Reality research: Norovirus in restaurant bathrooms

Long-time friend and friend of the, Don Schaffner, a professor of microbiology at Rutgers University (right, sort of as shown) writes:

More than seven years ago I had the good fortune to be contacted by my colleague, Dr. Lee-Ann Jaykus (below, left, exactly as shown).

She asked if I wanted to be involved in what was at the time going to be a remarkable endeavor. She was going to lead a team of scientists competing to earn a $25 million grant from the USDA focused on understanding Norovirus.

Norovirus causes more foodborne disease than any other microorganism. Because it is often self-limiting, and seldom fatal, it gets a little attention. It’s also a remarkably difficult organism to study. One of the reasons it has been difficult to study is that there had no way to culture the organism outside it’s human host. This meant that anyone wanting to do research with the organism had to have a supply of frozen poop containing the virus.

One of the goals of the ambitious project lead by Dr. Jaykus was to finally crack the code which would allow scientists to culture the virus in the laboratory. Spoiler alert, we got the grant. We were all excited to learn recently that thanks in large measure to the USDA Grant, that riddle has been solved.

This USDA Grant also allowed a number of other research projects too numerous to recount here, but I do want to tell the story of one.

Early on in our efforts on the ground, my colleague, Angie Fraser reached out and asked if I wanted to be part of an extensive survey of restaurant bathroom for Norovirus. I was delighted to say yes, and we began

I have to express my sincere appreciation to Cortney Leone whom led the project. She had the unenviable task of having to oversee researchers in three U.S. states, charged with collection of the data for this project. I also owe huge debt of gratitude to my graduate student Hannah Bolinger who led our data collection efforts in New Jersey.  Thanks also to the NJ team of graduate students, undergraduate students and significant others who visited public bathrooms around the state (Louis Huang, Pierce Gaynor, Sarah Hossain, Sneha Sreekumar, Jenny Todd-Searle, and Arthur Todd-Searle).

(Schaffner, this isn’t an Academy Award acceptance speech, on with it — dp.)

Because we wanted to ensure that our data were representative, we collected data from nine different counties in New Jersey. This turned out to be a lot harder than you might think. New Jersey is a home rule state.

This means that public health operates at the municipal level, with minimal oversight from the state. Our first task was to compile lists of food service establishment from the three regions of New Jersey.

This was easy to do in less densely populated regions, where the municipal entity was the county. We could simply contact the county, and through appropriate freedom of information act paperwork, obtain a list of all of the restaurants in the county. This was far more difficult in the densely settled parts of the state, where obtaining a list require contacting each and every municipality in the county, and filling out each municipalities’ different paperwork, in some cases mailing a paper check to cover photocopying costs, and then eventually taking the information we received back, and putting it into a standardized database. All of this was required before we could even begin to collect the first piece of data.

Thanks to Hanna’s outstanding work, we were eventually able to produce a robust enough database that we could proceed with data collection. Hannah lead a team of students that visited restrooms in and around New Jersey. This was harder than it sounded, as often the information provided by the  municipalities was out of date, and the students arrived at a location, only to learn that the restaurant was closed, or the location was incorrect.

Eventually, the teams in all three states had collected enough swabs and sent the samples back to CDC for analysis.

The end result of all this work was published in the Journal of Food Protection. Although our goal was to visit 750 commercial food establishments, we actually visited 751 establishments, in which 1,044 bathrooms and 4,163 surfaces were swabbed.  Four swab samples were collected from each bathroom: (i) the underside of the toilet seat where it connects to the toilet bowl, (ii) the flush handle of the toilet, (iii) the inner door handle of the stall door or, when there was no stall door, the inner door handle of the outer door, and (iv) the hot water knob of the sink faucet.

In the end 61 (1.5%) of 4,163 swabs were presumptively positive for human norovirus, and 9 of these were confirmed by sequencing.  This is similar to what others have found.

Almost half (30) of positive swabs were found on the underside of the toilet seat. About 20% were found on the toilet flush handle (13) and the inner handle of the stall or outer door (11). Only 11% (7) of positive swabs were found on the sink faucet handle.  Our results suggest that areas further away from the toilet are less likely to harbor norovirus contamination; toilet surfaces (especially the underside of the seat) would be closest to vomiting and diarrheal events during which high numbers of norovirus particles could be shed.

Chain restaurants had significantly more positive samples than non-chains (p = 0.0273). Unisex bathrooms had significantly more positive samples than female bathrooms (p = 0.0163).  Bathrooms with bar soap had significantly more positive samples than liquid soap bathroom (p = 0.0056) and foam soap bathrooms (p = 0.0147), but note that only 3 bathrooms out of 751 actually used bar soap. Bathrooms containing a trash can attached to the paper towel dispenser had significantly more positive samples than bathrooms with a free-standing trash can (p = 0.0004).

Although[the NoroCORE grant recently ended,I know there will be continued publications coming for many years, several of which will come from my lab, that will serve to further advance our understanding of Norovirus, and the means by which it can be controlled.

Prevalence of Human Noroviruses in Commercial Food Establishment Bathrooms


Although transmission of human norovirus in food establishments is commonly attributed to consumption of contaminated food, transmission via contaminated environmental surfaces, such as those in bathrooms, may also play a role. Our aim was to determine the prevalence of human norovirus on bathroom surfaces in commercial food establishments in New Jersey, Ohio, and South Carolina under nonoutbreak conditions and to determine characteristics associated with the presence of human norovirus. Food establishments (751) were randomly selected from nine counties in each state. Four surfaces (underside of toilet seat, flush handle of toilet, inner door handle of stall or outer door, and sink faucet handle) were swabbed in male and female bathrooms using premoistened macrofoam swabs. A checklist was used to collect information about the characteristics, materials, and mechanisms of objects in bathrooms. In total, 61 (1.5%) of 4,163 swabs tested were presumptively positive for human norovirus, 9 of which were confirmed by sequencing. Some factors associated with the presence of human norovirus included being from South Carolina (odd ratio [OR], 2.4; 95% confidence interval [CI], 1.2 to 4.9; P < 0.05) or New Jersey (OR, 1.7; 95% CI, 0.9 to 3.3; 0.05 < P < 0.10), being a chain establishment (OR, 1.9; 95% CI, 1.1 to 3.3; P < 0.05), being a unisex bathroom (versus male: OR, 2.0; 95% CI, 0.9 to 4.1; 0.05 < P < 0.10; versus female: OR, 2.6; 95% CI, 1.2 to 5.7; P < 0.05), having a touchless outer door handle (OR, 3.3; 95% CI, 0.79 to 13.63; 0.05 < P < 0.10), and having an automatic flush toilet (OR, 2.5, 95% CI, 1.1 to 5.3; 0.05 < P < 0.10). Our findings confirm that the presence of human norovirus on bathroom surfaces in commercial food establishments under nonoutbreak conditions is a rare event. Therefore, routine environmental monitoring for human norovirus contamination during nonoutbreak periods is not an efficient method of monitoring norovirus infection risk.