Responsiveness of PROMIS® to change in chronic obstructive pulmonary disease

Background: Chronic obstructive pulmonary disease (COPD) is a progressive chronic disease characterized by airflow obstruction that leads to shortness of breath and substantial negative impacts on health-related quality of life (HRQL). The course of COPD includes periodic acute exacerbations that require changes in treatment and/or hospitalizations. This study was designed to examine the responsiveness of Patient-Reported Outcomes Measurement Information System® (PROMIS®) measures to changes associated with COPD exacerbation recovery. Methods: A longitudinal analysis using mixed-effects models was conducted of people who were enrolled while stable (n = 100) and those who experienced an acute exacerbation (n = 85). PROMIS (physical function, pain interference, pain behavior, fatigue, anxiety, depression, anger, social roles, discretionary social activities, Global Health, dyspnea severity and dyspnea functional limitations) and COPD-targeted HRQL measures were completed at baseline and at 12 weeks. Results: We administered PROMIS measures using computer adaptive testing (CAT), followed by administration of any remaining short form (SF) items that had not yet been administered by CAT. Examination of the difference between group differences from baseline to 12 weeks in the stable and exacerbation groups revealed that the exacerbation group changed (improved) significantly more than the stable group in anxiety (p <.001 to p < .01; f² effect size [ES] = 0.023/0.021), fatigue (p <.0001; ES = 0.036/0.047) and social roles (p <.001 to p < .05; ES = 0.035/0.024). All effect sizes were small in magnitude and smaller than hypothesized. Depression was also statistically significant (p <.05, SF only) but the ES was trivial. For all other PROMIS domains, the differences were not significant and ES were trivial. Conclusions: This longitudinal study provides some support for the validity of the PROMIS fatigue, anxiety, and social roles domains in COPD, but further evaluation of responsiveness is warranted.