Restaurant-Associated Pontiac Fever—Will That Be Chills and Fever To Go?

Abstract & Commentary

Synopsis: Shortly after patronizing a restaurant, 117 individuals developed an illness characterized by fever, myalgia, chills, and headache. Respiratory symptoms were present in fewer than half. Legionella anisa was found to be the cause of this outbreak of Pontiac fever.

Source: Jones TF, et al. Epidemiologic investigation of a restaurant-associated outbreak of Pontiac fever. Clin Infect Dis. 2003;37:1292-1297.

Over a period of several days in April 2002, a number of people reported becoming ill shortly after eating at a restaurant in Nashville, Tenn. Predominant symptoms in the initial group of patrons were fever, headache, nausea, vomiting, and diarrhea, but later reports included chills and myalgia, with lesser gastrointestinal symptoms. County and state health departments initiated a case-control study to determine the cause and possible means of transmission of the outbreak.

After reviewing the restaurant’s reservation lists and credit card receipts and receiving reports of additional cases, health department investigators identified 117 individuals who reported being ill with fever and associated symptoms within 5 days of eating at the restaurant. An unidentified number of restaurant patrons who experienced no illness after a similar interval served as a control group. Among the ill individuals, fever (present in all by case definition), myalgia (93%), chills (92%), and headache (87%) were the most common symptoms; diarrhea and vomiting occurred in 30% and 16%, respectively. The incubation period was 49 hours (range, 4-120 hours). Illness resolved after a mean of 3 days (range, 4-192 hours).

Further epidemiologic investigation could identify no link to any food consumed at the restaurant. Enterovirus, adenovirus, respiratory syncytial virus, influenza virus, parainfluenza virus, and herpesvirus cultures were negative.

Because the restaurant contained many fountains, pools, and misting machines, as well as a waterfall, large collecting pool, and several large fish tanks, environmental samples were submitted for Legionella cultures. Legionella anisa was isolated from 2 water samples and a swab specimen from a large ornamental pool. Fifty-eight percent of ill persons (vs 18% of controls) recalled sitting near the fountain and pool. Half of ill individuals who provided acute and convalescent serum specimens had a 4-fold or greater antibody titer rise to > 1:256 to the suspect strain of L anisa; none of a healthy group of individuals from Nashville had significant levels of antibody.

Inspection of the water handling system in the restaurant revealed several areas where poor water circulation may have prevented adequate disinfection, and bromine levels and pH deviated from generally recommended standards. The air conditioning system appeared to be well maintained and functioning normally.

Comment by Jerry D. Smilack, MD

Jones and associates are to be congratulated for a nifty piece of epidemiologic work. Any outbreak of illness associated with a restaurant immediately raises the specter of contaminated or spoiled food or beverage. Jones et al, initially anticipating a food-borne etiology, did what all epidemiologists would have done: launch a case-control study to inquire what foods or beverages might have been the source of the illness. Surprisingly, they ran into a blind alley. However, recalling that several outbreaks of Pontiac fever have been associated with whirlpools and hot tubs and 1 with a hotel lobby decorative fountain, they pursued the possibility of Legionella infection by obtaining water and other environmental samples from the restaurant, where fountains, pools, and misters abounded. Their efforts were rewarded by isolating L anisa from several sites. They then demonstrated that many of the ill persons seroconverted to the isolated bacterial strain, clinching the case. Nice bit of detective work!

Is Legionella-induced Pontiac fever a common cause of what would otherwise appear to be viral-like outbreaks? Since fewer than 2 dozen have been recognized over the past quarter-century, one can only speculate. This fascinating report by Jones et al may stimulate others to look for Legionella in other outbreaks.