Demographic and clinical characteristics associated with the attrition in a cohort of adult asthma

Background In observational studies, unexpected attrition of patients can significantly reduce statistical power and comprise study objectives. In this study, we aimed to assess the factors predicting attrition in a cohort of adult asthma.

Methods In an adult asthma cohort (COREA) from 11 university hospitals in Korea, we estimated the proportion of completers and dropouts among the patients who enrolled more than 18 months ago. Completers were defined as the patients who had one or more visit records in the next 12-18 months after registration. Demographic and clinical features were compared between drop-outs and completers.

Results Of 1,499 patients, 670 (44.7%) patients dropped out in the study period. The dropouts showed lower mean age (43.1 ± 15.8 vs. 51.1 ± 14.3, P<0.001), shorter duration of treatment (3.5 ± 5.1 vs. 5.5 ± 6.0 years, P<0.001) and positive skin test (28.2% vs. 23.6%, P=0.044) compared with the completers. While there was no significant difference in lung function between two groups, drop-outs had lower severity of asthma based on physician-assessed severity (P=0.005) and used fewer controllers (P<0.001). Regarding the previous use of health care utilities (HCU), the dropouts had less use of HCU for the treatment of asthma.

Conclusion The dropouts in an adult asthma cohort showed younger age, shorter duration of treatment, higher sensitization rate to allergens, lower severity of asthma, and less use of HCU for asthma. These findings suggest the factors predicting attrition in the management of asthma.

Keywords: attrition, drop-outs, asthma, characteristics