Exploration of differentiated service models for hospital at home from the perspective of elderly patients: Based on latent class analysis.

BACKGROUND

This study analyzes Hospital at Home (HaH) service utilization among community-dwelling elderly patients and explores its influencing factors.

METHODS

This cross-sectional study employed face-to-face questionnaires with elderly patients enrolled in the Hospital at Home service at four community health centers in Minhang District, Shanghai. Service utilization items from the checklist served as observed indicators for latent class analysis. Variables covered predisposing, enabling, and need factors based on Andersen's behavioral model. Descriptive analysis, chi-square tests, and multinomial logistic regression were used to examine differences across latent classes, with a significance level of P = 0.05.

RESULTS

511 elderly patients receiving Hospital‑at‑Home (HaH) services were included. Among them, 58.3% were female, 44.6% were aged over 80 years, and 82.4% had multiple chronic comorbidities. Latent class analysis delineated three distinct service utilization profiles: a high‑intensity, multidimensional integrated care pattern (C1, 17.0%), a medication‑focused and health‑guided pattern (C2, 31.9%), and a basic follow‑up and supportive care pattern (C3, 51.1%). Significant differences existed across predisposing, enabling, and need factors. Multinomial logistic regression showed class membership was associated with staff attitude, medication count, pain, dysphagia, illness perception, elderly capability assessment, and IADL, indicating tailored matching between patients and service patterns.

CONCLUSION

Latent class analysis identified three HaH service utilization patterns, overcoming limitations of homogeneous services and providing evidence for precise service design targeting different patient groups.