Validation of the World Health Organization Disability Assessment Schedule in Refugees of Kiryandongo Settlement in Uganda

Background and Objective. This study examines the underlying factor structure of the World Health Organization Disability Assessment Schedule (WHODAS) with data from a sample of refugees living in Kiryandongo refugee settlement, Uganda. The WHODAS is a widely used instrument grounded in the International Classification of Functioning, Disability, and Health. Social and functional challenges are particularly pronounced in refugee settlements. However, the performance of this instrument is largely untested in the African context (Habtamu et al. 2017). Therefore, we aim to assess its applicability and validity with this unique population by employing exploratory and confirmatory factor analyses.

Method. We drew secondary data from a quasi-experimental study which evaluated the effectiveness of a psychosocial intervention among refugee caregivers in Uganda (Stark et al. 2024). We employed a split-sample psychometric validation of the 12-item WHODAS, among refugee caregivers. The dataset (n = 1,129) was randomly divided into two halves (n = 564 for EFA; n = 565 for CFA). Sampling adequacy was assessed using the Kaiser-Meyer-Olkin (KMO) test and Bartlett’s test of sphericity prior to extraction. Exploratory Factor Analysis (EFA) was conducted on the first subsample using principal axis factoring with oblimin rotation. Factor retention was guided by eigenvalues > 1, scree plot, and theoretical coherence. Confirmatory Factor Analysis (CFA) was performed on the remaining sample using robust maximum likelihood estimation (MLR), and model fit was evaluated using the Comparative Fit Index (CFI), Tucker-Lewis Index (TLI), Root Mean Square Error of Approximation (RMSEA), and Standardized Root Mean Square Residual (SRMR).

Results:
The overall KMO measure was .89, indicating excellent sampling-adequacy, with individual MSA values ranging from .71 to .94. Bartlett’s test of sphericity was significant (χ²(66) = 5694.59, p < .001), confirming that the correlation matrix was not an identity matrix. EFA supported a three-factor structure; (a) physical and cognitive functioning, (b) basic self-care, and (c) social functioning, accounting for 54% of the total variance. Communalities were moderate to high, and the model demonstrated excellent fit (TLI = .975; RMSEA = .043; RMSR = .02). Factor score adequacy was strong for the first two factors (R² = .89) and weaker for the third (R² = .66). The three-factor model was then tested via CFA on the second sub-sample. While factor loadings for most items were statistically significant and theoretically aligned, model-fit was suboptimal (CFI = .794; TLI = .734; RMSEA = .117; SRMR = .091). The Social Functioning factor was particularly unstable, with one item loading weakly (λ = .24). High correlations between latent factors (e.g., F1–F3 = 1.10) suggested potential multicollinearity or structural overlap.

Discussion and Implications:
This analysis provides partial support for the construct validity of WHODAS among refugee caregivers in Uganda. The EFA results revealed a coherent and interpretable structure, but the CFA indicated limitations in the measurement model, particularly in the social functioning domain. These inconsistencies may reflect contextual or cultural differences in how disability manifests or is reported in refugee populations. The findings potentially underscore the importance of validating global health instruments in local contexts, rather than assuming conceptual equivalence.