Published in:
Open Access 01-12-2020 | Research article
Validation of the interview-based life-space assessment in institutionalized settings (LSA-IS) for older persons with and without cognitive impairment
Published in: BMC Geriatrics | Issue 1/2020
Login to get accessAbstract
Background
Self-reported life-space assessment methods so far focus on community-dwelling persons, with a lack of validated assessment methods for institutionalized settings. This study evaluated construct validity, test-retest reliability, sensitivity to change, and feasibility of a new Life-Space Assessment for Institutionalized Settings (LSA-IS) in geriatric patients.
Methods
Psychometric properties of the LSA-IS in 119 hospitalized geriatric patients (83.0 ± 6.2 years) with and without cognitive impairment (CI) [Mini-Mental State Examination: 22.4 ± 4.9 scores] were evaluated within a comprehensive validation design. For the total group and subgroups according to cognitive status, construct validity was assessed by calculating Spearman’s rank correlation coefficients (rho) with established construct variables, test–retest reliability by intra-class correlation coefficients (ICCs), sensitivity to change by standardized response means (SRMs) calculated for effects of early ward-based rehabilitation during hospital stay.
Results
The LSA-IS (total score) demonstrated good test–retest reliability (ICC = .704), and large sensitivity to change (SRM = .806), while construct validity was small to high indicated by significant correlations of the LSA-IS to construct variables (rho = .208–716), depending on relative construct association. On average results of LSA-IS sub-scores confirmed results of the total score. Subgroups according to cognitive status did not differ for most analyzed variables. A completion rate of 100% and a completion time of 3.2 ± 1.2 min documented excellent feasibility.
Conclusions
The interview-based LSA-IS has proven to be valid, reliable, sensitive, and feasible in hospitalized, multi-morbid, geriatric patients with and without CI documenting good psychometric properties for institutionalized settings.