Breakout Group: Best Methods for Studying Aerosol Transmission November 4, 2010 and November 5, 2010 – Atlanta, GA “Understanding the Modes of Influenza Transmission” Workshop

Aerosol Transmission: Person-to-person transmission of influenza through the air by aerosols in the inspirable (inhalable) size range or smaller. Particles are small enough to be inhaled into the oronasopharynx and distally into the trachea and lung

What are the key questions/ gaps remaining in understanding the contribution of aerosol transmission to the spread of influenza among humans?

Issues Provided for Breakout Group Consideration - Level of documentation of influenza transmission via aerosol transmission - Influenza virus infectious dose for this route of transmission relative to other routes - Duration of infectivity from aerosolized infectious particles - Evidence on distance(s) over which transmission by this route may occur - Host factors that may increase/decrease transmission risk by aerosols - Intervention or exposures that may isolate the role of droplet nuclei in the spread of influenza - Impact of environmental factors (temperature, humidity, ambient UV, etc.) on influenza airborne transmission - How might changes in air exchange rates/ ventilation mitigate transmission via this route Breakout group felt that all of these were important!

Aerosol Transmission Breakout Group: Remaining Key Questions/ Gaps  Level of documentation of influenza transmission via aerosol transmission  All breakout group members felt that aerosol transmission, as defined earlier, plays a role  Key question is how big a role relative to contact and droplet spray  Why are other organisms (e.g., TB) able to spread via aerosol route over longer distances, while influenza seems limited to short range aerosol transmission?  TB is very different from influenza  Comparison with viruses capable of long-range airborne transmission, such as measles or varicella is probably more appropriate  Issue raised: TB is rare in the US, so TB transmission over short or long distances is easier to recognize than transmission of influenza, which is common, making it harder to clearly identify the source

Aerosol Transmission Breakout Group: Remaining Key Questions/ Gaps  How long do aerosolized particles remain viable?  Ability to detect and/or quantify viable/infectious particles in the air is uncertain  What factors influence infectivity of particles in the air?  How can we better measure exposures to viable/infective virus – example: current measuring instruments for airborne influenza (e.g., impingers) may damage the virus in an unpredictable manner  Dose-Infection relationships – many host and virus variables  Is one of the gaps we need to fill the clinical impact of virus particles of different sizes?  Data presented in meeting (Henle et al 1946) suggests that inhalation might cause more severe disease with systemic effects such as fever than intranasal deposition of virus.

Aerosol Transmission Breakout Group: Remaining Key Questions/ Gaps  Animal studies (e.g., ferrets) look at impact of viral factors but do not characterize host factors; role of host factors should be investigated with respect to aerosol transmission (and other routes of transmission) (temp, humidity also affect the host)  Lack of literature describing human-to-human transmission via sneezing (most is coughing). Should investigate mechanism by which aerosol particles are produced by humans as well as animals (ferrets, for example, don’t cough much).

Aerosol Transmission Breakout Group: Remaining Key Questions/ Gaps  In general, we also don’t know the contribution of different mechanisms to generate aerosols (e.g., sneezing, blowing nose, coughing, breathing)  Need to understand impact of variability in:  Donor  Recipient  Virus  Environment – weather, seasonality, temp, humidity  Social Dynamics – crowding, habits, etc

What are the best study designs and their pro’s/ con’s? What study designs would be best for understanding the contribution of aerosol transmission to the other transmission routes?

What are the best aerosol transmission study designs and their pro’s/ con’s? What study designs would be best for understanding the contribution of aerosol transmission to the other transmission routes? Thoughts to consider - What are the scientific approaches to answer key questions that will lead to a better understanding of the relative contribution or role of airborne transmission in the spread of influenza among people? - How should temperature, humidity and air exchange be considered in study designs? - Consider how a study could examine aerosol transmission, while evaluating other transmission routes (i.e. respirator study designed to examine the reduction in aerosol transmission, which could also assess the contribution of contact spread if masks reduce the number of times that people touch their mouth or nose).

General Issues  Need multidisciplinary study designs to capture contribution of aerosol transmission  All relevant partners (e.g., DOD)  Are intervention studies perhaps ideally suited to answer transmission questions? Well-designed efficacy studies?  Choice of study design somewhat dependent on outcome goal– the most valuable studies may be the ones that include outcomes that have intervention applications (i.e., effectiveness)  Existing animal models can (should) be used to answer transmission questions in humans

Challenges  Unable to discern contribution of aerosol transmission under conditions of widespread disease  Under these conditions, aerosol transmission may be attributed to other routes  Human studies involving experimental exposures or preventive interventions can be confounded by community exposures outside of the study conditions. If not controlled, these exposures can drown out effects of experimental conditions.  Within 1-2 meters, susceptible people are exposed to potentially infectious particles of a large size range  Difficult to tease out contribution of aerosol transmission vs. droplet spray in that situation

Aerosol Transmission Breakout Group: Study Design pro’s/ con’s  Animal studies  Should characterize the aerosols that are produced by animals and compare to those produced by humans (animal models need to be validated)  Concern raised that animal studies alone, no matter how compelling, won’t change clinical practice. Their relevance to humans will need to be convincingly validated!  May be well-suited to test pre-symptomatic or asymptomatic transmission  Should recognize and address potential confounders of inoculating with larger doses of virus than those that cause natural infection or use of laboratory virus strains instead of low passage virus from human infection  Need to account for variability in clinical presentation that different virus strains have (i.e., some elicit more GI symptoms and perhaps contaminated diarrhea) since these will permit different routes of transmission

Aerosol Transmission Breakout Group: Study Design pro’s/ con’s  Animal studies  An important advantage of animal models is that it is really possible to control the study conditions. For example, there is the potential for great control over manipulations, such as air disinfection of ducts between animals.  Experimental conditions in animal studies could be established to differentiate between droplet spray and aerosol transmission. For example, setting cages very far apart in a biobubble and at different heights can be a useful design overcoming the problem of differentiating between droplet spray and aerosol transmission. Studying infector and recipient animals placed at varying distances apart, with varying levels of ventilation was suggested for better understanding the impact of particle size.  It was suggested that animal models could be adapted to simulate clinical procedures in an effort to better understand the biology of aerosol-generating procedures. Some participants saw this as an especially high priority area for investigation.

Aerosol Transmission Breakout Group: Study Design pro’s/ con’s  Human challenge studies  Concern raised that these may be only be hypothesis- generating due to differences from natural infection Site of infection: Deep lung inhalation probably unethical; must use upper airway for infection Other issues – dose, strain of virus used in model  Should characterize patients with experimental infections induced via inoculation and compare to individuals infected naturally to validate that experimental and natural infection are similar or at least that differences are understood  Expensive, may be issues of power unless large studies are performed (given attack rate in subjects exposed to those with experimental influenza infections of ~ 20%)  Should consider and document air exchange dynamics in settings where subjects are exposed to those with experimental influenza. It may be necessary in some cases to limit the effect of air dilution in the study area in order to fully characterize (and perhaps facilitate) experimental transmission

Human Challenge Studies  Consider human challenge study with 2 rooms connected only via an air duct from one room to another (after the classic studies of Riley)  Consider human challenge study addressing using upper-room UV and fans to create good air mixing as an intervention. Would be expected to affect short-range aerosol transmission, but not droplet spray transmission.  Consider human challenge study with infected individual exposed to susceptibles wearing either (i) full face respirator (preventing droplet spray and aerosol exposure) or (ii) face shield (preventing only droplet spray exposure, superior for this purpose to face mask, which might be able to filter some aerosol) (less enthusiasm for this design)  Human challenge studies might be especially good for studying viral shedding and transmissibility in early and asymptomatic disease  Baseline immunity of study volunteers is an important consideration because it would be expected to profoundly affect results.

Human Challenge Studies  Ethical considerations include: Tendency for vulnerable populations to volunteer Compensation issues How and when do investigators communicate with public? Are investigators receptive to public input? Proper informed consent procedures and risk communication Integrate ethical evaluation into the study itself. Consider using a pre and post survey to better understand who volunteered, why they volunteered, what the experience was like, what they expected it to be like, whether they were they informed about risks, and whether information provided about risks was well understood

Real World Intervention / Exposure Assessment Studies  These types of studies were felt to be very desirable, because of their potential both to answer scientific questions and to have impact on public health prevention practices.  These studies might have the greatest chance to be informative about routes of transmission if they are well-controlled “efficacy” studies rather than studies of intervention effectiveness.  Main role of exposure assessment to validate impact of interventions  It is critical to account for and control for competing sources of infection from outside the study site when investigating aerosol transmission  Otherwise, competing risks from exposures outside of the study site might overwhelm the effects of interventions  Ideally, study “confined populations” with limited opportunity for exposures outside of the study site  Molecular epidemiology was suggested for use in studies. Sequencing viruses could help match infections to index cases, helping to identify exposures that lead to infection. A problem raised with this approach was that, in many studies, exposures would occur to many strains of virus, complicating the use of molecular epidemiology techniques

Intervention / Exposure Assessment  It is important to have surveillance data about which groups are at highest risk for acquiring influenza infection. Knowing which workers and tasks are at higher risk can help with targeting study interventions.  There was great enthusiasm for studies of engineering controls.  Near the top of the hierarchy of controls, because individual adherence is not an issue.  Not necessary to know the exact infectious source for the intervention to have an effect  use of engineering control interventions in “captive populations” would assure that these populations were covered all the time and eliminate adherence as a confounder of results.  One concern expressed about engineering controls was the ability to use them in homes; studies might be limited to institutional settings.  Because of the confounding effects of adherence and the impact of respirators and face masks on multiple modes of transmission, it was felt that respirator studies would not be as revealing about modes of transmission as studies using engineering controls such as upper room UVGI or air filtration or directional airflow as interventions.

Intervention / Exposure Assessment  Much discussion and enthusiasm for use of upper-room UV germicidal irradiation with good air mixing as an approach to dissect out the role of aerosol transmission from droplet spray transmission.  Livermore VA study cited  UVGI would be expected to impact on aerosol transmission, but not on droplet spray transmission  Aggressive air mixing with ceiling fans was recommended to knock out local high concentrations of virus aerosol and allow differentiation between aerosol (affected by upper room UV) and droplet spray (not affected by upper room UV)  Properly implemented upper-room UVGI should not be associated with confounding effects such as decontaminating surfaces or stimulating vitamin D production in subjects  UVGI is potentially applicable to low resource settings  Other options for study of engineering controls  Do optimally designed HVAC systems that route infectious particles away from susceptible individuals reduce transmission?  Alternative options to air disinfection were raised, such as air treatment with triethylene glycol or propylene glycol vapor (literature from the 1950s).  Not an engineering control, but point control using inhaled antivirals in an effort to prevent transmission was proposed.

Intervention / Exposure Assessment  PPE / Respirator Intervention Studies  Less informative about transmission than engineering intervention studies due to issues of adherence, unidentified exposures, and simultaneous effects on multiple routes of transmission  Thus, impetus for effectiveness studies of PPE (masks/respirators) is driven by policy concerns separate from need to understand transmission routes  Intervention during natural outbreaks  Logistics are often a prohibitive barrier  Intervention studies conducted in developed countries may not be applicable to most of the rest of the world  Also need to evaluate preventive interventions appropriate to low resource settings