Professor Craig Smith

Stroke experts hear about oral healthcare treatments to prevent pneumonia

A leading stroke researcher has shared the findings of a key study aiming to prevent pneumonia in stroke unit care.

Professor Craig Smith, theme lead for stroke at the Geoffrey Jefferson Brain Research Centre and a consultant in stroke medicine at Northern Care Alliance (NCA) NHS Foundation Trust, presented the findings from the CHOSEN trial at the UK Stroke Forum in Birmingham.

CHOSEN (Chlorhexidine or toothpaste, manual or powered brushing to prevent pneumonia complicating stroke) is a multicentre, randomised clinical trial in the North West of England  funded by the NIHR and led by Professor Smith. CHOSEN aimed to determine the feasibility of different oral healthcare treatments, supported by staff education and training, to prevent pneumonia in UK stroke unit care.

Pneumonia affects around one in 12 people after a stroke and increases their chances of death and disability. Poor oral health increases the risk of pneumonia, as inflammatory material and bacteria in the mouth can enter the lungs when there are swallowing problems (dysphagia). Improved oral healthcare is therefore a plausible approach to preventing pneumonia.

Education programme

The trial team developed an online training and education programme to help stroke unit staff deliver the oral healthcare treatments to people after a stroke.

The main feasibility trial involved a total of 101 people admitted to hospital within 24 hours of a stroke and with dysphagia. Participants were randomly allocated to one of four oral healthcare treatments: Chlorhexidine gel (with either a powered toothbrush or a manual toothbrush) versus non-foaming toothpaste (with either a powered toothbrush or a manual toothbrush).

Recruitment of participants to the feasibility trial at Salford Royal, Fairfield General – both part of NCA, and two other North West hospitals completed in January 2023 with final follow-up in April 2023.


All of the main trial feasibility criteria were met in terms of setting up and training the participating sites, recruitment and retention of participating patients, adherence to allocated treatment and collection of the outcome measures. There were no safety concerns overall, or between the different oral healthcare treatments.

In exploratory secondary analyses, there were no differences in clinical outcomes between the different allocation groups. A parallel process evaluation will provide further context and help us understand how best to implement the training and oral healthcare treatments.

Professor Smith and the team now anticipate undertaking a much larger phase 3 trial across the UK to test cost-effectiveness of the oral healthcare treatments, which would change clinical practice.

Skip to content