cost of across-the-board restrictions and economic shutdowns.
How accurate is wastewater sampling for coronavirus?
The US study estimated wastewater surveillance could detect one SARS-CoV-2 infection for every 100 people at the very least, in a monitored water-treatment catchment. At best, it could detect up to one infection per two million people.
The University of Arizona has used wastewater-based epidemiology at the individual dormitory level to manage the risk of COVID-19 among the returning student population. Early detection of SARS-CoV-2 in wastewater from one dormitory led to immediate individual testing of all 311 dormitory residents, with two asymptomatic students testing positive. These two were immediately isolated, preventing further spread of the virus and demonstrating the effectiveness of the approach.
Our own researchers have also confirmed that SARS-CoV-2 RNA seems to persist long enough in warmer conditions to permit detection – in other words, wastewater testing for the virus is reliable across a range of temperatures.
How long does it take to get a result?
SARS-CoV-2 can appear in faeces within two or three days of infection. Wastewater samples can be analysed within one to two days.
By contrast, it usually takes from five days to two weeks for people to develop symptoms severe enough for them to be tested.
In another recent study, our team was able to identify the most cost-effective and rapid sampling approach, which takes only 15–30 minutes per sample. We’re continuing to refine our sampling and analysis methods to improve their accuracy.
What about the future?
Some experts say we’re a long way from having a vaccine. So a national wastewater surveillance program or network could help Australian authorities detect and contain emerging COVID-19 clusters faster and more cost-effectively.
The technique can also be used to detect the virus in smaller populations, like aged-care facilities, schools and prisons, where COVID-19 prevention is particularly critical.
We recently demonstrated the effectiveness of on-board testing of wastewater on cruise ships for detecting SARS-CoV-2 before passengers disembark. This could play an important role in the resumption of cruises in future.
With biosecurity, nations tend to be reactive rather than proactive. Scientists, disease ecologists, and epidemiologists have been predicting the emergence of other pandemics following the SARS and MERS outbreaks earlier this century.
We need to prepare for future pandemics. The economics of permanent sewage surveillance stacks up – it costs very little in comparison to closing down entire economies.
As a leading public official in the US commented recently, if investing in sewage surveillance enabled his county to open even half a day earlier, it would be the best investment he ever made.
September 6, 2020 at 6:18 am
Thanks for the thorough but understandable explanation of this technique! Nice.
September 6, 2020 at 11:55 am
Thanks Eric! It’s great to hear it’s useful.
September 29, 2020 at 9:55 am
Fabulous information! Something I will be sharing with my Year 12 Biology students as we discuss application of genetic technologies
September 29, 2020 at 10:00 am
Great article. Wonderful to see the potential in really useful pro-active applications. Yes, let’s test the effluent of facilities and cruise ships. I imagine that if a new potent virus comes along, once we know its genetic sequence, an existing regime of testing wastewater can quickly swing across to monitoring for that as well
September 29, 2020 at 10:10 am
I concur with Eric. Very good article..
September 29, 2020 at 10:10 am
Hopefully a large number of people read this article, particularly decision makers. Of course that requires those decision makers to believe that science is relevant……
September 29, 2020 at 10:15 am
Exciting research. Pleased to hear of CSIRO’s involvement in this pandemic
September 29, 2020 at 10:20 am
Yes, great to hear about this approach. Thanks!
September 29, 2020 at 10:22 am
A comprehensive sewage testing program should be rolled out across the country asap especially as we move to relaxation of internal movement controls and fewer people present themselves for testing. This will be very helpful in providing a level of confidence that any new community transmission can be detected before it builds up too far. The costs of implementing such a program would be far less than further stop/start responses to new outbreaks.
September 29, 2020 at 10:53 am
Helen, a comprehensive summary of the need for federal funding for a national initiative this week. Perhaps 100% federal funding for all local government initiatives operational in October, declining to 50% of spending from January going forward.
September 29, 2020 at 11:02 am
Are you or the Colossus program putting forward a funding proposal through the Australian Health Protection Committee? Sounds as if now is the time to do it.
September 29, 2020 at 10:44 am
This is a game changer that technically validates decision making rather than the emotional responses and political points scoring coming from our political leaders. If we know theres CV19 in the wastewater, then take the sampling up the network until the source is in a group that can be 100% tested and we will crack it.
September 29, 2020 at 10:44 am
If you are taking your samples from waste water plants, how can you isolate the areas from where the samples come? If you were testing from the plant in a large rural town, would it be correct to say that, from your samples, there is 1 person in a town of 30,000 who may be, infected, but you have no idea who it is until they show symptoms? I can understand better if you were taking the samples from pumping stations and could therfore isolate a much smaller area of the town.
September 29, 2020 at 11:03 am
Thanks for your interest. You can test at any point in the system, and if you get a signal, it can trigger public health management decisions, including individual testing, in these population centres. The method can be used at various scales – pumping stations, individual sanitation tanks, or even at dormitory-level, as has been done at the University or Arizona.
September 29, 2020 at 5:10 pm
Thanks for your reply Helen. I am afraid don’t know enough about the sampling methodology, locations and frequency to fully understand the value of this exercise. If there are large numbers of people using the system two or three times a day how can you predict the numbers and exact location of the infected people?
September 29, 2020 at 5:54 pm
Hi, we’ve more information online, including the transcript of a briefing we offered when we published the world’s first peer-reviewed proof-of-concept trial. You can find more here: http://www.csiro.au/en/News/News-releases/2020/Australian-researchers-trace-sewage-for-early-warning-COVID-19-spread. If you want to dive deeper, we’re part of a global research collaboration: COVID-19 WBE Collaborative where you’ll find studies from across the world: https://urldefense.com/v3/__http:/www.covid19wbec.org/__;!!FvZmfVE!USLue3n0GgYRjr05Ewhp-dXusi-V3DkjWIzF4ONRmmEezQonH8rjIr8B6cjm-Fnhpw$
September 29, 2020 at 7:45 pm
Helen, I really don’t want to dive deeper. I have spent enough time submersed in sewerage ponds to last a lifetime (quite accidently of course). However I will continue to follow this up. It is the sampling methodology which interests me.
September 29, 2020 at 11:11 am
Excellent article and research. We need a multi-pronged approach to monitoring and detecting this virus, a silver bullet could be a long way off yet. This will play a vital part in helping to control the early detection and spread of the virus.
September 29, 2020 at 11:20 am
Thank you for this informative piece. Wastewater analysis is used to quantify drug use in the country and it is good to see the technique being used for human health as well.
Is it possible to use this technique to monitor diseases in feral animal popuations by testing creek and lake waters? African Swine fever, African Horse sickness and others would have a major effect on our domestic population is they were to enter Australia and the large feral populations would be major carriers. Grateful your thoughts?
September 29, 2020 at 11:41 am
Thanks David – interestingly, our lead researcher on this project Dr Warish Ahmed, also conducts research in areas such as identifying the microbial composition/biodiversity of aquatic ecosystems using next generation sequencing, tracking the sources of microbial contaminants (known as MST) in environmental waters using state-of- the-art technologies and quantifying bacterial, viral and protozoa pathogens in sand/sediment, wastewater and environmental water sources.
September 29, 2020 at 11:52 am
Roll it out!
September 29, 2020 at 4:04 pm
Is it known whether or not the virus sheds in urine? I ask because I’m an Australian expat in the USA where we are wrestling with when and how to reopen elementary (primary) schools. Testing here is still haphazard and not organized with an efficient sampling strategy. If the wastewater from a school could be tested it would yield a huge benefit. However, at day schools the great majority of students do not deposit faeces in the toilets, but almost all of them urinate once or more per day while at school, so to be effective it would be necessary to detect SARS-CoV-2 in diluted urine.
September 29, 2020 at 5:46 pm
Hi Steve, and thanks for your question. I’ve spoken with our experts who have advised that the literature provides evidence for shedding of SARS-CoV-2 in urine from some patients, but appears to be far less clear than shedding in faeces, with many studies indicating no evidence of shedding in urine (whereas SARS was shown to be shed in urine from most patients). There is also evidence that animal models, such as the ferret, which was used in the early work on vaccine screening, did shed in faeces and urine, demonstrating differences between species.
We’re art of a global collaboration sharing science on the topic, so you may find more information at the COVID-19 WBE Collaborative site: https://urldefense.com/v3/__http:/www.covid19wbec.org/__;!!FvZmfVE!USLue3n0GgYRjr05Ewhp-dXusi-V3DkjWIzF4ONRmmEezQonH8rjIr8B6cjm-Fnhpw$
September 29, 2020 at 7:38 pm
Helen,
Thank you for the reply. Too bad that the virus isn’t shed much in urine. That would have been great for schools.
In any case, keep it up!
Regards,
Steve
September 29, 2020 at 4:32 pm
Congratulations , This superb work combines practical and pure research and illustrates the importance of funding science and CSIRO.
September 29, 2020 at 4:49 pm
Great article, but the most important part is missing. Is there a plan in place to move beyond proof of principle ? Is there a national commitment to move very quickly from first detection of the virus in waste water to accurate localisation by rapid up-stream testing, then local containment and eradication by fast mobilisation of test-track-and-trace teams ?
September 29, 2020 at 5:58 pm
Thanks for your comment. We’re continuing to work on refining the methods, which provides information to support decision makers who have the remit to fund larger programs.
September 29, 2020 at 10:23 pm
Interesting article!
If you were to use it to monitor places like aged care centres and quarantine hotels, would catching the virus depend on whether you sample at the right time to get the waste from the infected resident, or is there enough mixing of the sewerage that it isn’t so time dependent? Wondering if you get a lot of false negatives.
September 29, 2020 at 10:33 pm
Thank you CSIRO.
Might I ask about the use of activated sludge from sewerage plants? I have heard that workers injuring (cut) themselves on site, will expose themselves to this as it will kill (in theory) any pathogen that has already come onto site. I’m not sure if this is more bacteria than virus.
Also, will we be using it to trace HIV-AIDS?
September 30, 2020 at 8:23 am
Hi Andrew, I’ve checked with our researchers who confirm that there is always a risk in working with activated sludge. If the workers wear appropriate PPE and wash their hands it should not be a problem. Workers are equally exposed to bacteria and viruses (in general pathogens). We are not planning to use this surveillance for HIV-AIDS but we can use it for other enteric viruses and also viruses associated with the pandemic.
September 30, 2020 at 8:40 pm
Thank you for a very comprehensive easily understood article. Very comforting to know that some way of detecting this virus is available. Congratulations and again grateful thanks
October 1, 2020 at 10:04 am
It would be great to know what the sewer sampling actually tells us about population infection rate and about the rate of asymptomatic infections.
October 1, 2020 at 12:38 pm
A great article – hopefully those making corona virus decisions use this tool to its fullest.
October 8, 2020 at 2:44 pm
Really excellent work. Just a technical question: Sewage is loaded with so many biological fragments from people, bacteria, & virus particles plus much other detritus, that separating out the virus particles for further work on the RNA by RT-PCR must be a difficult task.
How is that achieved? it sounds like it’s a complex process, so if it’s too lengthy to put here, are the details of the separation techniques published anywhere?
October 8, 2020 at 4:35 pm
Thanks Peter, I’ll share links to some of our recent papers with you.
Proof of concept
• A paper outlining the proof of concept was published in the Science of the Total Environment scientific journal https://www.sciencedirect.com/science/article/pii/S246858442030060X?fbclid=IwAR3jbpwvSRy1AB5yWx7YRA656pRd4HOZOKBKGd8GihIndJBRlEVudUzO25g#sec2
Methods paper
• CSIRO researcher Dr Warish Ahmed led the findings published in The Science of the Total Environment, which evaluated the concentration, recovery and detection of SARS-CoV-2 RNA (its genetic code)https://urldefense.com/v3/__http:/www.sciencedirect.com/science/article/pii/S004896972033480X__;!!FvZmfVE!USLue3n0GgYRjr05Ewhp-dXusi-V3DkjWIzF4ONRmmEezQonH8rjIr8B6chnn18Y1w$
Optimising the method and quality control
• This latest paper led by Dr Warish Ahmed was also published in late September. It looks at Methods optimisation and quality control are crucial for generating reliable public health information:https://www.sciencedirect.com/science/article/pii/S246858442030060X?fbclid=IwAR3jbpwvSRy1AB5yWx7YRA656pRd4HOZOKBKGd8GihIndJBRlEVudUzO25g#sec2
I hope these links are of value!
October 8, 2020 at 5:12 pm
Helen, Thank you for getting those links, they are exactly what I was looking for the other day Regards.
October 20, 2020 at 11:47 pm
“The hope is eventually we will be able to not just detect the geographic regions where COVID-19 is present, but the approximate number of people infected – without testing every individual in a location. This will give the public a better sense of how well we are containing this pandemic,” Dr Marshall said. Water headed for wastewater treatment plants can be testing for the presence of the novel coronavirus gene to develop an early warning surveillance system to track COVID-19 prevalence in the community.
October 24, 2020 at 9:06 am
You mentioned that virus levels can be detected. Is there a correlation between the amount of virus detected and the presence of known cases in a particular location?
September 30, 2021 at 12:22 pm
The sewage is a soup of many chemicals including bleach, detergent, cleaning agents, organic matters, medicine, etc. How can the COVID RNA survive in such a mixture? Any ‘fragments” could be elements or molecules of anything, not necessarily of COVID. It seems this kind of testing is very far fetched. I can believe the testing accuracy from a pure sample such as urine, blood or faeces, but not from a mixture of many chemicals that can react with each other.
September 30, 2021 at 1:13 pm
Hi Sam, and thanks for your comment. Our researchers advise that we most likely pick up partially damaged virus with intact RNA. It is possible that there may be some free floating RNA as well which will be degraded. We extract RNA in the lab and immediately test with PCR to avoid degradation.
Here are our two papers on virus decay and RNA decay, where you’ll find more detail:
https://www.sciencedirect.com/science/article/pii/S0013935120309890
https://pubs.acs.org/doi/abs/10.1021/acs.estlett.0c00730