Household Transmission of Pandemic (H1N1) 2009 Virus, Taiwan

During August–November 2009, to investigate disease transmission within households in Taiwan, we recruited 87 pandemic (H1N1) 2009 patients and their household members. Overall, pandemic (H1N1) 2009 virus was transmitted to 60 (27%) of 223 household contacts. Transmission was 4× higher to children than to adults (61% vs. 15%; p<0.001).


P andemic (H1N1) 2009 was fi rst identifi ed in 2 southern
California counties in April 2009 (1), and the World Health Organization declared a global pandemic on June 11, 2009 (2). In Taiwan, the government suggested that persons with pandemic (H1N1) 2009 remain home until 24 hours after they were symptom free (3). In some infl uenza epidemics, ≈50% of households have >1 members who become infected (4). Further investigation into the transmission of pandemic (H1N1) 2009 virus among household members is needed to help control and prevent additional infections. We investigated the transmission of pandemic (H1N1) 2009 virus and clinical outcomes of infection within households of persons with laboratoryconfi rmed infection.

The Study
During August-November 2009, we enrolled patients at the National Taiwan University Hospital who were infected with pandemic (H1N1) 2009 virus and their household members. The following samples were obtained from patients with clinical signs and symptoms suggestive of pandemic (H1N1) 2009 infection who visited the emergency department, outpatient clinics, or inpatient wards: nasopharyngeal swab specimen for rapid infl uenza antigen testing (QuickVue A+B test; Quidel, San Diego, CA, USA), throat swab specimen for virus isolation and novel subtype H1N1 reverse transcription PCR (RT-PCR), and blood specimen for serum hemagglutination inhibition (HI) assays. Laboratory-confi rmed pandemic (H1N1) 2009 infection was defi ned in 3 ways: 1) isolation of infl uenza A virus, followed by positive RT-PCR result for pandemic (H1N1) 2009 virus; 2) positive rapid infl uenza A test result, followed by positive RT-PCR result for pandemic (H1N1) 2009 virus; or 3) pandemic (H1N1) 2009 virus HI titer >40. None of the participants had received an infl uenza subtype H1N1 vaccine before this study.
Persons with laboratory-confi rmed pandemic (H1N1) 2009 and their household members were sent a letter and/ or received a telephone call inviting them to participate ( Figure). After they accepted the invitation, we collected their case report forms, which contained data regarding the source of infection, fi nal diagnosis, clinical manifestations, and course of the disease. The index patient was defi ned as the fi rst person in a household to have laboratory-confi rmed pandemic (H1N1) 2009 (body temperature >38.0°C and/ or cough and/or sore throat). All enrolled index patients and their household members provided blood samples for further HI assays. A mean of 45 days (SD 26, median 36, range 12-107 days) elapsed between the fi rst day of illness in the index patient and household investigations, including blood sampling. The household transmission infections were traced to schools, 5 (6%) infections were traced to daycare centers or other child care situation (when 1 babysitter cared for a few children), and 1 (1%) infection was traced to a summer camp.
As shown in Table 1, the mean ages of the index patients and their household contacts were 10.6 and 33.8 years, respectively; only 6 (7%) of the 87 index patients were adults. Households contained a mean of 1.9 children (SD 0.8, median 2, range 0-4 children).
Pandemic (H1N1) 2009 virus was transmitted to 60 (27%) of the 223 household contacts. The virus was transmitted to 35 (63%) of 56 child-aged siblings (but not to 1 cousin), to none of 5 adult-aged siblings, to 20 (14%) of 138 parents, to 4 (22%) of 18 grandparents, and to 1 (20%) of 5 aunts and uncles. Percentage of transmission among the different groups of household contacts differed signifi cantly: the virus was transmitted to 35 (61%) of the 57 children and to 25 (15%) of the 166 adults (p<0.01 by χ 2 test). However, percentage of transmission among different adult groups did not differ signifi cantly (p = 0.86 by χ 2 test). Mean interval between the onset of illness in the index patient and household members was 3.3 days (SD 2.6, median 3, range 1-6 days).
Of the 147 patients with pandemic (H1N1) 2009, 119 (81%) received a diagnosis of infl uenza-like illness; 10% received a diagnosis of upper respiratory tract infection; 3% each received a diagnosis of bronchitis, bronchopneumonia, asthma, or acute gastroenteritis; and 2% received a diagnosis of pneumonia. Of the 147 patients (all children), 10 (7%) were hospitalized and discharged without sequelae. Seventy-seven (89%) of the 87 index patients and 29 (48%) of the 60 household members received oseltamivir. Table 2 shows attack rates and odds ratios for pandemic (H1N1) 2009 virus infection among the 223 household contacts by patient characteristics (sex and age) and signs and symptoms. Age <18 years, fever, cough, sore throat, rhinorrhea, myalgia, and malaise were signifi cantly associated with pandemic (H1N1) 2009 infection, but age <18 years, fever, and cough most signifi cantly predicted the transmission of pandemic (H1N1) 2009 virus in multivariate analysis with a multiple logistic regression model. However, we did not fi nd a signifi cant relationship between index patient characteristics, specifi c symptoms, lower respiratory tract infection, or the need for hospitalization and the rate of household transmission of pandemic (H1N1) 2009 virus.

Conclusions
We found children to be >4× more susceptible than adults to the secondary transmission of pandemic (H1N1) 2009 virus within households (61% vs. 15%). Furthermore, 93% of our index patients were children, and for ≈60% of them, the source of exposure to the virus was a school or daycare center. Thus, children play major roles in the introduction and spread of infl uenza within families. Vaccination and other measures will prevent susceptible children from becoming infected and reduce infl uenza virus transmission among families and communities. This study has limitations, however, for example, the potential for nonresponse bias and possible preferential recruitment of families with sick children as index patients. Thus, adults may be relatively underrepresented as index patients in this study. Also, some adults may be less likely to go to the hospital with infl uenza-like symptoms.
In our study, the secondary attack rate in households was 27%, which is similar to rates in studies by Komiya et (5)(6)(7)(8)(9). The secondary attack rate found in this study may have been relatively high because, without a vaccine against pandemic (H1N1) 2009, there were more susceptible children in the households and because most index patients were children who may shed virus for a longer period (10). Our fi ndings show the key role that children play in introducing and spreading pandemic (H1N1) 2009 virus within households. Public health measures, such as vaccination and community health education, can prevent infections among children and help reduce virus transmission among families and the larger community.