AMERICAN JOURNAL OF EPIDEMIOLOGY

Vol. 131, No. 4

Copyright © 1990 by The Johns Hopkins Univernty School of Hygiene and Public Health All right* reserved

Printed m U S A.

EPIDEMIC CHOLERA IN WEST AFRICA: THE ROLE OF FOOD HANDLING AND HIGH-RISK FOODS MICHAEL E. ST. LOUIS,1 JOHN D PORTER,1 ANTOINETTE HELAL,3 KANDJOURA DRAME,3 NANCY HARGRETT-BEAN,' JOY G. WELLS,1 AND ROBERT V. TAUXE1 St Louis, M. E. (COC, Atlanta, GA 30333), J. D. Porter, A. Helal, K. Drame, N. Hargrett-Bean, J. G. Wells, and R. V. Tauxe. Epidemic cholera in West Africa: the role of food handling and high-risk foods. Am J Epidemiol 1990;131:719-28. During an epidemic of cholera in Guinea, West Africa, in 1986, the authors conducted two studies of risk factors for transmission. In the capital rity, 35 hospitalized cholera patients were more likely than 70 neighborhood-matched controls to have eaten leftover peanut sauces (odds ration (OR) = 3.1, 95% confidence interval (Cl) 1.2-8.2), but less likely to have eaten tomato sauces (OR = 0.2, 95 percent Cl 0.1-0.9). Hand washing with soap before meals by all family members protected against cholera (OR = 0.2, 95 percent Cl 0.02-0.96), suggesting that persons asymptomatically infected with Vibrio cholerae 01 may have been the initial source for contamination of the leftover foods. Laboratory studies demonstrated that V. cholerae multiplied rapidly in peanut sauce (pH 6.0), but not in the more acidic tomato sauce (pH 5.0). In an outbreak of cholera-like illness after a rural funeral, illness was strongly associated with eating a rice meal served over many hours without reheating. These studies demonstrated that, in this epidemic, many cases of severe cholera were associated with eating specific cooked foods that could support bacterial growth after contamination of these foods with V. cholerae within the household. Epidemic control efforts should include identification of high-risk foods and promotion of simple changes in food handling behaviors to lower the risk of foodbome transmission. cholera; food handling

Epidemic cholera was not recognized in demic resulting in more than 150,000 cases West Africa until 1970, when a major epi- and 20,000 deaths spread from coastal West Africa to involve much of the continent (1). Received for publication February 3, 1989, and in Cholera has since become endemic in Affinal form October 5, 1989. rica, and epidemics occur repeatedly (2). Abbreviations: Cl, confidence interval; OR, odds Epidemic cholera has historically been asratio, RR, relative risk. . , ... , .. 1 Enteric Diseases Branch, Division of Bacterial SOCiated With contaminated water supplies, Diseases, Center for Infectious Diseases, Centers for but growing evidence indicates t h a t foodDisease Control Atlanta, GA. ' fame transmission can also play an impor1 Division of Field Services, Epidemiology Program . , /o> T .,,, • , , , ,, tant role about t h e Office, Centers for Disease Control, Atlanta, GA. w e recently investigated risk Control, Atlanta, GA. factors for cholera during a large urban 719

720

ST LOUIS ET AL

epidemic and in a cluster of rural cases after the funeral of a cholera victim. We found a common link in the consumption of foods cooked thoroughly at first, but then subject to potential contamination with Vibrio cholerae and storage for several hours at ambient temperatures that would allow bacterial multiplication. These findings suggest that intrahousehold foodborne transmission of cholera may be common in West Africa and also that specific control measures may be useful in future epidemics. BACKGROUND

Guinea, a coastal West African country of 6.1 million persons (figure 1), recorded epidemics of cholera in 1970 and 1978. Beginning in January 1986, at the peak of the dry season, localized outbreaks of cholera were reported in Forecariah Prefecture and other rural areas with case-fatality ratios of 10-60 percent. An epidemic swept the capital city of Conakry (population, 600,000) in the wet months of June to September (figure 2). From January to October 1986, 297 (13.0 percent) fatalities were reported among 2,279 cases. Deaths were reported in 71 (4.9 percent) of 1,448 cases in

Conakry and 226 (27.2 percent) of 831 rural cases. Conakry is supplied by three water mains providing chlorinated water from different sources. However, water pressure is erratic and may be absent for several hours each day. Many persons also drink or bathe with well water. During the heavy rains of midsummer, sewers may overflow, and wells may be flooded with surface water. Attack rates for cholera in the 1986 epidemic in Conakry did not vary be sex, ethnic group, or residential zones receiving water from any of the three water mains. The overall attack rate for Conakry residents treated for cholera in a hospital was 2.4/1,000 residents. The principle meal of the day in Guinea is typically prepared in late morning for consumption at noontime. An extensively boiled sauce or soup is usually poured over a pot of rice. Household members scoop food with fingers from this common pot either directly to their mouths or less frequently to individual plates; some persons use a spoon rather than fingers for eating. Extra sauce may either be poured or scooped with fingers or a spoon from the pot holding the sauce. Refrigeration is not

CONAKRY

FIGURE 1. Map of Guinea showing areas primarily affected by cholera.

721

CHOLERA IN WEST AFRICA 700 -,

I 1 Rural arvaa K

Conakry

COM—contro* study In Conakry

200-

Jan

Fob

Uar

Apr

Uay

Jun

Jul

Aua

S«p

Oct

FIGURE 2. Reported cases of cholera-like illness in Guinea, January to October 1986.

commonly available for leftovers. The evening meal is less formal and is often eaten individually rather than collectively; one or two individuals may eat the leftovers from the noon meal either reheated or not, usually 6-10 hours after the noon meal. METHODS

surface), use of hand washing with soap before meals and after defecation, usual practices and methods for storing leftover foods, the usual market at which commodities were purchased for the household, and consumption within the 5 days before illness (or before interview for controls) of common foods prepared in the home and/ or purchased ready to eat. During home visits before and during the case-control study, the pH of sauces and soups was routinely measured with litmus paper. We attempted to collect stool or rectal swabs for isolating V. cholerae from all patients. Attempts were also made to collect suspect foods for culture during home visits, but no foods eaten by case-patients before the onset of illness were still available.

Investigation of the epidemic in Conakry For a case-control study in Conakry, we defined a case of severe cholera as vomiting and watery diarrhea that began abruptly in a resident of Conakry aged 14 years or older and led to hospitalization between September 6 and 13 with administration of intravenous fluids required. Family members of these patients led us to their homes, from which point we systematically selected Laboratory investigations dwellings in search of two controls matched for sex and for age within 10 years (within Rectal swabs of patients were trans5 years for patients aged less than 25) who ported in Cary-Blair medium to the Public had not had watery diarrhea, defined as Health Laboratory in Conakry, where they three liquid stools within 24 hours, in the were inoculated into thiosulfate citrate bile preceding 30 days. A questionnaire was ad- salts sucrose medium. Colonies typical of ministered to study subjects in their pri- V. cholerae were subcultured and tested for mary language, covering the source for agglutination with V. cholerae 01 antidrinking and bathing water (piped, well, or serum. At the Centers for Disease Control,

722

ST LOUIS ET AL.

Atlanta, Georgia, selected isolates were identified and tested for antimicrobial resistance by disk diffusion and for toxin production using the Y-l adrenal cell assay. To study the growth of V. cholerae in leftover sauces, we prepared tomato and peanut sauces according to recipes from Conakry. Tomato sauce was made from fresh ripe tomatoes, canned tomato paste, cooking oil, onions, cloves, garlic, chili peppers, sugar, salt, black pepper, chicken bouillon cubes, and water. Peanut sauce contained the same ingredients but with fewer tomatoes, and peanuts ground to a coarse paste were added. The sauces were boiled for 30 minutes, then allowed to cool to 30*C (the approximate afternoon temperature in Conakry in September) and inoculated with a culture of V. cholerae 01 El Tor Ogawa recovered from a patient in Conakry. The pH of the sauces was measured with litmus paper. Plate counts were performed by doing 10-fold serial dilutions after inoculation and after 6 and 10 hours. The thermal kill of V. cholerae in reheated peanut sauce was investigated by warming the leftover, contaminated sauce with occasional stirring and determining the temperature at which V. cholerae could no longer be recovered despite enrichment in alkaline peptone broth before plating onto thiosulfate citrate bile salts sucrose medium. Investigation of the funeral-associated outbreak To investigate an outbreak of choleralike illness after a funeral in a rural district in Forecariah Prefecture, we visited all households in the village in which the funeral took place and recorded the household members' participation in and food consumption at the funeral. We defined a case of cholera-like illness as vomiting and watery diarrhea (three or more stools per day) in a person 3 years of age or older that resulted in a visit to a health facility. No cultures were obtained from persons associated with this outbreak.

Statistical methods The case-control study was analyzed using a conditional logistic regression model with estimates of the odds ratio (OR) and 95 percent confidence interval (CI) calculated by the maximum likelihood method (4, 5). Variables significant at the 95 percent level in the univariate analysis were included in the multivariate model. Etiologic fractions were calculated by the method described in Schlesselman (6). For the outbreak after the funeral, the confidence intervals for the relative risk (RR) were calculated by the method described in Thomas and Gart (7). RESULTS

The case-control study in Conakry The case-control study was conducted after the peak of the epidemic in Conakry (figure 2). A total of 35 (19 men and 16 women) of 40 patients who met our case definition were enrolled in the case-control study. The median age was 28 years (range, 14-60 years). Of the remaining five patients, three were discharged, and two died before interview. Three (9 percent) of the study patients reported other cases of diarrheal illness in the household within the preceding 2 weeks. Rectal swabs in 23 (88 percent) of 26 study patients yielded V. cholerae 01. Culture-negative patients and those not cultured were clinically indistinguishable from patients whose stools yielded V. cholerae; most patients had received antimicrobials before specimens were collected. Isolates were confirmed as V. cholerae El Tor Ogawa; all were toxigenic and sensitive to all antimicrobials tested. In univariate analyses, eating leftover peanut sauce without reheating it was significantly associated with severe clinical cholera, but eating peanut sauce at the primary (noon) meal was not associated with an increased risk (table 1). Eating tomato sauce at the noon meal appeared to be protective against cholera, but no asso-

723

CHOLERA IN WEST AFRICA TABLE

1

Case-control study of potential nsk factors for seuere cholera during an urban epidemic, Conakry, Guinea, 1986 )exposed

Univanate analysis Habitual exposures Exclusive use of piped water For drinking For bathing Habitual consumption Leftover rice Leftover sauces Routine hand washing with soap* Before meals After defecation Use of utensils* for the noon meal Exposures within the past 5 dayst Tomato sauce Eaten at the noon meal Eaten unreheated as a leftover Peanut sauce Eaten at the noon meal Eaten unreheated as a leftover Meat sauce Other sauce Smoked fish Fresh fish Shellfish (clams) Fresh vegetables Raw fish Any prepared foods in the marketplace Multivariate analysis^ Routine hand washing with soap* before meals Tomato sauce at noon meals§ Peanut sauce leftover, cold§

Odds ratio

interval

74 57

1.36 1.66

0.35-5.17 0.61-4.52

83 80

77 87

1.48 0.48

0.49-4.45 0.14-1.55

9 14 23

26 27 36

0.21 0.35 0.44

0.04-0.95 0.09-1.29 0.15-1.31

83 54

96 47

0.19 1.32

0.04-0.95 0.59-3.00

91 71 14 14 94 77 14 17 11

90 46 17 7 99 84 3 9 13

1.17 3.21 0.79 2.50 0.25 0.64 5.00 2.16 036

0.30-4.51 1.24-8.31 0.24-2.56 0.57-10.9 0.24-2.56 0.23-1.75 0.97-25.8 0.65-7.19 0.23-3.25

34

29

131

0.54-3.19

9 83 71

26 96 46

0.22 0.14 3.10

0.04-1.16 0.01-0.96 1.17-8.24

Cues (n-35)

Controls (n - 70)

77 65

* By all household members, not just study subjects. t Within 5 days of onset of illness for cases and of interview for controls. % Includes the three variables below in the model. § Within 5 days preceding illness (cases) or interview (controls).

ciation was evident between eating leftover associated with eating in restaurants or at tomato sauce and the risk of cholera. We vendor stalls, with the type of water used found no association between cholera and • for drinking or bathing, or with the coneating other leftover or fresh foods except sumption of ice or iced drinks. Reported for a borderline association between illness hand washing with soap by all family memand eating shellfish (clams) harvested from bers before eating meals was protective tidal estuaries in Conakry (OR = 5.0, 95 against cholera (table 1). Hand washing percent CI 0.97-25.8, p = 0.055); only five with soap after defecation by all family (14 percent) of the 35 patients ate clams, members was also associated with a deand all reported that the clams were well- creased risk of cholera, but was practiced cooked. No increased risk was found to be in few households in our study and was

724

ST LOUIS ET AL.

not statistically significant. Multivariate analysis yielded little change in the odds ratio for the three variables that were significant in the univariate analysis, indicating that they were independent risk factors (table 1). The proportion of cases of severe cholera attributable to consumption of leftover peanut sauce (the etiologic fraction) was estimated from the odds ratio of 3.2, using the 46 percent rate of exposure to leftover peanut sauce among controls in the casecontrol study (table 1) as an estimate of the proportion of the population exposed. For hospitalized cases of cholera in Conakry, the population etiologic fraction for eating leftover peanut sauce was 50 percent, suggesting that approximately half of the cases of severe cholera during this time period in the epidemic in Conakry could be attributed to consumption of unreheated, leftover peanut sauces.

about the participation of 263 household members in the funeral and their subsequent history of illness. A total of 11 cases of cholera-like illness and three deaths in nine households occurred in the week after the funeral (figure 4). Cases of cholera-like illness occurred in eight of 16 households in which persons ate a rice meal with goat meat sauce, but in only one of 24 households in which no household member ate the rice meal (RR = 12.0, 95 percent CI 1.87-467.6, p < 0.01). Three women who had cleaned the bed sheets and body of the index cholera victim (including evacuation of bowel contents with enemas) helped to prepare the food for the funeral shortly thereafter. The goat meat sauce did not contain peanuts. DISCUSSION

These two studies of the transmission of cholera during an epidemic in Guinea demonstrated foodborne transmission by Growth and thermal kill of V. cholerae in cooked foods in which V. cholerae had the leftover sauces opportunity to multiply because of proThe pH of tomato sauces and other longed exposure of food to ambient tempersauces used as condiments for rice in atures. The peanut sauces implicated as Guinea ranged from 4.5 to 5.0, while sauces vehicles for transmission of cholera in Concontaining ground peanuts ranged from pH akry differ fundamentally from those re6.0 to 7.0. The tomato sauce prepared in ported in past outbreaks. Previously idenAtlanta was pH 5.0 and the peanut sauce tified vehicles of epidemic transmission, inwas pH 6.0; the pH of each remained con- cluding contaminated water (8-10) and stant over the 8 hours of holding time. V. seafood (11-13), have generally been items cholerae 01 multiplied rapidly in the exper- that were contaminated with V. cholerae imental peanut sauce, while no substantial organisms outside the home, usually at a growth was evident in tomato sauce (figure single common source. By contrast, peanut 3). Despite the presence of greater than 107 sauces in Conakry were prepared de nouo organisms per gram of peanut sauce after 8 in individual households, were thoroughly hours of holding at ambient temperatures, cooked at initial preparation, and were not no viable organisms could be recovered associated with an increased risk of cholera from the sauce after it was reheated to when eaten at a first meal. Therefore, these approximately 65 °C. This was substan- sauces almost certainly became contamitially below the temperature at which the nated with V. cholerae within individual sauce began to boil visibly (86 °C). households rather than at a common source. The growth of V. cholerae to high levels in experimental preparations of peaFuneral-associated outbreak nut sauce but not in tomato sauce suggests A total of 40 of 42 households in the that the different risks of developing severe village where the funeral had taken place cholera after eating these leftover sauces were successfully contacted for information

725

CHOLERA IN WEST AFRICA 10 -I Peanut sauce Tomato sauce 10 01

as |

10 2

4 8 Hours after inoculation

8

10

FIGURE 3. Recovery of Vibno cholerae El Tor Ogawa from leftover sauces. 4-.

Fun oral

a

^

Deaths

Death

l

s m

O 1•

0-

i

26 ' 27

W/< 28 29

* 30

August

31

1

2 3 September

4

'Secondary caao In household FICURE 4. Cases of cholera-like illness after the funeral of a cholera victim, Forecariah, Guinea, 1986.

may have been due to their differing abili- isms. Cooked foods are a more favorable ties to support the rapid growth of V. chol- milieu than raw foods for growth of V. erae. cholerae (16, 17) because cooking destroys These epidemiologic findings are consis- ' many competing bacteria. The growth of V. tent with the microbiology of V. cholerae. cholerae is inhibited by low pH and is enIngestion of a larger inoculum of organisms hanced in substances of neutral or alkaline is associated with more severe diarrhea (14, pH. Peanuts and other lentils contain pep15). Multiplication of V. cholerae 01 in a tides that neutralize and buffer acidic foods potential vehicle may therefore be impor- to which they are added (18). It is therefore tant in determining its likelihood of causing plausible that adding ground peanuts to severe illness, particularly if the vehicle is sauces increases the risk of transmitting initially inoculated with only a few organ- cholera by raising the pH of those foods,

726

ST LOUIS ET AL.

permitting small numbers of inoculated organisms to grow to numbers sufficient to cause illness. Even subtle modifications of the chemical milieu of foods may substantially change bacterial growth: In one experiment, rice gruels that differed only in the season during which the rice was harvested yielded substantially different growth rates for V. cholerae (19). In addition, the capacity of the peanut sauces to buffer stomach acidity may have lowered the dose of organisms it was necessary to ingest to cause clinical illness, an effect seen with other foods (20) and with antacids (14, 15). However, in evaluating these findings, we are unable to control for the additional effect on the severity of cholera of known host factors for which we have no information, such as blood group (21) and hypochlorhydria (22).

amplification in a nutrient-rich medium (3, 27). Such an inoculation of cooked foods by contaminated water has been incriminated in the past as the cause of a large outbreak of cholera in the United States (28). Our inability to demonstrate an association between cholera and the type of drinking water (piped vs. well) may have been due to a true lack of association, to insufficient power in the case-control study, or to a situation in which—at different times and places during the epidemic in Conakry— either piped or well water could have been contaminated with V. cholerae, so that neither type of water supply could be shown to be a risk factor in the aggregate by comparison with the other.

Although only 14 percent of patients with severe cholera reported eating shellfish, the elevated odds ratio and borderline statistiAlthough almost all (91 percent) patients cal significance (table 1) of this association with severe cholera in the case-control raise several possibilities. First, shellfish is study in Conakry were the initial cases of a well-recognized vehicle of transmission diarrheal illness in their households, the for V. cholerae 01 in other geographic areas, high frequency of asymptomatic infections (11, 13) and may have been contributing to during epidemics of El Tor cholera (25-100 a subset of cases in Conakry too few for our infections: one severe case of cholera (23- case-control study to definitively associate 25)) constitutes a substantial reservoir of with shellfish consumption. Second, alV. cholerae 01 for the intrahousehold con- though the shellfish were reportedly eaten tamination of foods. The protection against well-cooked, it ia possible that they served cholera afforded by hand washing with soap as one means for introduction of V. cholerae by all family members before meals sug- into households, after which other foods gests that V. cholerae shed by asymptomat- became cross-contaminated, eventually ically infected household members could leading to illness from consumption of spehave been one source for-eontamination of cific foods that supported the growth of V. leftover foods. In previous cholera epidem- cholerae to a high inoculum. Finally, it is ics, increased risk of cholera-like illness has possible that shellfish or some other massbeen demonstrated in persons who ate contaminated food or water source was a foods prepared by ill food handlers (26). principal vehicle of transmission of cholera Since household members in Guinea gen- early in the epidemic, while at the later erally share in the main noon meal by stage of the epidemic during which we conscooping food from one common pot by ducted the case-control study (figure 2), hand, each family member effectively be- widespread environmental contamination comes a "food handler" with respect to the and asymptomatic carriage of V. cholerae leftovers. Contaminated household water is leading to intrahousehold transmission via another plausible source of V. cholerae for foods had become a predominant mode of inoculation of high-risk foods, particularly transmission. if the water contains a low number of orAfter the funeral for a cholera victim in ganisms unlikely to cause illness without rural Guinea, an outbreak of cholera-like

727

CHOLERA IN WEST AFRICA

illness was shown to be associated with a rice meal served during the course of the burial vigil. Although no cultures were available to document V. cholerae 01 as the cause of this outbreak, clusters of deaths from dehydration in adults are highly likely to be due to V. cholerae infection. In this outbreak, the same persons who cleaned the index victim's body later prepared the meal, providing an obvious means for contamination of the food. Although the sauce in this instance did not contain peanuts and the pH was unknown, the foods were held at ambient temperature and served over 14 hours, an ample time for multiplication of V. cholerae. V. cholerae multiplies rapidly in cooked rice (17), and cooked rice in which V. cholerae has had the opportunity to grow has been associated with past outbreaks (28). In the neighboring West African country of Mali, a village outbreak of cholera was associated with leftover millet gruel in which the acidic, soured milk traditionally added had been unavailable because of drought (25). Moist, leftover, nonacidic grains may in general represent specific high-risk foods for cholera transmission in West Africa. Better understanding of the modes of transmission of cholera will foster more specific and effective means of control. Cholera outbreaks after funerals and in hospitals in Africa have frequently been attributed to person-to-person transmission (29-32). Distinguishing foodborne from direct contact transmission may help focus prevention efforts more precisely. Determination of the pH of common foods and practices for storing leftovers may be useful to quickly assess the risk associated with intrahousehold foodborne transmission. The etiologic fraction calculated for peanut sauce in the case-control study suggests that approximately half of the hospitalized cases in Conakry were due to consumption of this one high-risk, low-acid food, but only when eaten as a leftover without being reheated. Transmission of this kind could be reduced by reheating

leftovers to a boil before eating. Hand washing with soap before meals may lessen the risk of contamination of foods. The addition to high-risk foods of acidifying agents such as tomatoes, tamarind, sour milk, or vinegar that are locally available and acceptable may be considered when outbreaks occur, particularly if reheating foods is not feasible. Modifying food handling behavior at funerals during epidemics may represent a more practical and acceptable approach to disease control than would prohibiting funerals entirely (29), which may conflict with strong religious mores and traditions in the community. REFERENCES

1. Goodgame RW, Greenough WB. Cholera in Africa: a message for the West. Ann Intern Med 1975;82:10l-6. 2. World Health Organization. Cholera in 1986. Weekly Epidemiol Rec 1987;62:141-2. 3. Feachem RG. Environmental aspects of cholera epidemiology III. Transmission and control. Trop Dis Bull 1982;79:l-47. 4. Breslow NE, Day NE. Statistical methods in cancer research. Vol 1. The analysis of case-control studies. Lyon: International Agency for Research on Cancer, 1980:248-79. 5. Lubin JH A computer program for the analysis of matched case-control studies. Comput Biomed Res 1981;14:138-^t3 6. Schlesselman JJ. Case-control studies; design, conduct, analysis. New York: Oxford University Press, 1982:43-5. 7. Thomas D, Gart J. A table of exact confidence limits for differences and ratios of two proportions and their odds ratios. J Am Stat Assoc 1977,72:736. 8. Snow J. On the mode of communication of cholera. In: Snow on cholera. London: Oxford University Press, 1936 9. Blake PA, Rosenberg ML, Florencia J, et aL Cholera in Portugal, 1974. II. Transmission by bottled mineral water. Am J Epidemiol 1977;105:344-8. 10. Sinclair GS, Mphahlele M, Duvenhage H, et al. Determination of the mode of transmission of cholera in Lebowa. S Afr Med J 1982;62:753-5. 11. Baine WB, Mazzotti M, Greco D, et aL Epidemiology of cholera in Italy in 1973. Lancet 1974;2:1370-4. 12. Mclntyre RC, Tira T, Flood T, et al. Modes of transmission of cholera in a newly infected population on an atoll: implications for control measures. Lancet 1979;1-311-14. 13. Blake PA, Rosenberg ML, Bandeira Costa J, et al Cholera in Portugal, 1974. I. Modes of transmission. Am J Epidemiol 1977; 105:337-43. 14. Hornick RB, Music SI, Wenzel R, et al. The Broad

728

15. 16.

17. 18.

19.

20.

21.

22.

ST LOUIS ET AL.

Street pump revisited; response of volunteers to ingested cholera vibrios. Bull NY Acad Med 1971;47:1181-91. World Health Organization Scientific Working Group. Cholera and other vibrio-associated diarrheas. Bull WHO 1980:58:353-74. Teng PH. The role of foods in the transmission of cholera. In: Bushnell OA, Brookhyser CS, eds. Proceedings of the Cholera Research Symposium. Washington, DC: US Department of Health, Education, and Welfare, 1965:328-31. Kolvin JL, Roberts D. Studies on the growth of Vibrtc cholerae biotype El Tor and biotype Classical in foods. J Hyg Camb 1982:89:243-52. US Department of Agriculture. The composition of foods: legumes and legume products. Raw, processed, and prepared. Agriculture handbook no 8 16. Washington, DC: US GPO, 1986:112. Benenson AS, Ahmad SZ, Oseasohn RO Personto-person transmission of cholera. In: Bushnell OA, Brookhyser CS, eds. Proceedings of the Cholera Research Symposium. Washington, DC: US Department of Health, Education, and Welfare, 1965:332-6. Levine MM, Black RE, Clements ML, et aL Volunteer studies in development of vaccines against cholera and enterotozigenic Escherichia coU: a review. In: Holme T, Holmgren J, Merson MH, et aL, eds. Acute enteric infections in children, new prospects for treatment and prevention. Amsterdam: El«evier/North Holland Biomedical Press, 1981:443-59. Glass RI, Holmgren J, Haley CE, et aL Predisposition for cholera of individuals with O blood group. Possible evolutionary significance. Am J Epidemiol 1985:121:791-6. Sack GH, Pierce NF, Hennessey KN.etaL Gastric

23.

24.

25.

26. 27.

28.

29. 30.

31. 32.

acidity in cholera and non-cholera diarrhea. Bull WHO 1972:47:31-6. Bart KJ, Huq Z, Khan M, et al. Seroepidemiologic studies during a simultaneous epidemic of infection with El Tor Ogawa and classical Inaba Vibrio cholerae. J Infect Dis 1970;121 (Suppl)-17-24. Gangarosa EJ, Mosley WH. Epidemiology and control of cholera. In: Barua D, Burrows W, eds. Cholera. Philadelphia, PA: W. B. Saunders Co., 1974-381-403. Tauxe RV, Holmberg SD, Dodiri A, et al. Epidemic cholera in Mali: high case mortality and multipte routes of transmission in a famine area. Epidemiol Infect 1988:100:279-89. Holmberg SD, Harris JR, Kaye DE, et al. Foodborne transmission of cholera in Micronesian households. Lancet 1984:1:325-8. Spira WM, Khan MU, Saeed YA, et aL Microbiological surveillance of intraneighborhood El Tor cholera transmission in Bangladesh. Bull WHO 1980:58:731-40. Johnston JM, Martin DL, Perdue J, et aL Cholera on a gulf coast oil rig. N Engl J Med 1983;309:5236. Viguelloux J, Gausse G. Reflexions sur l'epidemiologie du cholera en Afnque occidentale. Med Trop 1971;31:678-84. Mandara MP, Mhalu FS. Cholera control in an inaccessible district in Tanzania: importance of temporary rural centers. Med J Zambia 1980; 15:10-13. Mhalu FS, Mtango FDE, Msengi AE. Hospital outbreak of cholera transmitted through close person-to-person contact. Lancet 1984:2:82-4. Cliff JL, Zinken P, Martelli A. A hospital outbreak of cholera in Maputo, Mozambique. Trans R Soc Trop Med Hyg 198630:473-6.

Epidemic cholera in West Africa: the role of food handling and high-risk foods.

During an epidemic of cholera in Guinea, West Africa, in 1986, the authors conducted two studies of risk factors for transmission. In the capital city...
634KB Sizes 0 Downloads 0 Views