A. Khan, S.F. Hoch, A. Merchant and J.B. McCormick.

Department of Community Health Sciences and Pathology The Aga Khan University, Karachi

BACKGROUND: More cases of cholera were reported in 1991 in more countries than ever before. In 1992 emergence of the new strain of Vibrio cholera, serogroup 0139, further increased epidemic potential of the disease since it was able to infect persons immune to previous strains. Late in 1993 we found that this new strain had become endemic in Karachi.

SUBJECTS AND METHODS: We sought to establish the epidemiology and the risk factors for infection with strains of V. cholera in Karachi, by conducting a case control study. A case-patient with cholera was a patient of either sex, above 5 years of age, seeking treatment at one of the 2 identified tertiary - care hospitals, with a history of 5 or more loose watery stools in the previous 24 hours, who also had V. cholera isolated from stool. We randomly selected 116/263 cases of cholera thus identified between July and September, 1994. For each case-patient we recruited two neighborhood controls matched for age (within 5 years) and sex. We analyzed matched pairs for risk factors associated with infection with V. cholera serogroups 01 and 0139 both together and independently.

RESULTS: We studied 92 cases and 170 neighborhood controls. Cholera patients were more likely to drink unboiled water (Matched Odds Ratio [MOR] 4.2, 95% Confidence Interval [CI] 1.8-9.6), eat food that was not cooked in the home (MOR 2.4, CI 1.3-4.1) and eat food from a street vendor (MOR 5.0, CI 1.7-14). Multivariate analysis concluded that case-patients were more likely to drink unboiled water and eat from a street vendor (p<0.05). Eating chicken was an independent risk factor for V. cholera serogroup 0139 (MOR 13.0, CI 1.5-110) but not for serogroup 01 infection. Infection with serogroup 0139 was also associated with drinking unboiled water (MOR 4.0, CI 1.4-11) and eating food not cooked at home (MOR 3.3, CI 1.2-8.5). Though drinking unboiled water was significantly associated with infection with both strains, patients infected with seroguroup 01 were more likely to have done so. All ages were affected, but patients with serogroup 0139 were significantly older. Furthermore, the two epidemics were separate, the early part of the year featuring serogroup 01, followed by replacement with serogroup 0139.

CONCLUSIONS: Cholera is widespread in Karachi, where both serogroup 01 and 0139 strains circulate. Differences were detected in both the age group affected and the risk factors associated with infection in what were two separate epidemics. Though municipal water appears to be a major source of infection, a strong association with eating chicken in those patients infected with 0139, but not those infected with serogroup 01 suggests that the organisms may occupy different ecological niches.

BACK TO AKU-SYM 95

BACK TO RESEARCH INDEX.

BACK TO MAIN PAGE.