From primary health center records and house-to-house searches, we identified 42 case patients from Sithalikuppam (40 [95%]) and Verupachi (2 [5%]) villages; 24 (57%) were male, and the median age of all case patients was 50 years (range: 22–80 years). Of the 42 case patients, 38 (90%) had diarrhea and 23 (55%) had vomiting. All patients recovered completely within 24 h; no deaths occurred. Cases were uniformly distributed by location, and we detected no geographic clustering. The overall AR for Sithalikuppam and Verupachi villages was 5% (42/855). The ARs were 65% (36/55) for those who attended the rally and 0.8% (6/800) for those who did not attended the rally.

The rally attendees left Cuddalore district at 9 a.m. on July 24 to travel to Tiruchirappalli district. All attendees ate only food prepared for the trip; lunch was at 3:30 pm. and dinner at 9 pm. Thirty-six (65%) of 55 attendees reported becoming ill between 5 a. m. and 6 p. m. on July 25 [Figure 1].

None of the food items eaten by rally attendees at lunch on July 24 was associated with illness [Table 1]. Analysis of the four dinner items revealed that only eating lemon rice was significantly associated with illness (AR among exposed and unexposed was 89% and 21%, respectively, with relative risk 4.2 (1.8–10.2) [Table 1]. The median incubation period from lemon rice ingestion at dinner and illness onset for rally attendees was 14 h (range: 9–20 h).

Six additional case patients who did not attend the rally reported eating leftover lemon rice, the only remaining food item from the rally, at 9 a. m. on July 25 [Figure 1]. All six persons reported vomiting between 8 p. m. on July 25 and 1 a. m. on July 26. The median incubation period from lemon rice ingestion to illness onset was 12 h (range: 10–15 h).

All ten (24%) stool specimens collected from the 42 identified case patients were negative for Shigella, Salmonella, and V. cholera by culture and microscopy.

The cook reported no diarrhea, vomiting, or other gastrointestinal illness symptoms during the month before the outbreak. Lemon rice was prepared on the ground in an open space from 3 a.m. to 5 a.m. on July 24. The cooked rice was spread over a dry cotton cloth (placed directly on the earthen ground) for approximately 1 h to cool. The rice was then placed in a vessel and mixed with lentils and spices that had been tempered in oil and peeled lemon. The preparation was then packed in four stainless steel containers and placed near the bus’s engine during the 6–7-h journey. The highest ambient recorded temperature on the rally day was 40°C (104°F).

We investigated a foodborne outbreak of diarrhea and vomiting among political rally attendees, as well as nonattendees who ate leftovers from the rally, in Cuddalore district, Tamil Nadu. The rapid response and detailed epidemiologic investigation identified illness associated with ingestion of lemon rice.

Even without laboratory confirmation, the epidemiologic, clinical, and environmental findings suggest the most likely etiologic agent was Bacillus cereus. Outbreaks of diarrheal B. cereus typically have a clinical presentation of mostly diarrhea and some vomiting after an incubation period of 8–16 h.^[5] Consistent with these characteristics, this outbreak had a similar incubation period and a predominance of diarrhea (90%) with some vomiting (55%). Sources of B. cereus include the soil, where end ospores are abundant.^[5] Many outbreaks of B. cereus have been associated with ingestion of boiled or fried rice kept at a high ambient temperature for a prolonged time, which results in bacterial growth and production of heat-resistant enterotoxins.^[6] In this outbreak, the lemon rice was not refrigerated after possible contamination with spores during preparation on an earthen surface. The B. cereus spores probably germinated as the rice cooled, and transport and storage at high ambient temperatures promoted bacterial growth with resultant enterotoxin production.

Without laboratory confirmation of B. cereus toxin, we cannot rule out other possible etiologies. The Staphylococcal aureus toxin was another possible etiology in this outbreak given the large proportion of vomiting among case patients,^[7] but S. aureus outbreaks generally have shorter incubation periods (<6 h).^[7] Clostridium perfringens is another possible etiologic agent. However, its incubation period is usually at least 24 h, and although clinical presentation is also predominantly diarrhea, vomiting rarely occurs. In addition, the source of infection is usually of animal origin.^[8]

Our findings are subject to several limitations. First, recall bias is possible because the cook and case patients were not interviewed until 7 days after the outbreak. There was no toxin testing available. Finally, at the time of the investigation, food samples were no longer available for testing. Nevertheless, the strength of the epidemiologic findings provides sufficient data to conclude an association between ingestion of contaminated lemon rice and development of illness.

Foodborne illness is an underestimated public health problem in India. Robust foodborne disease surveillance combined with high-quality outbreak investigations is needed to better clarify the extent, sources, and risk factors associated with foodborne illnesses. We recommended continued expansion of district-level capacity for laboratory-based surveillance of diarrheal disease pathogens to help detect outbreaks and identify their etiologic agents.