Household Transmission of Vibrio cholerae in Bangladesh

Abstract

Background: Vibrio cholerae infections cluster in households. This study's objective was to quantify the relative contribution of direct, within-household exposure (for example, via contamination of household food, water, or surfaces) to endemic cholera transmission. Quantifying the relative contribution of direct exposure is important for planning effective prevention and control measures.

Methodology/principal findings: Symptom histories and multiple blood and fecal specimens were prospectively collected from household members of hospital-ascertained cholera cases in Bangladesh from 2001-2006. We estimated the probabilities of cholera transmission through 1) direct exposure within the household and 2) contact with community-based sources of infection. The natural history of cholera infection and covariate effects on transmission were considered. Significant direct transmission (p-value<0.0001) occurred among 1414 members of 364 households. Fecal shedding of O1 El Tor Ogawa was associated with a 4.9% (95% confidence interval: 0.9%-22.8%) risk of infection among household contacts through direct exposure during an 11-day infectious period (mean length). The estimated 11-day risk of O1 El Tor Ogawa infection through exposure to community-based sources was 2.5% (0.8%-8.0%). The corresponding estimated risks for O1 El Tor Inaba and O139 infection were 3.7% (0.7%-16.6%) and 8.2% (2.1%-27.1%) through direct exposure, and 3.4% (1.7%-6.7%) and 2.0% (0.5%-7.3%) through community-based exposure. Children under 5 years-old were at elevated risk of infection. Limitations of the study may have led to an underestimation of the true risk of cholera infection. For instance, available covariate data may have incompletely characterized levels of pre-existing immunity to cholera infection. Transmission via direct exposure occurring outside of the household was not considered.

Conclusions: Direct exposure contributes substantially to endemic transmission of symptomatic cholera in an urban setting. We provide the first estimate of the transmissibility of endemic cholera within prospectively-followed members of households. The role of direct transmission must be considered when planning cholera control activities.

Figure 1. Survey data and specimen collection schedule for each study household, relative to the enrollment date of the household's index cholera infection (study day 1).

The “*” denotes the day on which stool/rectal swab specimens were only collected from the index cholera infections.

Figure 2. Schematic illustration of the transmission model.

The probability represents the daily risk that a susceptible contact (hollow figure) will subsequently develop cholera infection by serogroup-serotype v after exposure to household surfaces or water/food supplies contaminated by a member infected with and shedding v (black figure). The probability represents the susceptible contact's daily risk of cholera infection resulting from exposure to sources of infection located outside of his/her household. The corresponding epidemiologic summary measures for and are the household secondary attack rate, or , and the community probability of infection, or (see the text for parameter definitions).

Figure 3. The number of days between the illness onset dates of household primary and non-primary symptomatic Vibrio cholerae cases (all serogroup-serotypes).

Primary symptomatic cholera cases are defined as enrolled household members meeting the case definition for a symptomatic cholera case and whose symptom onset date was on or before that of the household's index infection. All other symptomatic cholera cases are classified as non-primary. The horizontal line represents the interquartile range (25^th through the 75^th percentile), with the median denoted by the vertical crossbar.