Cost effectiveness of a novel swallowing and respiratory sensation assessment and a modelled intervention to reduce acute exacerbations of COPD.

Abstract

Swallowing impairment observed in ~ 20% of people with Chronic Obstructive Pulmonary Disease (COPD) may increase the risk of aspiration pneumonia and acute exacerbations. We designed a decision analytic model to assess the cost-effectiveness of the Swallowing and Respiratory Sensation Assessment (SwaRSA) tests and swallowing rehabilitation to reduce COPD exacerbations. We believe that swallowing rehabilitation to improve coordination of swallowing and breathing may reduce exacerbations in people with COPD.From the Australia health system perspective, we assessed the cost effectiveness of four tests relative to standard of care, or no testing, over a time horizon of one year. The SwaRSA tests assessed relative to a standard of care arm of no testing: included the Eating Assessment Tool (EAT-10) score, Swallowing Capacity of Liquids, Tongue Strength Assessment, and Respiratory Sensation Assessment, in people with moderate to severe COPD. Outcome measures were COPD exacerbations per year, which were converted into quality adjusted life years (QALYs). Model inputs including costs, test sensitivities and specificities, COPD exacerbation risks, and exacerbation-related utilities were derived from published sources. Our assumptions on the costs, recovery, and risk reduction are based on the available data on pulmonary rehabilitation in COPD.Relative to no-SwaRSA, three individual testing strategies were found to be cost-effective at incremental cost effectiveness ratio per QALY ranging from $27,000 to $37,000 AUD assuming a willingness to pay of $50,000 AUD. The EAT-10 and the tongue strength were the two dominant options on the cost-effectiveness frontier. Model results were robust to variations in one-way and probabilistic sensitivity analyses.In COPD, SwaRSA modelling suggests that self-assessment with the EAT-10 and subsequent intervention is highly cost-effective relative to no-SwaRSA.