Safety, Feasibility, and Exploratory Functional Changes During GRILLO© Gait Trainer Use in Adults with Severe...

Background/Objectives: Assisted verticalization and supported upright activity are relevant components of rehabilitation in adults with severe acquired brain injury (sABI), although patient selection and implementation remain challenging. This retrospective observational study aimed primarily to describe the implementation feasibility and documented safety of GRILLO-based training in routine inpatient multidisciplinary rehabilitation, and secondarily to report exploratory pre-post functional changes. Methods: We reviewed clinical records of 34 adults screened or considered for GRILLO-based training at Centro Cardinal Ferrari KOS, Italy, between June 2022 and December 2024. GRILLO training was delivered as part of standard care and not as an experimental intervention. Functional outcomes included the Barthel Index (BI), Trunk Control Test (TCT), Tinetti Balance Scale, and Tinetti Gait subscale, extracted from routine documentation. Non-parametric descriptive analyses were used. Results: Of 34 screened patients, 4 did not meet diagnostic criteria for ABI, 5 interrupted training because of pain or poor tolerance to prolonged upright positioning, and 3 were not included because of poor compliance/motivation or an incomplete clinical pathway. The paired functional-analysis cohort comprised 22 patients: 20 (91%) completed 15 sessions and 2 (9%) completed 10 sessions. No serious device-related adverse events were documented in available clinical records, although minor adverse events were not systematically monitored. Among patients with paired observations, median BI increased from 16 to 22.5 (median change, +3; p = 0.008; n = 20), median TCT from 72 to 74 (median change, +12; p < 0.001; n = 21), and median Tinetti Balance Scale from 1 to 2 (median change, +1; p = 0.006; n = 22). Individual responses were heterogeneous and floor effects were evident, especially for balance and gait-related measures. Conclusions: In this retrospective real-world cohort, GRILLO-based training could be implemented in selected severely impaired inpatients, but feasibility may be overestimated if interrupted and non-completing cases are not considered. The non-completion cases may suggest that feasibility depends not only on initial clinical indication, but also on the appropriate timing of introduction, tolerance to prolonged upright physical effort, pain/discomfort, motivation, and behavioral engagement. The retrospective design, survivorship bias, non-systematic adverse-event monitoring, concurrent multidisciplinary rehabilitation, and absence of a comparator group preclude conclusions regarding device-specific safety or efficacy. Nevertheless, these preliminary findings support further prospective controlled studies.