Healthcare Infection Prevention-The night Norwalk virus crossed the scrimmage line

'Did you say sic'em or sick them?'"

How did Karen Becker, DVM, MPH, an epidemiologist with the Centers for Disease Control and Prevention, solve a puzzling case of apparent transmission of Norwalk virus between opposing college teams on the football field? In the tradition of the sport, she studied films.

"I watched this game on tape several times in an attempt to understand it," she says. "There was a lot of hand-to-mouth contact. I saw people being pushed in the face. One can also get it not just through fecal-oral [contact], but in the areosolization of vomitus. If vomit is on the [jersey] and they smash into somebody, viral particles can aerosolize and can be ingested. They also wiped their hands with their mouth pieces."

The saga begins with a turkey sandwich

The story of this highly unusual outbreak begins on Friday, Sept. 18, 1998, when, after a locker-room box lunch of turkey sandwiches, the members and the staff of the Duke University Blue Devils football team flew to Tallahassee, FL, for a showdown with perennial powerhouse Florida State University.

Toe hit leather at 7 p.m. the following night, and the outmanned Duke squad began battling in earnest to bring down the mighty Seminoles. But as the game proceeded, several Duke players began to rush off the field with symptoms of vomiting and diarrhea. Some went back to the locker room, but as the game proceeded, players were seen throwing up on the sideline and reentering the game.

When the dust cleared, FSU had triumphed by a score of 62 to 13, but severe gastrointestinal distress would be the price of victory for 11 Seminoles. Ironically, that is the number needed to field a team without penalty. On first impression, no one surmised that Norwalk virus had crossed the line of scrimmage. But Becker, then a CDC Epidemic Intelligence Service (EIS) officer stationed at the North Carolina Department of Health in Raleigh, sensed something was amiss.

'The only possibility was the playing field'

"I was thinking we ought to focus on foods and water that were shared. Were there any common sources?" she says. "But I couldn't find any possible source of linkage between the two [teams]. I realized that the only possibility was on the playing field. These people had nothing in common otherwise."

The investigation began to focus on that locker-room box lunch the Duke team ate before flying south. The rate of attack among those who ate the box lunch was 62%. Of the 36 North Carolina players who were defined as having primary or secondary cases, 19 (53%) began to have vomiting, diarrhea, or both before or during the game.

The 18 patients from North Carolina who were not football players included coaches, trainers, and team physicians. Four stool samples obtained from North Carolina patients were positive for Norwalk virus on electron microscopy. All samples as well as one of two stool samples from players on the Florida team were positive for a genetically identical Norwalk virus of genogroup 1.¹

"They were identical strains," she says. "We continued to do the epidemiological analysis. The thing that came out as a clearly higher risk for those that got sick was eating the turkey sandwiches. That was the first meal in North Carolina. That made sense, because people who didn't go to the game still got sick if they ate that box lunch."

Becker and co-investigators narrowed the search to food handlers who prepared the turkey sandwiches. The two persons who prepared the box lunches had no reported history of illness. Investigators took stool samples from the workers, but knew it was probably too late in the game for them to be shedding virus. So they drew blood looking for antibodies. Indeed, one worker had a clinically significant rise in the titer of antibody against recombinant Norwalk virus antigen — suggesting a recent infection with the bug.

"We drew blood samples because the immune response in the body is a little delayed," she says. "We took a baseline and then [drew again] three weeks later, and we saw more than a four-fold rise in antibody titers. A lot of people get exposed to this, so having one measure and seeing antibodies to Norwalk virus wouldn't tell us much. But by seeing a more than four-fold rise in the antibody to Norwalk virus, we were able to say with confidence that this was recent exposure."

However, the worker could have been an asymptomatic carrier or the sandwich ingredients also could have been contaminated from another source in the kitchen. A previous outbreak of the virus was linked to the consumption of a tossed salad that may have been cross-contaminated by raw shellfish, which had been rinsed in a sink that was later used to separate lettuce, Becker reminds. This type of contamination might have occurred in the football outbreak because the lettuce and tomatoes used for the sandwiches were washed in the same sink that was used for other foods, utensils, and pots and pans.

Regardless, shoe-leather epidemiology and microbiological evidence all point to Duke having an extra player on the field that night — a strain of Norwalk virus that is actually relatively rare in the United States. "Most of us probably have immunity to genotype 2 because we have had a case of the typical vomiting and diarrhea 24 to 36 hours after eating a meal," she says "This was a genotype 1, which is uncommon in the U.S."

Infectious tackling

One thing continued to puzzle investigators. With time of possession of the ball roughly equal — creating similar opportunities among all players for contact and infection — why did only FSU offensive players become infected? Investigators theorized that some of the Duke defensive players may have been more "efficient transmitters" of the virus. Moreover, some of the defensive players who had vomiting and diarrhea during the game made most of the tackles.

"During the time of acute illness, Norwalk is most infectious," she says. "That is the time it is easiest to get infected through secondary means from person to person."

This is apparently the first time such an outbreak has been documented in a football game, but Becker suspects it has occurred but gone unreported. "There is a lot of contact in sports," she says. "I think blood-to-blood transmission is very rare. It just doesn't happen. However, in terms of anything that is GI tract [related] — especially a virus that has shown to be as contagious as this one — it probably happens more than we think. When you watch the number of times they have hand-to-mouth, hand-to-body [contact] and just tumble into each other, it doesn't take many viral particles to ingest and become infected. It is very low infectious dose, and there were many opportunities."

The recommendation to football teams is to bench players with severe gastrointestinal symptoms because, as has now been shown, transmission could occur to the opposing team. "It is really the best recommendation we can make," she says. "You should use another string of players. You don't want people to go out there and get dehydrated and get really sick. A lot of these folks ended up on intravenous fluids. They couldn't drink anything without vomiting."

Reference

1. Becker KM, More CL, Southwick KL. Transmission of Norwalk virus during a football game. N Eng J Med 2000; 343:1,223-1,227. n

---