Thirty-count breathlessness score: reliability, sensitivity, specificity and validity

The fifteen-count breathlessness score (15CBS) has been reported to quantify breathlessness; however, a ceiling effect limits its ability to discriminate between subjects with varying degrees of breathlessness. The aim of this study was to determine the reliability, sensitivity, specificity and validity of the thirty-count breathlessness score (30CBS) in adults with COPD.

Using an observational correlation design, subjects with clinical signs and symptoms of COPD were videotaped performing the 30CBS on two occasions. Respiratory related quality of life questionnaires, self-report shortness breathlessness measures and pulmonary function tests were completed by all subjects. Asymptomatic subjects of a similar age completed the 30CBS for comparison.

Thirty-eight COPD subjects and 24 control subjects completed the protocol. The 30CBS was reliable within and between assessors (intraclass correlation coefficients >0.7). The 30CBS demonstrated a higher sensitivity (55.8% sensitivity, 82.3% specificity) than the 15CBS (33.2% sensitivity, 93.2% specificity); however, the sensitivity of the 30CBS was still not great enough to discriminate between those with and without COPD. Regression analysis calculated significant relationships between the 30CBS and a majority of measures of pulmonary impairment (flows and volumes), age, height, Visual Analogue Scale for breathlessness and physical aspects of the Short Form-36.

While reliable, the 30CBS did not display sufficient discriminative ability to predict subjects with COPD though subjects requiring greater than two breaths may warrant further investigation. The 30CBS may be a useful indicator of physiological impairment, but was not significantly correlated with measures of breathlessness.