Do facemasks and/or hand hygiene prevent the transmission of influenza to household contacts?
Hand washing and facemasks decrease the transmission of influenza to household contacts if applied within 36 hours of symptom onset in the index patient.
To say that influenza awareness is heightened this season due to the H1N1 strain would be a gross understatement. Non-pharmacologic measures can be taken to prevent the spread of influenza such as hand-washing, gloves, surgical facemasks, and lifestyle education.
Many countries as part of their pandemic planning have facemasks, improved hand hygiene, cough etiquette, isolation measures and travel restrictions as their primary means to prevent an influenza pandemic. One preliminary study done by this research group showed a reduction in influenza spread among households with these simple, cost-free interventions, although compliance was variable. The purpose of the current CDC-funded cluster randomized trial was to see if household transmission could be reduced with hand hygiene or hand hygiene plus facemasks.
From January to September 2008, the investigators assessed 2750 potential index cases and randomly allocated 407 (245 Influenza A, 162 Influenza B) individuals (each representing a unique household) to one of the three intervention arms: hand-hygiene, face-mask and hand-hygiene, or control. After randomization a home-visit was scheduled within 12-hours to 2-days to initiate the intervention. The control group received education about the disease-specific importance of healthy diet and lifestyle. All family members in the hand-hygiene intervention group received instructions to wash their hands after bathroom use and following coughing or sneezing. They were provided specific liquid hand soap for the kitchen and every bathroom in the household, as well as small containers of alcohol based hand cleanser with instructions to use when returning home or after touching potentially contaminated surfaces. The hand-hygiene group received the same soaps and cleansers in addition to 50 surgical facemasks for each family member with instruction on appropriate wear and disposal methods. All subjects were asked to keep an illness log for one-week following enrollment to assess subjective symptoms and self-reported adherence. Soap samples were collected and measured to assess compliance rates. Approximately 25% of subjects in each arm of the study received oseltamivir.
The overall transmission rate was not decreased with either of the interventions compared to control. However, subgroup analysis planned before the study evaluated the impact of timing. This analysis demonstrated that if the intervention was started before 36-hours of index patient symptom onset than influenza transmission was significantly reduced (Table). The investigators used reverse-transcription polymerase chain reaction (RT-PCR) to define infections so we know that secondary cases truly acquired influenza. In addition, they reported clinically implicated influenza as defined by at least two of the following signs and symptoms: temperature >37.80C, cough, headache, sore throat or myalgia.
Table Secondary Attack Ratio of RT-PCR Confirmed Influenza virus Infection and Clinical Influenza if Intervention Started within 36-hours
RT-PCR = reverse-transcription polymerase chain reaction
SAR = secondary attack ratio (95% CI)
HH = hand hygiene group
FM = face mask
Facemask plus hand hygiene also decreased the odds ratio of RT-PCR confirmed influenza when the intervention was applied within 36-hours OR=0.33 (95% CI 0.13-0.87). There was no difference found between facemask and hand hygiene and hand hygiene alone OR=0.72 (95% CI 0.21-2.48).
Adherence to the intervention varied among groups. Across groups 75% of household contacts used the liquid soap, but only 25% used the alcohol cleansers. Facemask compliance was even worse with only 49% of the index patients and 26% of household members reporting regular use. A median of 9 facemasks were used by index cases compared with 4 facemasks per family member.
This study did not specifically address the H1N1 influenza. Although our learning curve for this novel virus continues, we expect it has a similar mode of transmission. The current study was conducted in Hong Kong households and initiated from outpatient office settings. Patients presenting to the emergency department may differ in important, unmeasured ways from outpatient settings limiting external validity.
Additionally, in Hong Kong the patient population lived in apartments with an average size of 700 square feet and a median household size of 4 people. In contrast, two-thirds of American households are single unit detached or semi-attached units with only one-third living multi-unit (apartment) settings. The average American new single-family home is 2,349 square feet. American apartments are also larger with an average size of 1,043 square feet. The average American family size in the latest census was 3.19. This results in 0.1 and 0.3 persons per 100 square feet for homes and apartments respectively. Since crowding can increase influenza transmission it is possible that in less crowded American homes non-pharmacologic interventions may not be as effective.
Regardless of the differences in household size promoting good hand washing early in the disease course is a reasonable and inexpensive non-pharmacologic method to decrease transmission of influenza to household contacts. Wearing a facemask can also prevent influenza transmission, but adherence is problematic. In summary, inability to begin such measures within 36 hours of symptom onset for the index case and lack of compliance with the recommended steps substantially limit the effectiveness of hand hygiene and facemasks to prevent transmission.
Dr. Milne is an adjunct professor in the division of emergency medicine at the University of Western Ontario and is a BEEM Faculty Member, McMaster University