Predicting the Outcome of Pneumonia for People who Aspirate (PrOPPA) Study: Predicting the outcome of pneumonia for individuals who aspirate when eating and drinking as a result of oropharyngeal dysphagia:- determining the validity, reliability and acceptability of a prognostic clinical prediction model to support clinical management.

Background Oropharyngeal dysphagia occurs in multiple health conditions [7-15] and is highly prevalent throughout health and social care settings [16]. It is thought that 51% of people with oropharyngeal dysphagia aspirate [1], and 50% of those develop an aspiration pneumonia [2]. Aspiration pneumonia is life threatening and treating pneumonia is very costly to the NHS [17]. For quality of life, to promote nutrition and hydration, or as a step in swallow rehabilitation, people may choose to eat and drink knowing that they are aspirating [18-19]. It is not clear from the current literature who is likely to develop aspiration pneumonia and who is not. This study aims to refine and validate the published prognostic clinical prediction model [4] (Figure 1) that was developed in response to this evidence-base gap. The model was developed using seven years of retrospective data from a cohort of in-patients and out-patients, from a variety of settings, with varying health conditions (neurology, oncology, respiratory, spinal, gastroenterology, cardiac, mental health and adults with learning disability). Each patient had had a videofluoroscopy which showed aspiration. Each patient was eating and drinking with acknowledged risk of developing aspiration pneumonia. Logistic regression analysis was used to produce the model. Aim To evaluate the validity, reliability and acceptability of the clinical risk predication model of aspiration pneumonia in a prospective, multi-centre study. Objectives To determine the validity and reliability of the model in predicting the development of aspiration pneumonia for people who aspirate (WP1). To maximise the acceptability of the model for future users (professionals and patients) (WP2, WP3). Methods WP 1 (Year 1-3): Data will be collected from ten NHS sites with a target sample size of 400. Inclusion criteria: People who have had an instrumental swallowing assessment that shows aspiration on what they are eating and drinking. Variables: Risk factors for developing aspiration pneumonia (Figure 1). Outcome variable: Whether pneumonia is developed in the subsequent three months following eating and drinking with acknowledged risk recommendations commencing following an instrumental assessment showing aspiration. Analysis: a) Reliability and refinement: Multiple logistic regression analysis of the prospective data when compared with the previous study's retrospective data b) Validation: V-fold cross-validation of the prospective data [20]. Key outcome: a refined and internally and externally (geographically) validated model. WP 2 (Year 1-2): A national multidisciplinary stakeholder survey to explore barriers and facilitators to implementation. Analysis: Thematic analysis [21]. Key outcome: Identified themes concerning implementation barriers and facilitators. WP 3 (Year 2-3): A co-produced animation (implementation) that facilitates decision-making regarding eating and drinking with known aspiration. Method: 1) Evidence review and stakeholder consultation (informed by WP1 and WP2) 2) Co-production 3) Prototyping [22]. Key outcome: A facilitative and inclusive animation. Impact and dissemination The research (including the clinical tool; model and animation) will be disseminated via peer-reviewed journal publications, international conferences and targeted (professional and patient) social media. Future research will evaluate the impact of the model including its anticipated influence on policy, patient experience and clinician experience.