From Validation to Risk Stratification: Psychometric Evaluation and Latent Class Analysis of the Pediatric Psychosocial Assessment Tool (PPAT)

Background and Purpose

Pre-transplant psychosocial assessment is a critical component of determining pediatric transplant candidacy, yet few standardized tools exist with demonstrated psychometric validity or clinical utility. The Pediatric Psychosocial Assessment Tool (PPAT) was developed to provide a structured, multidimensional evaluation of psychosocial risk across domains relevant to transplant readiness. While initial work supports its reliability and predictive value, less is known about whether psychosocial risk operates as a single continuum or reflects distinct risk profiles. This study evaluates the psychometric properties of the PPAT and extends prior work by examining heterogeneity in psychosocial risk through latent class analysis (LCA), with direct implications for medical review board (MRB) decision-making.

Methods

Data were drawn from 189 pediatric heart transplant evaluations conducted at a large pediatric heart transplant center. Confirmatory factor analysis (CFA) using WLSMV estimation assessed factor structure across nine ordinal PPAT domains. Internal consistency reliability was evaluated using Cronbach’s alpha and McDonald’s omega. Criterion-related validity was examined through logistic regression predicting MRB listing decisions, and construct validity was assessed via known-groups comparisons across patient demographic and clinical characteristics. To examine heterogeneity in psychosocial risk, LCA was conducted using the nine PPAT domains as ordinal indicators. Model fit was evaluated using entropy and likelihood-based indices, and class interpretation focused on clinically meaningful risk patterns.

Results

CFA supported a unidimensional structure of psychosocial risk with acceptable model fit (CFI = 0.950, TLI = 0.933, SRMR = 0.071), although RMSEA suggested some model misfit (RMSEA = 0.128). All domain loadings were significant (λ = .690–.872), and reliability was strong (α = .885; ω = .886). PPAT scores demonstrated strong criterion-related validity, significantly predicting MRB listing decisions (χ²(1) = 11.10, p < .001; Nagelkerke R² = .162). Each one-point increase in PPAT score was associated with a 17.5% decrease in odds of being listed for transplant (OR = 0.825), indicating clinically meaningful predictive utility. Construct validity was supported by higher PPAT scores among patients with greater medical complexity, including those with genetic syndromes and prior transplant history.

LCA identified three distinct psychosocial risk profiles with strong classification accuracy (entropy = 0.855). A high-risk class (≈19%) demonstrated elevated difficulties across multiple domains, including adherence, literacy, and coping. A low-risk class (≈30%) showed minimal concerns across domains, reflecting stable psychosocial functioning. A moderate-risk class (≈50%) exhibited targeted challenges, particularly in financial strain, understanding of the transplant process, and coping. These findings indicate that psychosocial risk is not purely continuous but reflects distinct, clinically meaningful risk strata.

Conclusions and Implications

Findings support the PPAT as a psychometrically sound and clinically meaningful measure of psychosocial risk in pediatric transplant populations. Beyond validation, results demonstrate that psychosocial risk is heterogeneous, with distinct profiles that may not be captured by total scores alone. Integrating PPAT-based risk stratification into MRB processes has the potential to enhance decision-making, tailor interventions, and improve equity in transplant evaluation. This study highlights the role of standardized psychosocial measurement in advancing precision assessment within social work and interdisciplinary clinical care.