Validation and Further Analysis of the COPD Exacerbation Recognition Tool ( CERT )

ABSTRACT Background and Objective The COPD Exacerbation Recognition Tool ( CERT ) contains five symptoms that may change at the onset of an exacerbation. It was designed to help patients with COPD recognise the onset of an exacerbation and seek medical attention. This study tested its sensitivity and specificity and further examined its properties. Methods Stable and exacerbating patients presenting to three tertiary hospitals in China were enrolled. Patients' responses to the CERT were compared with physicians' diagnoses made using GOLD 2021 criteria. The CERT was assessed as recommended, and as a total score, using scores assigned to the degree of change for each of the five symptoms (total score 15). Exploratory factor analysis (EFA) was used to identify the key components of the CERT and potentially reduce the number of items. Results Sensitivity and specificity of the original CERT were 92.4% and 72.1%. Using a cut‐off value of 6, the score‐based CERT had the better sensitivity (94.3%) and specificity (86.9%). Concordance between CERT and physician diagnosis was higher with the scored version (kappa 0.833) versus the original version (kappa 0.666). Using EFA, two components were identified ‘cough and sputum’ and ‘breathlessness and activity limitation’ with a cumulative variance of 82.3%. Conclusion The original and scored‐based CERT both had good sensitivity and specificity, but the score‐based version showed better concordance with physician diagnosis. The use of two representative items in a 2‐item scored version was also effective in identifying acute exacerbations. Trial Registration ChiCTR2400090589 registered with https://www.chictr.org.cn