BACKGROUND AND OBJECTIVE:Exacerbation(s) of chronic obstructive pulmonary disease (eCOPD) entail important events describing an acute deterioration of respiratory symptoms. Changes in medication and/or hospitalization are needed to gain control over the event. However, an exacerbation leading to hospitalization is associated with a worse prognosis for the patient. The objective of this study is to explore factors that could predict the probability of an eCOPD-related hospitalization.
METHODS:Data from 128 patients with COPD included in a prospective, longitudinal study were used. At baseline, physical, emotional, and social status of the patients were assessed. Moreover, hospital admission during a one year follow-up was captured. Different models were made based on univariate analysis, literature, and practice. These models were combined to come to one final overall prediction model.
RESULTS:During follow-up, 31 (24.2%) participants were admitted for eCOPD. The overall model contained six significant variables: currently smoking (OR = 3.93), forced vital capacity (FVC; OR = 0.97), timed-up-and-go time (TUG-time) (OR = 14.16), knowledge (COPD knowledge questionnaire, percentage correctly answered questions (CIROPD%correct)) (<60% (OR = 1.00); 60%-75%: (OR = 0.30); >75%: (OR = 1.94), eCOPD history (OR = 9.98), and care dependency scale (CDS) total score (OR = 1.12). This model was well calibrated (goodness-of-fit test: p = 0.91) and correctly classified 79.7% of the patients.
CONCLUSION:A combination of TUG-time, eCOPD-related admission(s) prior to baseline, currently smoking, FVC, CDS total score, and CIROPD%correct allows clinicians to predict the probability of an eCOPD-related hospitalization.