Network Analysis of Factors Influencing ICU Patient Transfer Readiness: A Social-Ecological Model Approach

ICU transfer readiness is a multidimensional state influenced by informational, functional, emotional and support-related needs. Network analysis can clarify how readiness indicators and emotional distress co-occur as a system, rather than as independent factors.

To (1) model conditional associations among ICU transfer-readiness indicators and anxiety/depressive symptoms and (2) identify highly connected and highly predictable targets that may be prioritised in transitional-care interventions.

Transfer readiness was measured using the ICU Transfer Readiness Scale (15 items; 0-10) and emotional distress using the Hospital Anxiety and Depression Scale (HADS; 14 items; 0-3). A regularised mixed graphical model (MGM) was estimated to accommodate mixed/ordinal data. Edge-weight accuracy, centrality robustness and node predictability (R²) were assessed using bootstrap procedures. Age and gender were included as background variables.

A total of 276 ICU patients were included in the analysis. The estimated network showed a clear modular structure: transfer-readiness indicators clustered into personal status, disease knowledge, coping after transition and expected-support domains; HADS symptoms clustered into anxiety and depression domains. Within transfer readiness, the strongest association occurred between two expected-support indicators, suggesting perceived support is tightly interlinked. Within HADS, multiple strong within-subscale connections were observed for both anxiety and depression symptoms. Node predictability ranged from 0.224 to 0.633 (mean 0.434), with the highest predictability for key readiness indicators in the knowledge/coping domains (R² ≈0.63). Bootstrap analyses supported acceptable stability of the main edge and centrality patterns. Demographic nodes showed comparatively weaker and more cautious-to-interpret associations.

ICU transfer readiness and emotional symptoms form structured conditional interdependencies, indicating readiness should be conceptualised and addressed as a connected system rather than isolated determinants.

Transitional-care strategies may be strengthened by prioritising highly connected/predictable readiness targets-particularly expected support and key knowledge/coping indicators-while concurrently screening for and addressing anxiety/depressive symptoms to improve transfer preparation and early post-transfer adjustment.