Abstract
Background: Disability and loss of function are acknowledged as important problems for people living with COPD, but there is a need for validated tools to assess them.
Research question: The Late-Life Function and Disability Instrument (LLFDI) was originally validated for community-dwelling older adults. The full instrument has not been validated to assess disability and loss of function in people with COPD.
Methods: People with COPD from 6 European countries completed the LLFDI as part of an observational study. Its validity was assessed in terms of 1) levels and distribution of LLFDI domain and subdomain scores; 2) floor and ceiling effects; 3) instrument structure (3 domains, 7 subdomains) by confirmatory factor analysis; and 4) construct validity by (i) convergent validity, based on Spearman correlation with COPD-relevant and related constructs (functional exercise capacity, severity of dyspnea and COPD-related health status), and (ii) known-groups validity, based on the distribution of LLFDI scores according to COPD-meaningful groups (disease severity, age groups and use of a walking aid).
Results: The study included 605 participants (aged 68±8 years, 37% female, FEV1 54±20%pred.). Most had impaired disability and function levels. We observed no floor effects and a ceiling effect in only two subdomains. Confirmatory factor analysis showed a moderate model fit for all LLFDI domains. Most of the correlations met our hypotheses (73%), with moderate to strong correlations for function domain (r min-max 0.25-0.70), followed by disability-limitation domain (r min-max 0.15-0.54), and weakest correlations in the disability-frequency domain (r min-max 0.04-0.41). The disability-limitation and function domains differed by disease severity, age group and use of a walking aid. The disability-frequency domain differed by disease severity and use of a walking aid, but not by age groups.
Conclusion: The LLFDI, a valid patient-reported outcome to investigate disability and function, has proven good construct validity in people with COPD.
Research question: The Late-Life Function and Disability Instrument (LLFDI) was originally validated for community-dwelling older adults. The full instrument has not been validated to assess disability and loss of function in people with COPD.
Methods: People with COPD from 6 European countries completed the LLFDI as part of an observational study. Its validity was assessed in terms of 1) levels and distribution of LLFDI domain and subdomain scores; 2) floor and ceiling effects; 3) instrument structure (3 domains, 7 subdomains) by confirmatory factor analysis; and 4) construct validity by (i) convergent validity, based on Spearman correlation with COPD-relevant and related constructs (functional exercise capacity, severity of dyspnea and COPD-related health status), and (ii) known-groups validity, based on the distribution of LLFDI scores according to COPD-meaningful groups (disease severity, age groups and use of a walking aid).
Results: The study included 605 participants (aged 68±8 years, 37% female, FEV1 54±20%pred.). Most had impaired disability and function levels. We observed no floor effects and a ceiling effect in only two subdomains. Confirmatory factor analysis showed a moderate model fit for all LLFDI domains. Most of the correlations met our hypotheses (73%), with moderate to strong correlations for function domain (r min-max 0.25-0.70), followed by disability-limitation domain (r min-max 0.15-0.54), and weakest correlations in the disability-frequency domain (r min-max 0.04-0.41). The disability-limitation and function domains differed by disease severity, age group and use of a walking aid. The disability-frequency domain differed by disease severity and use of a walking aid, but not by age groups.
Conclusion: The LLFDI, a valid patient-reported outcome to investigate disability and function, has proven good construct validity in people with COPD.
| Original language | English |
|---|---|
| Journal | Annals of the American Thoracic Society |
| Early online date | 23 Sept 2024 |
| DOIs | |
| Publication status | E-pub ahead of print - 23 Sept 2024 |