A retrospective pilot study of the use of a new algorithm to improve quality control in bronchodilator studies

Reversibility testing is used to identify a positive or negative response to bronchodilators. Results from a reversibility test can not only support a diagnosis of asthma but can alter a patient's treatment plan, so its clinical importance should not be understated. With multiple guidelines published classifying a 'positive response' it becomes unclear on how to categorise certain individuals. This study looks into the discrepancies between the guidelines, and introduces a new algorithm to help clinicians. This retrospective pilot study was completed across four hospitals in South Wales. Data were collected from a total of 117 patients referred for a reversibility study during November 2013 and April 2014. An algorithm was created to improve flow-volume loop (FVL) quality control when assessing airways bronchodilation in symptomatic patients. Each patient result was placed through four major reversibility guidelines [British Thoracic Society (BTS), National Institute for Clinical Excellence (NICE), Association for Respiratory Technology Physiologists (ARTP) and Global Lung Initiative (GLI)] and the new algorithm. When comparing published guidelines, 75% of patients would receive the same bronchodilator response decision, positive or negative, irrespective of the guideline followed. Variability between the numbers of positive responders in each guideline varied by up to 58%, with NICE found to give the least number of positive responses (7%), and BTS giving the greatest (65%). Using the new algorithm, over one third (38%) of patients required a repeat FVL, as baseline and/or post-bronchodilator FVLs did not meet the quality control specification. Further investigation is needed to establish the clinical impact of the new algorithm, and its approach to using the whole of the FVL in bronchodilator analysis; however, quality control during reversibility testing needs to be improved to ensure that bronchodilator responses are correctly identified.