Reliability and Minimum Important Difference of Sputum Weight in Bronchiectasis

Despite the widespread use of sputum weight to assess the effect of airway clearance interventions, its psychometric properties have not been evaluated. The purpose of this ad hoc analysis was to determine the test-retest reliability of 24-h sputum weight in clinically stable individuals with bronchiectasis. This study also aimed to estimate the minimum important difference of 24-h sputum weight after an airway clearance session in subjects with bronchiectasis.

Sixty subjects were included in the 24-h test-retest cohort, 42 of whom were part of the airway clearance cohort. For the 24-h test-retest cohort, spontaneously expectorated sputum was collected over 24 h on 2 different days without any airway clearance interventions. For the airway clearance cohort, sputum expectoration was also collected during 3 airway clearance sessions and over the 24 h following these interventions. Intraclass correlation coefficient (ICC_3,1) and Bland-Altman analysis were used to assess reliability. The minimum important difference was calculated using distribution-based and anchor-based methods, with cough impact as assessed with the Leicester Cough Questionnaire and the global rating of change as anchors.

The reliability was acceptable (ICC_3,1 = 0.75) for sputum weight over 24 h without any intervention. The agreement level was wide, particularly for high levels of sputum expectoration. The minimum important difference of the sputum collected in the 24 h after the intervention from baseline was -6.4 g (about -17%), determined using distribution-based methods. There was no correlation between sputum weight and the anchors, thus the anchor-based methodology could not be used.

Multiple measurements should be considered to increase the agreement when sputum weight is used as an outcome measure for short periods in people with bronchiectasis. A reduction of 6.4 g (or 17% from baseline) in sputum collected during the 24 h after the airway clearance intervention may be considered the minimum important difference in people with bronchiectasis. (ClinicalTrials.gov registration NCT02392663; NCT01854788; NCT02614300.).