Stochastic modelling of intra-household transmission of hepatitis C virus: evidence for substantial non-sexual infection.

OBJECTIVE: To simulate the probability of HCV transmission from an HCV seropositive index patient to susceptible household contacts through non-sexual exposures. METHODS: A modified Reed-Frost stochastic simulation model was used to assess the probability of HCV transmission from an HCV seropositive index patient to susceptible household contacts through non-sexual exposures. This mathematical model does not require the specification of infection onset times for individual, nor is it necessary to identify the chains of household infections. Therefore, this model can be used with serologic data on detected asymptomatic infections. The HCV serological data on 341 non-sexual household contacts of 86 HCV seropositive index patients were used in this simulation study. The frequency distribution of HCV infection of susceptibles for each household size of 4-8 initial susceptibles was calculated. A maximum likelihood procedure was used to estimate the non-sexual household transmission parameter for HCV infection for the range of household sizes studied and was used in 1000 stochastic iterations. The goodness-of-fit test was carried out to compare the observed proportions of households where HCV transmission occurred to one or more initial susceptible with mean expected simulated proportions of such households with varying sizes ranging from 4 to 8 initial susceptibles. RESULTS: The maximum likelihood estimates (90% probability interval (PI)) of binomial probability of HCV transmission within households with varying number of initial susceptible non-sexual household contacts ranged from 0.248 (90%PI: 0.031, 0.560) to 0.164 (90%PI: 0.011, 0.440) for household size of 4 and 8 respectively. The χ(2) goodness-of-fit test of observed and mean expected simulated proportions of households wherein at least one of the susceptibles was infected revealed good fit for households of all sizes examined (P ≥ 0.96). In a household, the probability of HCV transmission from the index HCV seropositive patient to susceptible via non-sexual contacts tended to decrease linearly as the household size increased from four to seven. CONCLUSION: Intra-household HCV transmission through non-sexual contacts may have substantial impact on HCV transmission and needs to be considered in an HCV control program.