Project Details

Report now available.
View the executive summary (pdf format)
You can view the full text from the H1N1 influenza and pandemic flu themed issue 1 publication details page on the HTA website

Pandemic influenza is likely to be spread by droplets expelled when infected individuals cough, sneeze and breathe normally. A small proportion of these individuals develop breathing complications such as pneumonia especially if they have underlying chronic health problems. The therapies used to treat these people, such as non-invasive ventilation (NIV), oxygen therapy, physiotherapy and nebulised drugs may increase droplet spread around the patient/and or generate aerosols of infected material. In the epidemic of Severe Acute Respiratory Disease (SARS) in 2003 a high proportion of healthcare workers were infected and there was spread of infection on the ward which was associated with a range of factors which included use of NIV, nebuliser therapy etc. However it is not clear whether these therapies were required in sicker patients with a higher viral load, or whether droplet transmission is increased. Some studies have attempted to visualise the spread of droplets but these have either used patient simulators or employed smoke particles as a marker of dissemination around the devices. Patient simulators may not represent the breathing patterns of patients accurately and the physical dispersion properties of very small smoke particles are unlikely to mimic the distribution of exhaled water droplets.

We will therefore study droplet and aerosol distribution in normal subjects, those with cold or flu-like symptoms and those with a flare-up of a chronic breathing problem such as chronic bronchitis or cystic fibrosis such that they require NIV, oxygen therapy and nebuliser treatment.

The study will be carried out in a ward side room at Royal Brompton Hospital. 
The outcomes will be the number and size of particles generated next to, and at fixed distances from the subject/patient. These particles will be measured by 2 small counting devices which will be mounted at fixed distances from the individual and will not interfere with therapy in any way.

Limitations: The patients studied during a flare up of their condition will not have swine flu but most cough and the particles produced are likely to be similar regardless of the trigger for the flare up. We will not know how many viruses are present in the droplets, but as viruses are thought to be spread by droplets we will gain a idea of likelihood of dissemination.

Ethical aspects: We will study patients in whom treatment with NIV and O2 is already clinically indicated and ensure they are stable. the settings of the ventilator will be those that have been shown to be indicated for the patient. Patients will be monitored throughout the study so we know that they are stable.
In the group with cold-like of flu symptoms there is a risk of transmission of infection in research staff, so aerosol generating protective measures will be taken in infection control side room. The study will be subject to a full ethics application.

Team: The research team comprises of members who are experienced managing individuals with respiratory problems on a daily basis and familiar with the ventilatory, oxygen, nebuliser and physiotherapy techniques, and have carried out previous crossover trials. Team members from the Bioengineering dept at Imperial College have expertise in droplet visualisation and mathematical modelling of droplet dispersion. Both teams have worked together on successful projects previously.

Justification costs: the costs consist of fulltime research fellow, supervisor time and mathematical modelling time plus travel cost for subjects and equipment costs to include 1 ventilator, masks, circuitry filters. For the patient studies there are no additional NHS costs not covered by equipment/consumable costs as patients will already be in hospital undergoing monitoring.

Applicability: we believe the study results will be immediately applicable to the respiratory management of patients with H1N1 influenza who develop respiratory complications, improve safety for healthcare workers and other patients. If this allows wider safer use of NIV, one consequence may be a reduction in pressure on Intensive Care Unit beds as NIV can be applied on respiratory and high dependency ward areas.

The key objective is to understand the characteristics of droplet and aerosol dispersion around delivery systems during non-invasive ventilation (NIV), oxygen (O2) therapy, nebuliser therapy and physiotherapy procedures.

We will examine: 
i) droplet size and count 
ii) geographical distribution of droplets
iii) rise and decay of droplets over time after the therapies are initiated and discontinued 
iv) the impact of modifications to the delivery system to reduce droplet/aerosol dissemination 
in a) normal subjects b) individuals with coryzal symptoms and c) patients with an acute exacerbation of chronic lung disease