• Users Online: 161
  • Home
  • Print this page
  • Email this page
Home About us Editorial board Ahead of print Current issue Search Archives Submit article Instructions Subscribe Contacts Login 

 Table of Contents  
Year : 2017  |  Volume : 19  |  Issue : 1  |  Page : 28-33

Food poisoning outbreak in a training establishment: A retrospective cohort study

1 Station Health Organization, Eastern Naval Command, Visakhapatnam, Andhra Pradesh, India
2 Office of Director General Armed Forces Medical Services, New Delhi, India
3 Headquarters Western Naval Command, Mumbai, India

Date of Web Publication17-Aug-2017

Correspondence Address:
Surg Cdr Maramraj Kiran Kumar
SHO (V), Malkapuram, Gandhigram Post, Near INHS Kalyani, Naval Hospital, Visakhapatnam - 530 005, Andhra Pradesh
Login to access the Email id

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/jmms.jmms_17_17

Rights and Permissions

Background: An outbreak of food poisoning occurred among recruits in a training establishment. Investigation of outbreak was undertaken with active preventive interventions concurrently to arrest the current outbreak as well as to avoid such incidents in future. Materials and Methods: A retrospective cohort study was undertaken among all recruits, regardless of presence or absence of symptoms. The risk ratios (relative risks) and attributable risks were calculated for each food item of the suspected meal to assess the association between consumption of individual food items and subsequent illness. An environmental survey was undertaken to investigate into the course of food processing and storage facilities at trainees' galley (cookhouse) and other relevant eating establishments. Results: A total of 494 recruits reported with symptoms of gastroenteritis in a span of 3 days. Of those affected, only 9 were admitted and rest recovered with treatment on OPD basis. The overall attack rate was 22.9%. No deaths were reported. It was a classical point source, single exposure gastroenteritis outbreak. When food histories and sickness histories were analyzed, the attributable risk (24.17) and relative risk (5.11) were highest for the “Flavoured milk,” which was an outsourced item. The statistical findings were substantiated with environmental and epidemiological evidence. Conclusion: Epidemiological investigation incriminated dinner of the previous day as the meal responsible for the outbreak with flavored milk as the most attributable food item.

Keywords: Food poisoning, food safety, outbreak, Salmonella

How to cite this article:
Kumar MK, Bhaskar V, Ray S. Food poisoning outbreak in a training establishment: A retrospective cohort study. J Mar Med Soc 2017;19:28-33

How to cite this URL:
Kumar MK, Bhaskar V, Ray S. Food poisoning outbreak in a training establishment: A retrospective cohort study. J Mar Med Soc [serial online] 2017 [cited 2023 Feb 6];19:28-33. Available from: https://www.marinemedicalsociety.in/text.asp?2017/19/1/28/213095

  Introduction Top

The term “Food poisoning” is usually used when a large number of persons are affected with similar gastrointestinal symptoms and signs, at the same time with a history of intake of a common meal.[1] The “in-living personnel” (personnel staying in the single accommodation without families) constitute a high-risk group for food poisoning because of community kitchen practices. In such settings, every meal is a common meal and hence detection of the item which caused the incidence of food poisoning may not be as simple a common-sense conclusion.[2]

The “recruit/trainee group” in Armed Forces is considered as one of the high-risk groups for outbreaks of communicable diseases including foodborne diseases. This is possibly attributed to the close communal living and training quarters, operational constraints, physical exertion, sleep deprivation, compromised hygiene, and psychological stress all resulting in physiological and immunological changes.[3] Prompt intervention to reduce the loss of training man-hours is therefore considered as vital in such settings as the training programs are highly structured and multicentric in nature.

A training establishment reported an occurrence of 494 cases of gastroenteritis among recruits on three consecutive days of a summer season (from 11 to 13 April). The number of cases reported was obviously much higher than the expected incidence, and warranted an immediate response. The investigation was started within 24 h of occurrence of the first case with an aim to arrest the outbreak and identify the factors responsible that might have led to this outbreak, to prevent such incidents in future.

  Materials and Methods Top

All recruits in the training establishment dine in a centrally located, common dining hall, and consume food prepared from a single cookhouse/galley. The weather conditions were hot, humid with no known outbreaks or increased incidence of foodborne diseases in the vicinity.

Outbreak of food poisoning was confirmed among the recruits when a large number of them reported with similar gastrointestinal symptoms and signs. All the available suspected food and water samples, stool samples of cases and food handlers were taken. Being a very remote area, all these samples were transported in a cold chain to a laboratory of the nearest service hospital (around 330 km away) for microbiological analysis.

Line listing of all the cases revealed the clinico-epidemiological profile and descriptive characteristics of the outbreak in terms of time, place, and person. The symptomatology matched with that of  Salmonella More Details spp and therefore tentatively Salmonella was suspected to be the cause of the outbreak. An epidemic curve was drawn and based on minimum, maximum and median incubation periods of Salmonella, the meal responsible for outbreak was traced.

A “case” for the outbreak was defined as any recruit presenting with three loose stools in past 24 h.[4] A retrospective cohort study was undertaken among all recruits, regardless of the presence or absence of symptoms. The whole study population (recruits) was available for investigation; therefore, requirement for adopting a sampling method did not arise. The historical data pertaining to the outbreak was collected from the recruits using an epidemiological case sheet containing items for personal information, details of food items of suspected meals taken, presenting symptoms, time of onset of symptoms, medical care provided, and time required for recovery.[5] The additional information as required was obtained from health record cards of recruits. Health record card is a document, designed by the unit to keep a track of the health record of the individual in the form of OPD visits and hospital admissions during the training period.

The attack rates, risk ratios (relative risks), and attributable risks were calculated for each food item of the suspected meal to assess the association between consumption of individual food items and subsequent illness.

An environmental survey was undertaken to investigate into the course of food processing and storage facilities at trainees' galley (cookhouse). All food-handlers were medically examined. The water distribution system from source to consumer end was inspected for any points of cross-contamination. The butchery/slaughtering house, the wet canteen attached to dining hall (which supply refreshments to recruits during stand-easy/tea-breaks) and the civil factory (which manufactures and distributes milk and flavored milk to recruits) were all inspected.

Preventive measures were also undertaken concurrently to control the outbreak such as super-chlorination of drinking water, health awareness sessions for the recruits and food-handlers on daily basis. All externally prepared food supplies from wet canteen and civil factory were suspended for the interim period.

  Results Top

A total of 494 recruits suffered from either of the symptoms of loose motions, fever, abdominal pain, and vomiting within the duration of 3 days. Out of 494 recruits, 379 met the case definition of the current outbreak. These 379 recruits reported to the casualty (MI room) with complaints of loose motions with or without fever. Of those affected, only 9 had moderate-severe dehydration requiring hospital admission. The rest had mild or no dehydration and recovered with treatment on OPD basis. The overall attack rate was 22.9%. No deaths were reported.

Clinico-epidemiological profile is shown in [Table 1]. The majority (76.72%) had presented with loose motions as the predominant symptom. About 74.6% had fever, 9.7% had pain abdomen, and 16.3% had vomiting as the predominant symptom. There was no history of blood in the stools. Majority had fever between 100°F to 102°F. There was no mortality and all the cases made an uneventful recovery. Majority of the cases recovered with oral rehydration solution. Antibiotics and intravenous fluids were required in some hospitalized/OPD cases.
Table 1: Clinico-epidemiological profile of the outbreak

Click here to view

From the clinico-epidemiological profile, it was ascertained that symptomatology (diarrhea predominantly with fever and without dysentery) was matching that of Salmonella species [5] and hence a tentative diagnosis of Salmonella food poisoning was made. The incubation period of nontyphoidal  Salmonellosis More Details is 6–72 h, but illness usually occurs within 12–36 h after exposure.

When the onset of cases was plotted on X axis against number of cases on Y axis, an epidemic curve as shown in [Figure 1] was obtained. As classically seen in any other point source-single exposure outbreaks, it has a sharp upslope and a well-defined peak. The index case reported at 08:00 h in the morning sick report on April 11 with a history of onset of symptoms at around 02:00 h. The cases reported throughout the day and the number reached the peak at around 20:00 h on 11 April. The downward slope was a little prolonged, and an unusual small secondary peak was seen due to late presentation of certain mild to moderate cases and also possible malingering cases, which could not be ruled out. In the epidemic curve, a tight clustering of cases in time is seen. Nearly, all cases occurred within a single maximum Incubation period of suspected organism, Salmonella spp.
Figure 1: Epidemic curve.

Click here to view

After plotting the epidemic curve [Figure 1], when the minimum, median, and maximum incubation periods of Salmonella were traced back from the onset of symptoms of the first case, case at peak and last case, they converged approximately at the dinner time on April 10, the day before occurrence of outbreak (recruits had given a history of consumption of the dinner at around 19:00 h). Therefore, this meal was presumed to be the most suspected meal, which might have led to the outbreak.

Outbreak description in terms of place and person is depicted in [Table 2]. All the recruits who were grouped into nine divisions were accommodated in six blocks. It is observed that higher number of cases had occurred in Golf division (28.8%), Charlie division (23.97%), and Foxtrot division (21.49%). However, from the spot map plotted, it was clear that none of these three divisions share the common block or common water supply. When outbreak in terms of person was further analyzed, it was observed that 53 trainees of “other forces,” who consumed the same food prepared from the common cookhouse of recruits, did not report any illness.
Table 2: Outbreak description in terms of place and person*

Click here to view

Identifying the food item attributable to food poisoning using attributable risk is illustrated in [Table 3]. In addition to the three main meals (prepared in the cookhouse), certain refreshments were provided to the recruits daily in the form of samosa during morning breaks and flavored milk postdinner. The samosa was prepared and supplied by the civil canteen, which is adjacent to the dining hall of recruits. The flavored milk was manufactured in a civil factory exclusively for these recruits, which is located at around 35 km from the training establishment and is supplied to the recruits through a middle agency, which is located around 3 km from the training establishment. When food histories and sickness histories were tabulated and analyzed, the attributable risk (24.17) and relative risk (5.11) were highest for the “Flavoured milk.” This statistical finding was corroborated by the history of change in color and smell of the flavored milk served that day, as given by some of the recruits. Fifty-three trainees of “some other security forces” share all the three meals with recruits, cooked in the common galley, however, refreshments such as samosa and flavored milk were served only to recruits. This fact further substantiated the statistical finding.
Table 3: Identifying the food item attributable to food poisoning

Click here to view

Overall hygienic conditions of trainees' galley (cookhouse) and dining hall were satisfactory. However, it was observed that on most occasions, dinner for all the recruits is usually prepared and kept ready by 14:30 h, but served at around 18:30 h. Since storage of such large quantities of prepared food in refrigerators or under cold conditions is not feasible, it is left at room temperature till it is served, which might have proved unsafe especially during hot weather conditions. All the food handlers were clinically found to be free from infection, and their monthly medical chart maintained in the cookhouse. The hygienic condition of the wet canteen was satisfactory. There was no rat nuisance or infestation found in the cook house and wet canteen. Water distribution and sewage disposal system were satisfactory and no cross-contamination found anywhere. The hygienic condition of butchery was satisfactory.

On inspection of the civil factory, which manufactures flavored milk for recruits, hygiene conditions were found to be unsatisfactory. The vanilla flavor was added to the chilling unit after pasteurization and from there the flavored milk was diverted to packaging unit. The possibility of cross-contamination of the flavored milk exists, as the flavor was added to the chilling unit after pasteurization. Further, it was found that the vanilla flavor which was used did not have any food safety and standard trademarks on the packing label. There was no rat nuisance or infestation found in the factory. The cold chain system of the flavored milk from the place of manufacturing through the middle agent before reaching the consumer end was also found to be highly unsatisfactory (the storage facility was also found to be grossly inadequate).

It was thus evident from the circumstantial inquiry that the manner in which the flavored milk was manufactured through the various stages, and transported to the consumer, afforded ample opportunity for growth of organisms. Therefore, the statistical finding commensurate with the history obtained from the patients as well as the findings of environmental survey, incriminating the contaminated flavored milk to be the most suspected cause of outbreak.

The laboratory findings of the samples collected are shown in [Table 4]. All the samples were sterile except that of fruit custard of lunch on April 10 and water sample from trainee's galley, where  Escherichia More Details coli growth found. Since there was no left-over sachet of flavored milk served on April 10, sample of the same was not available and could not be sent to the laboratory. However, the sample of flavored milk of next day was preserved and sent to laboratory, which did not show any microbial growth.
Table 4: Laboratory results of food, water, and stool samples

Click here to view

  Discussion Top

A total number of 494 cases of gastroenteritis reported among recruits in 3 days (from 11 to 13 April). The overall attack rate was 22.9% with no deaths. It was a classical point source, single exposure gastroenteritis outbreak. Epidemiological investigation incriminated the dinner of previous day as the meal responsible for the outbreak with flavored milk as the most attributable food item which was an outsourced item. The hygienic conditions of the manufacturing factory from where the food item was outsourced as well as the storage facilities (cold chain system) were found to be unsatisfactory. The validity of this statistical finding was also corroborated by the fact that 53 trainees of “some other security forces,” who had consumed the suspected meal (but not provided flavored milk) remained unaffected.

The clinical and epidemiological features provided important clues to etiology. In Clostridium perfringens food poisoning, abdominal cramps, and diarrhea within 8–16 h can occur. However, fever is uncommon, occurring in <10% of the patients. In the present outbreak, 75% had fever; therefore, C. perfringens food poisoning was excluded. Fever, abdominal cramps, and diarrhea within 16–48 h (as in the present outbreak), are usually due to Salmonella or Shigella. Shigella was excluded, as there was no blood in the stools in majority of the cases.[5],[6],[7]

Outbreaks due to E. coli are almost and always due to E. coli O157:H7.[7] Incubation period is usually 3–4 days, diarrhea is generally bloody.[8],[9] The patient is afebrile or has only a low-grade fever. The patient suffers from nausea and vomiting in 50% of cases. Therefore, E. coli was excluded to be the cause in the current outbreak.

Salmonella is the most commonly reported cause of bacterial food-borne illness. Typically, nontyphoidal Salmonella produces a self-limiting febrile gastrointestinal illness, usually with nonbloody stools.[10] Based on the clinical presentation, Salmonella species appeared to be the most possible causative organism.

Salmonella food poisoning is most often caused by improperly handled or cooked poultry or eggs.[11],[12],[13] However, it is not uncommon that milk products made from insufficiently pasteurized milk or contamination after pasteurization with Salmonella result in large outbreaks of salmonellosis in humans.[14] Milk is an ideal food for bacterial growth and could cause food poisoning. Salmonella serovars commonly isolated from milk include Salmonella typhimurium, Salmonella montevideo, Salmonella newport, Salmonella dublin, and others. After the invention of pasteurization, outbreaks due to the contaminated milk have come down significantly. However, some outbreaks which were reported in the past due to milk consumption were all due to cross contamination after pasteurization. For instance, in a major outbreak attributed to Umpqua dairy in Roseburg, it was found that a machine that washes cases carrying products on a conveyor belt around the plant was infected with salmonella and cross contaminated the milk after pasteurization.[8],[14]

The epidemiological finding could not be corroborated by laboratory analysis due to nonavailability of the suspected flavored milk sample. Further, laboratory analysis revealed E. coli growth in one of the water samples and a fruit custard, which could not be linked to epidemiological findings. The water sample was taken from one of the taps in galley, which is used for cooking rice. The rice cooking in the galley was found to be steam based, where a hot steam enters the boiler from one side and water is added to rice from another side. It is very unlikely that E. coli had sustained high temperatures to contaminate the steamed rice. E. coli was also found in the sample of fruit custard. However, if E. coli in the fruit custard (of April 10 lunch) is suspected to be the cause of outbreak, onset of symptoms in index case would have been on or after April 12 instead of April 11 (minimum incubation period is 48 h). Despite sterile precautions undertaken during sampling and transport, cross contamination cannot be ruled out because of multiple agencies involved.

The availability of whole study population for investigation and the feasibility to adopt retrospective cohort-type study design were important strengths of the study. However, nonavailability of serotyping and other laboratory facilities closer to the training establishment (being a remote locality) was one of the limitations in conduct of the study. The nonavailability of suspected flavored milk sample for laboratory analysis and certain degree of unavoidable recall bias when food histories were collected from the subjects were the other limitations of the study.

If resources permit, characterization of the epidemiology of outbreaks by performing serotyping would be highly useful in conduct of such outbreak investigations.[15] Serotyping assists in linking the epidemiological findings to the etiology of the outbreak, especially when the outbreak is due to multiple serotypes.

Health education of the food handlers and dissemination of food safety message needs to be stressed upon. The personnel, especially food handlers, should be made fully aware of the food safety standards prescribed by Food Safety and Standards Authority of India.[16] The principles of Hazard Analysis Critical Control System may be adopted to identify, prioritize, and control the potential problems.[17],[18],[19] Hand hygiene, washing vegetables and fruits properly, cooking or reheating the food thoroughly, and keeping cooked food at safe temperatures preferably below 5°C are the basic food safety keys to prevent any food borne disease.[20] Storage in cold conditions and reheating large quantities of prepared food may not be practically feasible in certain situations (like in training establishments). In such settings, the timings of preparation of breakfast, lunch and dinner for the recruits in the galley are to be planned in such a way that the storage time of the prepared food should not exceed 2 h before serving, as recommended by the WHO.[20],[21]

  Conclusion Top

Epidemiological investigation incriminated dinner on the previous day as the meal responsible for the outbreak with flavoured milk, which was an outsourced food item, as the most attributable. Our study emphasised that in addition to the internal food safety checks, periodic meticulous inspection of hygienic conditions and storage facilities of the outlets from where food items are outsourced is of paramount importance.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.

  References Top

Directorate General Armed Forces Medical Services. Food poisoning. In: Manual of Health for the Armed Forces. Vol. 1. New Delhi: Directorate General Armed Forces Medical Services; 2008. p. 605-12.  Back to cited text no. 1
Park K, editor. Food poisoning. In: Park's Text Book of Preventive and Social Medicine. 23rd ed. Jabalpur: M/S Banarsidas Bhanot; 2015. p. 227-34.  Back to cited text no. 2
Ho ZJ, Hwang YF, Lee JM. Emerging and re-emerging infectious diseases: Challenges and opportunities for militaries. Mil Med Res 2014;1:21.  Back to cited text no. 3
Public Health Surveillance: IDSP Guidelines for Medical Officers. Ministry of Health and Family Welfare, Government of India; 2005.  Back to cited text no. 4
Bhalwar R. Investigation of epidemic. In: Textbook of Public Health and Community Medicine. 1st ed. New Delhi: AFMC Pune in Collaboration with WHO-India Office; 2009. p. 184-196.  Back to cited text no. 5
Jadhav SL, Sinha AK, Banerjee A, Chawla PS. An outbreak of food poisoning in a military establishment. Med J Armed Forces India 2007;63:130-3.  Back to cited text no. 6
Mustafa MS, Jain S, Agrawal VK. Food poisoning outbreak in a military establishment. Med J Armed Forces India 2009;65:240-3.  Back to cited text no. 7
General Information on E. coli. Available from: http://www.cdc.gov/ecoli/general/index.html. [Last accessed on 2016 Nov 16].  Back to cited text no. 8
Factsheet on Enterohaemorrhagic Escherichia coli (EHEC). Available from: http://www.who.int/mediacentre/factsheets/fs125/en/. [Last accessed on 2016 Nov 16].  Back to cited text no. 9
Foodborne Disease Outbreaks: Guidelines for Investigation and Control. World Health Organization, Geneva; 2008.  Back to cited text no. 10
Butaye P, Michael GB, Schwarz S, Barrett TJ, Brisabois A, White DG. The clonal spread of multidrug-resistant non-typhi Salmonella serotypes. Microbes Infect 2006;8:1891-7.  Back to cited text no. 11
Chin J, Salmonellosis. In: Control of Communicable Diseases Manual. 17th ed. Washington: American Public Health Association 2000;440-4.  Back to cited text no. 12
Black RE. Diarrheal Diseases. In: Nelson KE, Williams CM, Graham NMH eds. Infectious disease epidemiology. Aspen Publishers, Gaithersburg 2001;497-517.  Back to cited text no. 13
Reported Outbreaks of Food Borne Illness. Available from: http://www.realmilk.com/foodborne.html. [Last accessed on 2016 Dec 05].  Back to cited text no. 14
Multiple-Serotype Salmonella Gastroenteritis Outbreak after a Reception, Connecticut, 2009. Morbidity and Mortality Weekly Report No. 34. Center for Disease Control and Prevention; 2010.  Back to cited text no. 15
Food Safety and Standards in India Prescribed by FSSAI. Available from: http://www.fssai.gov.in. [Last accessed on 2017 Jan 02].  Back to cited text no. 16
Bhalwar R, Tilak VW. Application of hazard analysis critical control system: Analysis of a food poisoning outbreak. Med J Armed Forces India 1996;52:100-4.  Back to cited text no. 17
Bryan FL. Hazard analysis critical control point evaluations. World Health Organization, Geneva; 2008.  Back to cited text no. 18
Gandhi AP. Development of HACCP protocols for the production of soy milk. Asian J Food Agro Ind 2009;2:262-79.  Back to cited text no. 19
Prevention of Foodborne Disease: Five Keys to Safer Food. Available from: http://www.who.int/foodsafety/consumer/5keys/en/index.html. [Last accessed on 2016 Dec 05].  Back to cited text no. 20
Dangerous Food Safety Mistakes. Available from: http://www.foodsafety.gov/keep/basics/mistakes/index.html. [Last accessed on 2016 Dec 05].  Back to cited text no. 21


  [Figure 1]

  [Table 1], [Table 2], [Table 3], [Table 4]


Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
Access Statistics
Email Alert *
Add to My List *
* Registration required (free)

  In this article
Materials and Me...
Article Figures
Article Tables

 Article Access Statistics
    PDF Downloaded423    
    Comments [Add]    

Recommend this journal