Mortality | Widely accepted; consistently defined, universally available in trials, observational studies and registries | Insensitive to small or medium treatment effects, those improving symptoms or accelerating recovery |
Clinical treatment success or failure | Frequently evaluated and reported in AECOPD trials A crude measure of treatment effect | Significant variability in the definition that limits comparability |
Improvement in health status or symptoms | Easy to complete questionnaires, frequently self-administered Some are designed to evaluate multiple features of an AECOPD | Significant variability in the utilised measures, which are of untested and doubtful validity Often have complex results that are challenging to interpret |
Length of hospitalisation | Easy to define and widely accepted outcome Universally available in trials, observational studies and registries | Cannot be used for moderate (non-hospitalised) AECOPD Also, its accuracy is limited by: 1) the availability and extent of community COPD care 2) non-medical delays in discharge as well as social care 3) the lack of consistent criteria to guide timing of hospital discharge |
Time-to-treatment success | May be more sensitive to small or medium treatment effects, especially acceleration of recovery | Infrequently reported May be limited by the subjectivity of patient-reported outcomes |
Microbiological response | Easily and consistently defined | Lack of sensitivity and specificity of sputum cultures in COPD exacerbations |
Spirometry | Consistently defined and universally available test | A substantial proportion of patients are unable to perform acceptable spirometry during AECOPD Lack of repeatability during AECOPD |