Characterization of orthogeriatric care models and association with mortality and health-economic outcomes in patients with hip fractures: a retrospective cohort study from Germany

Background Multidisciplinary collaboration between traumatologists and geriatricians has gained international recognition for improving outcomes for older patients after fragility fractures through various models of care delivery. To inform decisions about harmonizing existing models, a deeper understanding of their delivery and impact on patient-relevant and health-economic outcomes is needed. Methods We analyzed health insurance claims data from patients aged ≥ 80 years hospitalized with hip fractures between 2014 and 2019 at 121 certified German orthogeriatric centers. We defined hospital-level orthogeriatric care models by the geriatricians’ integration into the surgical wards (geriatric consult service (GCS) vs. integrated care model (ICM)), their weekly patient visit frequency in hospitals with GCS (> 2 (high) vs. 2 (low)), and the network structure between hospitals (single-site institution vs. multisite network cooperation). Outcomes included survival time, direct inpatient medical costs, length of stay (LOS), delivery of early complex geriatric rehabilitation (EGR), and transfer patterns. Results The four care models observed in \:≥\:5 hospitals were low-frequency GCS (26 hospitals, 1479 patients), high-frequency GCS (46 hospitals, 3451 patients), hospital networks (42 hospitals, 1832 patients), and ICM (7 hospitals, 457 patients). Covariate-adjusted death hazards revealed the lowest 30-day hazard in ICM hospitals and the lowest 30-to-180-day hazard in hospital networks. Statistically significant differences were found when comparing these hazards with those in low-frequency GCS hospitals, showing respective reductions of 28% and 20%. Although the care models involved different individual treatment paths, such as varying timings and rates of EGR or transfers to external wards or subacute facilities, their overall costs remained similar. Conclusions Our findings indicate that health outcomes in orthogeriatric care models depend not only on the availability of geriatricians but also on the extent of their involvement in patient management, whereas mean costs were similar across all models.