Exploring persistent throat symptoms

Abstract

Background Patients commonly present with a range of persistent throat symptoms. There is still much research to be done to understand how the individual symptoms relate to each other. An improved understanding of any symptom relationships could aid in identifying groups of patients for particular treatments. An opportunity to explore throat symptoms in detail was provided by a clinical trial to assess the effectiveness of stomach acid lowering medication on patients’ throat and voice complaints. The aim was to identify a clinically meaningful classification of patients’ symptoms. Methods Baseline data for all patients entering the Trial of Proton Pump Inhibitors in Throat Symptoms (TOPPITS) was provided by the Newcastle University Clinical Trials Unit. Data included: demographics, three separate symptom questionnaires and a scoring assessment of throat appearances. The relationships between patient demographics, symptom scores and throat appearances were explored with scatter plots and pairwise correlation. Exploratory factor analysis (EFA) of the combined symptom questionnaires was conducted. Cluster analysis of patients using the factors generated by the EFA was then performed. Results Data for 344 patients analysed. The distributions of the questionnaire scores were comparable with published literature. The total scores from the three questionnaires were positively related with each other. No relationship was observed between the throat appearances and any of the three symptom questionnaires in this population. The EFA led to a seven factor model comprising factors of: voice, cough, gastrointestinal symptoms, airway symptoms and dysphagia, throat clearing, life events, and lump in the throat sensation. Cluster analysis failed to identify clinically meaningful groups of patients. Conclusions The TOPPITS baseline data confirmed that patients recruited to the trial reflect the wider population of patients presenting with persistent throat symptoms. No evidence of an association between throat appearances and patient reported symptoms was found in this study. Dimension reduction offered a simplified classification of symptoms, but clusters of patients based on this classification could not be identified. These results imply that individual throat symptoms cannot be used to define patient groups and that the term “persistent throat symptoms” to encompass all symptoms is appropriate to use in clinical practice.