The eltor cholera epidemic in Dhaka in 1974 and 1975*

Bulletin ofthe World Health Organization, 61(4): 653 - 659(1983) World Health Organization 1983 The eltor cholera epidemic in Dhaka in 1974 and 1975* MOSLEM UDDIN KHAN,' MOHAMMAD SHAHIDULLAH,2 WASEQUE UDDIN AHMED,3 DoMINICA PURIFICATION,4 & MAKSUD ALI KHAN5 Surveillance of hospitalized cholera cases from 1970 to 1977 in Dhaka, a matched control study in 1974, and a neighbourhood control study in 1975 were carried out and show a change from classical cholera to the eltor biotype during this period. Of all the hospitalized cholera cases, 9.1% in 1972 and 99.9% in 1973 were due to the eltor biotype. In 1974 and 1975 the distribution of eltor cholera cases in the city was uniform, except for areas with modern sanitation whose residents were spared. The incidence rates of cholera per 1000 infants (under the age of 1 year) were 1.16 and 0.93for 1974 and 1975, respectively. On the whole, children below 10 years andfemales between 15 and 44 years ofage were the ones most affected with eltor cholera. Higher rates of diarrhoea and hospitalization were noted among the contacts with cholera cases, compared with non-cholera controls. Contracting cholera was significantly associated with eating in places away from home, especially at charitable feeding centres. In India, the classical Vibrio cholerae biotype cholerae was replaced by biotype eltor by 1966 (1). In Bangladesh, however, the eltor biotype first appeared in Chittagong in 1963 and in Dhaka in 1964; it then disappeared, but returned in 1968 (2, 3). In 1972 and 1973, there was a shift in Vibrio cholerae from the classical Inaba serotype to the eltor Ogawa serotype. By 1973 all cholera cases were due to the eltor biotype, and the biggest eltor cholera epidemic ever recorded in Dhaka occurred in 1974. After the war of liberation in December 1971, a number of unemployed and homeless rural people flocked to Dhaka for government assistance in 1972 and 1973. Heavy floods and a food crisis affected the country during 1974 (4, 5). The homeless and jobless people constructed thousands of huts near the mar- kets, railway stations, outskirts of the city, industrial parks and on unused government land. Their maxi- mum concentration was in Ramna and Moham- madpur police zones. These settlements had no waste disposal facilities and the few handpump tubewells available were not enough to provide safe drinking water. People were therefore forced to use water from nearby rivers, ponds and canals, as well as from the city supply and the handpump tubewells, for bathing, washing, and drinking. The vehicles of cholera transmission in urban epi- demics in Bahrain, Italy, Malaysia, the Philippines, * From the International Centre for Diarrhoeal Disease Re- search, GPO Box 128, Dhaka 2, Bangladesh. 'Scientist, Disease Transmission Working Group. 2 Field Research Officer. 3Senior Research Officer. 4Health Assistant. 5Senior Field Assistant. and Portugal have been identified (6-10). But although several studies (11-16) have been conducted in Dhaka, few specific suggestions have been made to define the factors relating to the occurrence of cholera epidemics in this urban area. During the epidemic of 1974, a study was planned in order to (1) characterize the epidemiology of this first eltor cholera epidemic in Dhaka city, and (2) carry out a matched control study using hospitalized cholera and non-cholera diarrhoea cases in order to look for factors favouring the trans- mission of cholera. Since eating in wayside places was identified as an important factor in 1974, a study of the foods responsible and of food stalls connected with cholera outbreaks, based on cases and neigh- bourhood healthy controls, was carried out in 1975. MATERIALS AND METHODS The former Cholera Research Laboratory (now the International Centre for Diarrhoeal Disease Re- search, Bangladesh) has been the main centre for treatment of diarrhoea cases in Dhaka. This research centre with its hospital serves the entire city, the sub- urbs and surrounding rural areas. Patients with diar- rhoea from other hospitals are also referred to this hospital. For the epidemiological study, the popu- lation of Dhaka city was given by the 1974 census (17). But in the absence of figures for the population by age (single years), population figures based on estimated percentages of children aged 0, 1, 2, 3 and 4 years old in the total population of Dhaka (urban districts) were used. All hospitalized, confirmed cholera cases from 1970 to 1973 were analysed for 4323 -653- M. U. KHAN ET AL. serotype and biotype and for comparing seasonal differences. The epidemic was defined by analysing all confirmed and hospitalized eltor cholera cases from the urban area for the years 1974 and 1975, with matched controls employing a randomized sample. All patients seeking treatment, on arrival, had either stool or rectal swabs taken for culture on Monsur's medium and bile peptone broth (18). On receipt of the results on the following morning, a team of trained field-workers interviewed these patients because to be included in the study, they had to have a family, had to be staying in the city, and had to be cooperative. A total of 203 cholera cases were selected (alternately with controls) in 1974. The information sought in- cluded the patient's name, age, sex, place of resi- dence, number of family members, sources of water, sanitary facilities, history of eating in places away from home within 5 days of the onset of cholera, and history of any diarrhoea occurring in family members within 5 days of the onset of illness in the index case. Identical data were collected from 203 patients hospitalized with non-cholera diarrhoea. These con- trols were matched according to place of residence, water and sanitation facilities and age groups (0-4, 5-9, 10-14, and over 14 years). The first non-cholera case matching the study case was taken as the control. Both cases and controls were selected from the city areas only. During the 1975 epidemic, 168 cholera cases and 168 neighbourhood controls were selected alternately and interviewed. The same criteria and study design were used as in 1974. The type of food consumed away from home and the type of eating place fre- quented were also recorded. For significance testing, the x2 test was used. RESULTS Table 1 shows the change from the classical Inaba serotype to the classical Ogawa serotype between 1970 and 1972, and that in 1973 the eltor Ogawa serotype displaced all other types. The highest number of cholera cases recorded by the Cholera Research Laboratory was in 1974 (incidence of 1.73 per 1000 population); this outbreak was caused mainly by V. cholerae eltor Ogawa. An epidemic of lesser inten- sity due to both eltor Inaba and Ogawa serotypes occurred in 1975 (incidence of 1.37 per 1000). The monthly variations in the frequency of hospitalized cholera cases in 1971, 1974 and 1975 are shown in Fig. 1. The distribution for 1971 is typical of classical cholera and shows a small rise in May/June, and a post-monsoon peak in November. In 1974, the distribution of eltor cases reached a peak in October, about 6 weeks earlier than with classical cholera. The usual fall in the number of classical cholera cases to ,o a0 0 0 .5 t 10 oZ IL OA c M 'a r_ D_ Ct 45 o----a 1971 _- 1974 1975 ,i Jon. Feb. Mar. Apr. May. Jun. Jul. Aug. Sep. Oct. Nov. Dec. Fig. 1. Percentage incidence of hospitalized cholera cases in 1971, 1974 and 1975. Table 1. Change from classical cholera to eltor cholera (hospitalized cases), 1970-75 Year Classical Classical Eltor Eltor Total %Inaba Ogawa Inaba Ogawa eltor 1970 2792 38 - 138 2968 4.6 1971 1042 1426 - 185 2653 7.0 1972 40 598 1 63 702 9.1 1973 1 - 10 1265 1276 99.9 1974 - 2 552 4496 5050 100.0 1975 - - 1827 1530 3357 100.0 654 I ELTOR CHOLERA EPIDEMIC IN DHAKA during July and August and the high peak between November and December were not observed during the eltor epidemics (13). This type of seasonal variation has continued in Dhaka city even after 1974 and 1975. Within the municipal areas the incidence of hospitalized cholera cases was 1.73 per 1000 popu- lation in 1974 and 1.37 in 1975, except in Ramna dis- trict where the rate was reduced from 1.75 in 1974 to 0.88 in 1975, and in Mohammadpur district from 1.37 in 1974 to 0.81 in 1975. This significant reduction (P < 0.001) occurred after the demolition of most of the slums in these districts. There were no cholera cases, however, in both these years among persons from the higher socioeconomic groups who were living in houses with a clean water supply and sewage systems. A total of 993 cholera cases occurred in children under 5 years old in 1974 and 1975. Table 2 shows the distribution of cholera cases by age (single years till 4) per 1000 population. For infants under the age of one year, the incidence rate per 1000 was 1.16 in 1974 and 0.93 in 1975. The rates in children aged 1 year were 1.95 and 1.75 respectively for 1974 and 1975, and they then increased with increase in age. The distribution of hospitalized eltor cholera cases (rate per 10 000 population) for 1974 and 1975 is shown for males and females and for all ages in Fig. 2. The pattern of distribution was similar in both years, during which a total of 4131 cholera cases from the Dhaka municipal areas were treated. In children aged 14 years and under, the rate for males was higher than that for females (P < 0.05). Between the ages of 15 and 44 years, females had a much higher rate of cholera than males (P < 0.001). This type of distri- bution by age and sex, was found with classical cholera also. Nearly 49%o of all the cases occurred in Table 2. Cholera in children in Dhaka city, 1974_75a 1974 1975 Age (years) Cases Rate/ 1 000 Cases Rate/ 1 000 0 35 1.16 28 0.93 1 69 1.95 62 1.75 2 107 2.75 113 2.90 3 166 3.66 113 2.49 4 179 4.33 121 2.93 ,_ 5 1750 0.91 1389 0.72 Total 2305 1.73 1826 1.37 a Population by single years was estimated from the percent- age of Dhaka district urban population. 1974 c 0 cx 35 o 30 0 00oO '- 20 w 10 c 35 .2 o 30 0.Z02 0. so 20 0 z 0~ 0 'a .-. Male *--o Female 1975 *-4 Male .--o Female k 0--0 Yrt5 5-9 10-14 15-1 20-24 2542 30-34 35-39 40-44 45-49 50-54 55-59 60+ Fig. 2. Rate of hospitalized cholera cases per 10 000 population, by age group and sex, during 1974 and 1975. children up to the age of 9 years and the highest rate was in children 5-9 years old. The case-control studies showed the extent to which the contacts of cholera patients and of control subjects had diarrhoea (Table 3). In both 1974 and 1975, there were higher rates of diarrhoea and hospitalization among the contacts of cholera cases, which might explain the occurrence of multiple cases in cholera-affected families. The differences of the rates of diarrhoea and hospitalization between con- tacts of cholera patients and of control subjects were highly significant (P < 0.001) in both years. Table 3. The eltor cholera epidemic in 1974 and 1975, showing diarrhoea occurring among family contacts of cases and controls No. of No. of No. of No. of families members diarrhoea hospitalized cases cases 1974: cases 203 1388 214 69 controls 203 1301 81 0 1975: cases 168 1149 70 24 controls 168 1077 20 3 os w s * s . s v v w 655 M. U. KHAN ET AL. The effect of eating in places away from home by cholera cases, within 5 days of the onset of illness, and by matched controls is shown in Table 4. There were 203 paired samples for this purpose in 1974. Those who ate away from home contracted cholera signifi- cantly more often (X2 = 44.3, P < 0.001). In all age groups, the relative risks were high and the differ- ences between those who ate and did not eat in places outside the home were statistically significant, except for persons in the 0-4-year age group who do not usually go to restaurants or food stalls for meals. Table 4. Matched pair analysis of cases and matched controls, by history of eating in places outside the home during the 5 days prior to the onset of cholera in 1974 Cases Ate out Did not Total eat out Controls: ate out 7 11 18 did not 73 112 185 eat out Total 80 123 203 The effect of eating out by cholera cases and healthy controls from the same neighbourhood in 1975 is shown in Table 5. Here too, the frequency of eating away from home by cases was significantly higher than among the controls (P < 0.01). The types of eating places frequented by cholera cases and controls outside their homes are shown in Table 6. Overall, 52.3% of cases and 29.60o of con- trols ate away from home. In instances of charity feeding, or meals taken at roadside food stalls or with families where there was a case of diarrhoea (which Table 5. Matched pair analysis of cases and matched controls, by history of eating in places outside the home during the 5 days prior to the onset of cholera in 1975 Cases Ate out Did not Total eat out Controls: ate out 113 55 168 did not 92 76 168 Total 205 131 336 might have been cholera), the differences compared with the controls were quite significant. Eating at charitable feeding centres was significantly associated with contracting cholera (P < 0.001). DISCUSSION This study reports the displacement of classical Vibrio cholerae by the eltor biotype in Dhaka. Although the classical biotype had disappeared earlier from other countries, it was still predominant (91%) in Dhaka until 1972. The epidemiology of eltor cholera differs in some respects from that of classical cholera. The peak inci- dence of eltor cholera, for example, has been reported to occur over one month earlier than classical cholera (2, 3). Though the peaks of eltor epidemics in the present study were still close to those of classical cholera (in October/November) in Dhaka, small peaks in April/May during the monsoon were noted during the eltor outbreaks. Other workers (16) reported a concentration of classical cholera cases in small urban communities, but the present study shows that eltor cholera was widespread in the urban areas of Dhaka in 1974 and 1975. Those in the highest socioeconomic groups were, however, always spared. The decrease in attack rates in Ramna and Moham- madpur zones during 1975 was probably due to demolition of most of the slums and open latrines, as a result of which a part of the population at risk was reduced. Many authors have found that cholera was a dis- ease of young adult males (19). The findings reported here and also in our previous studies (14, 20, 21), however, show that cholera in Dhaka is mainly a dis- ease of children up to the age of 9 years (49% of all cases). In 1974 and 1975, the rates in infants (under 1 year) were 67% of the average, and in children aged 1-4 years they were much higher than the total case rates. This shows the high susceptibility of infants and young children to cholera. We are not sure, however, whether the higher rates in children are related to low vibriocidal titres (22), or frequent exposure to infected surface water while playing. Although adult females in Bangladesh usually avoid having to go to hospital, the rate per 1000 population of hospitalized female cases (aged 15-44 years) was higher (34.9) than the rate for adult males (13.3; P = < 0.001). The higher incidence among women may be because they usually nurse and clean the sick and launder their soiled clothes, and also wash the faeces from children after defecation. The occurrence of diarrhoea was more frequent among persons who had contacts with cholera cases (within 5 days of hospitalization of the index case) 656 ELTOR CHOLERA EPIDEMIC IN DHAKA Table 6. Types of eating places outside the home that were frequented by cholera cases and controls, 1974-75 Cholera cases Controls Eating X2 P valuesplaces Ate out Ate out Ate out Ate out (No.) (%) (No.) (%) Ordinary 46 12.4 37 10.0 0.9 > 0.5 restaurants Charitable 33 8.9 3 0.8 24.5 < 0.001feeding centres Roadside 85 22.9 59 15.9 4.9 < 0.05 Rehatives' 1 7 4.6 8 2.1 2.6 < 0.10 Family with a 13 3.5 3 0.8 5.2 < 0.01 case Total: ate out 194 52.3 110 29.6 38.4 < 0.001 Total: did not eat out 177 47.7 261 70.4 than among non-cholera controls. Many of these diarrhoea cases could have been cholera, multiple cases of cholera in some families being the result of contamination of food or water. During 1974 and 1975, free meals were distributed by both voluntary and government agencies from special feeding centres (23, 24). As the price of food in Dhaka was high, thousands of people depended on these charitable feeding centres. The association be- tween consumption of these meals and contracting cholera might be due to unhygienic preparation and handling of food or contamination of the cooked food or drink with Vibrio cholerae during prolonged storage before serving. The attack rate was higher for adult males who ate these meals than for adult females, but exactly how many (of all who ate) contracted cholera could not be ascertained. This finding was consistent in both years. It is a custom for people in Bangladesh to drink a glass of water after every meal or snack, but in our studies it was not possible to ascertain whether it was the food or the water that was the vehicle of trans- mission of Vibrio cholerae. It is suspected that multi- plication of Vibrio cholerae can take place in either food or water before consumption. Several factors may be responsible for the spread of urban cholera. This study has shown that eating away from home during floods or in a food crisis, especially in places where free meals were served, was an impor- tant determinant of the cholera epidemics in Dhaka in 1974 and 1975. ACKNOWLEDGEMENTS We thank Dr W. B. Greenough III, Director of the International Centre for Diarrhoeal Disease Research, Bangladesh, for reviewing this manuscript, and Dr George T. Curlin and the late Dr W. F. Verwey for supporting the study. We are also grateful for the suggestions of Dr Tom Butler (Program Head, ICDDR,B), the editorial assistance of Mr Richard Doan (Yale University), and the secretarial assistance of Mrs Prabashi Mahmud (Disease Transmission Division). RtSUMt EPIDEMIE DE CHOLERA ELTOR A DHAKA EN 1974-1975 Tous les cas de cholera classique 6taient remplac6s en 1974 par des cas de cholera eltor et c'est en 1974 que s'est produite la plus grande epidemie de cholera eltor. Bien que plusieurs etudes aient et faites, peu de facteurs specifiques interessant la transmission du cholera urbain avaient et definis. Au cours de cette premiere epidemie de 657 658 M. U. KHAN ET AL. cholera eltor en 1974, nous avons donc cherche 1) a caracte- riser l'epidemiologie de l'6pid6mie de chol6ra eltor dans la ville de Dhaka et 2) a faire une 6tude avec t6moins appari6s en nous servant des cas hospitalises de diarrh&e cholerique ou non cholerique pour degager les facteurs de transmission du chol6ra. En 1975, nous avons recherche quels types d'aliments avaient ete consommes et dans quels genres d'6tablissements par les malades et par les temoins sains. Les patients 6tudi6s etaient tous des cas pris au laboratoire de recherche sur le cholera (ICDDR,B). Nous avons pris comme d6nominateur la population urbaine recens6e a Dhaka en 1974. L'epidemie de cholera eltor a 6te caracte- ris&e sur la base d'une analyse de tous les cas hospitalises a Dhaka en 1974 et 1975. Des 6chantillons de selles ou des frottis rectaux ont e mis en culture sur milieu Monsur et bouillon biWiU peptone. Au total, 203 cas de cholera et 203 temoins apparies ont e 6tudies en 1974 et, en 1975, 168 cas de cholera et 168 temoins sains du voisinage. Les donn6es concernant les ann6es 1970 a 1973 font appa- raitre un passage du cholera Inaba classique au cholera Ogawa classique, puis au cholera Ogawa eltor. Pour la periode classique, l'incidence accusait un pic important apres la mousson. La saison de la mousson ne s'accom- pagnait pas de chol6ra classique. Le chol6ra eltor, par contre, a continue pendant la mousson et son incidence pendant la periode post-mousson atteignait son maximum environ six semaines avant les pics de cholera classique. L'incidence des cas hospitalis6s, pour 1000 habitants, etait de 1,73 en 1974 et de 1,37 en 1975. Les taux etaient nettement plus bas en 1975 dans les districts de Ramna et de Mohammadpur apres la destruction des camps de refugies. Les cas nouveaux 6taient largement diss6min6s dans la ville a l'exception des quartiers disposant d'installations sanitaires modernes, qui etaient epargnes. Les incidences les plus fortes etaient observ&es chez les enfants et les femmes adultes. II y a eu de multiples cas de diarrh&e parmi les contacts des cas de cholera mais peu parmi les contacts des diarrheiques t6moins. Les risques de contracter le chol6ra etaient nette- ment plus eleves parmi les gens qui avaient mange hors de chez eux. La consommation d'aliments A des 6ventaires exte- rieurs et plus particulierement dans des centres caritatifs s'accompagnait d'un risque accru de contracter le cholera. REFERENCES 1. MUKERJEE, S. & BASU, S. Cholera El Tor in India: effect on epidemiology of classical cholera. Tropical and geo- graphical medicine, 19: 138-143 (1967). 2. MCCORMACK, W. M. ET AL. Endemic cholera in rural East Pakistan. American journal of epidemiology, 89: 393 404(1969). 3. KHAN, M. ET AL. The changing pattern of cholera in East Pakistan: the appearance of El Tor Vibrio cholerae. Journal ofthe Pakistan Medical Association, 20: 43-48 (1970). 4. BANGLADESH WATER DEVELOPMENT BOARD. Flood forecasting and warning. In: prefaces of Annual report onfloods in Bangladesh, 1974 (No. 2-75), published in April 1975; and 1975 (No. 1-76), published in 1976. 5. Bangladesh Observer. Reports on flooding from 5 July to 31 August 1974, and editorials on flooding on 5, 10 and 25 July and I August 1974. 6. GUNN, R. A. ET AL. Bottle feeding as a risk factor for cholera in infants. Lancet, 2: 730-732 (1979). 7. BAINE, W. B. ET AL. Epidemiology of cholera in Italy in 1973. Lancet, 2: 1370-1374 (1974). 8. DuTT, A. K. ET AL. A shellfish-borne cholera outbreak in Malaysia. Transactions of the Royal Society of Tropical Medicine and Hygiene, 65: 815-818 (1971). 9. JOSEPH, P. R. ET AL. Studies of Cholera El Tor in the Philippines. 2. A Retrospective Investigationi of an Ex- plosive Outbreak in Bacolod City and Talisay, Novem- ber 1961. Bulletin of the World Health Organization, 33: 637-643 (1965). 10. BLAKE, P. A. ET AL. Cholera in Portugal, 1974. I. Modes of transmission. American journal of epi- demiology, 105: 337-343 (1977). 11. OSEASOHN, R. ET AL. Clinical and bacteriological find- ings among families of cholera patients. Lancet, 1: 340-342 (1966). 12. KHAN, A. Q. Role of carriers in the intrafamilial spread of cholera. Lancet, 1: 245-246 (1967). 13. KHAN, M. & MOSLEY, W. H. The significance of Shigella as a cause of diarrhoea in a low economic urban community in Dacca. East Pakistan medical journal, 12: 45-51 (1968). 14. MARTIN, A. R. ET AL. Epidemiologic analysis of endemic cholera in urban East Pakistan, 1964-1966. American journal of epidemiology, 89: 572-582 (1969). 15. KHAN, M. & MOSLEY, W. H. Contrasting epidemiologic patterns of diarrhoea and cholera in a semi-urban com- munity. Journal of the Pakistan Medical Association, 19: 380-385 (1969). 16. BART, K. J. ET AL. Seroepidemiologic studies during a simultaneous epidemic of infection with El Tor Ogawa and Classical Inaba Vibrio cholerae. Journal of infec- tious diseases-, 121: (suppl.) S17-S24 (1970). 17. BANGLADESH BUREAU OF STATISTICS. Bangladesh population census report, 1974. National volume, Dacca, 1974, p. 56; and Dacca district census report, 1974, pp. 1-25. 18. MONSUR, K. A. A highly selective gelatin-taurocholate- tellurite medium for the isolation of Vibrio cholerae. Transactions of the Royal Society of Tropical Medicine and Hygiene, 55: 440-442 (1961). 19. DIZON, J. J. ET AL. Studies of Cholera El Tor in the Philippines. 1. Characteristics of Cholera El Tor in Negros Occidental Province, November 1961 to Sep- tember 1962. Bulletin of the World Health Organiz- ation, 33: 627-636 (1965). 20. KHAN, M. ET AL. Ten years' review of the age and sex of cholera patients, Dacca, Bangladesh. Cholera Research Laboratory, Scientific Report No. 14 (1978). ELTOR CHOLERA EPIDEMIC IN DHAKA 659 21. MERSON, M. H. ET AL. Epidemiology of cholera and enterotoxigenic Escherichia coli diarrhoea. In: Ouchter- lony, 0. & Holmgren J., ed., Cholera and related diarrhoeas. Stockholm, 43rd Nobel Symposium, 1978, p.34. 22. MOSLEY, W. H. ET AL. A serological survey for cholera antibodies in rural East Pakistan. 1. The distribution of antibody in the control population of a cholera-vaccine field-trial area and the relation of antibody titre to the pattern of endemic cholera. Bulletin of the World Health Organization, 38: 327-334 (1968). 23. ALAMGIR, M. Bangladesh: a case of below-poverty- level-equilibrium trap. Dacca, Bangladesh Institute of Development Studies, 1978, p. 2. 24. SATrAR, M. A. Speech published in Bangladesh Ob- server and the Ittefaq (daily newspapers), 18 October 1981, p. 1 (1981).