Kaplan–Meier estimates for median time until viral RNA was undetectable (<5 copies per reaction) were determined using right censoring at the last positive sample day, and compared for cases who took timely Oseltamivir versus late or no Oseltamivir by Log Rank (Mantel–Cox) test. Continuous variables are presented as median and interquartile ranges and compared using Rank sum test. Undetectable viral RNA levels were assigned a value of one to facilitate Log 10 transformation. Chi-squared or Fisher's exact test were used for proportions. All statistical tests were 2 sided, and probability less than 0.05 was considered significant. Univariate and multivariate

logistic regression was performed to determine factors

associated with A(H1N1)pdm09 infection among contacts. Generalized MEK inhibitor estimating equations were used to account for household clustering in the logistic regression model. Predictor variables included the age and sex of the contact and of the index case, number of people in the household and index case peak viral load, sum of daily scores for symptoms and antiviral treatment. Variables with a univariate P value <0.10 were included in multivariate analysis. The Box–Tidwell test was used to assess selleck the assumption of linearity. 5 and 6 Index cases were detected in 20 (7.4%) of 270 households (Table 1). Two households had two separate index case episodes resulting in 22 index cases. The second episode was excluded from analysis of transmission. The households contained 81 people including the 22 index cases with the remaining 59 classified as contacts. Households comprising four people were significantly more common than amongst all 270 cohort

Immune system households (p = 0.009). Accordingly, most households comprised nuclear families with similar numbers of mothers, sons and daughters whereas some households lacked fathers. 25% of sons and daughters were older than 15 years. The median age of people in index case households was 23.3 years (IQR 12.2–39.3) with significantly fewer in the youngest and oldest age categories compared to all 270 households in the cohort. Pre-pandemic blood was collected from 69 (85%) of the index case household members ( Table S1). HI titres against A(H1N1)pdm09-like virus were <10 in all but one who had a titre of 20 and was not infected. None reported ever having received influenza vaccine. Eleven of 59 contacts were infected, giving a household secondary infection risk (SIR) of 18.6% (95%CI 10.7–30.4%). The secondary cases were from eight (40%) of the index case households. Five households had one secondary case, three households had two and twelve households had none. Six of the secondary cases were symptomatic giving a household secondary confirmed influenza illness risk of 10.2% (95%CI 4.8–20.5%). Five were asymptomatic, representing 45% of secondary infections.