Pragmatic cluster-randomised trial of a multidomain intervention to prevent hospital-acquired disability: the OPTIMAge-IT study protocol Authors: Chukwuma Okoye, Maria Cristina Ferrara, Beatrice Tonus, Wenxiang Guo, Elena Page, Eleonora Cucini, Martina Marelli, Alberto Finazzi, Elena Pinardi, Filippo Fimognari, Carmen Ruberto, Paola Palestini, Roberta Daini, Francesca Gasparini, Stefano Calciolari, Eleonora Pagan, Luca Cuffaro, Federico Emanuele Pozzi, Gessica Sala, Andrea Corsonello, Fabio Guerini, Agostino Virdis, Alessandra Marengoni, Andrea Ungar, Antonella Zambon, Carlo Ferrarese, Giuseppe Bellelli, Emma Esposito, Emma Esposito, Anna Vanoncini, Flavia Sandi, Greta Tavecchi, Lorenzo Lodovici, Serena Bertocchi, Paola Negri, Luca Valsecchi, Maria Chiara De Ponti, Valeria Ortu, Lucia Ferrara, Roberto Galletti Published: 2026 ABSTRACT BackgroundHospitalisation frequently leads to hospital-acquired disability in older adults, with subsequent functional decline and loss of independence. However, the potential role of multidomain interventions (MDIs) integrating physical, nutritional and cognitive components, supported by digital technologies, remains largely unexplored in the acute care setting. OPTIMAge-IT therefore aims to evaluate the effect of a technology-supported MDI on motor function at hospital discharge and its feasibility in the post-discharge phase among frail older adults.MethodsOPTIMAge-IT is an ongoing, pragmatic, multicentre, cluster-randomised controlled trial conducted in eight Italian Acute Geriatric Units. Adults aged ≥ 70 years admitted from the Emergency Department with mild-to-moderate frailty are enrolled consecutively and allocated by centre to usual care or a structured, multidisciplinary, technology-supported MDI. The intervention starts after clinical stabilisation and continues for 12 weeks post-discharge, supported by a tracker and a tablet. The MDI includes Vivifrail-based physical exercise, computerised cognitive training (BrainHQ), nutritional counselling, health education, medication review, and promotion of social engagement. The primary endpoint is change in Short Physical Performance Battery at hospital discharge. Secondary endpoints include feasibility of the intervention and its home continuation (adherence, usability and retention), as well as the effect of the MDI on in-hospital and post-discharge outcomes. Follow-up assessments are performed at 3 and 6 months in the intervention arm and at 6 months in the control arm; blood biomarkers are collected at discharge and at 6 months.DiscussionOPTIMAge-IT will provide evidence on the feasibility and clinical value of a technology-supported MDI to prevent hospital-acquired disability and promote functional status in frail older adults. Findings may inform the development of scalable, person-centred post-hospital care models.Trial registrationNCT06611228 (https://clinicaltrials.gov/study/NCT06611228; 2025–01-08). Background In older adults, hospitalisation often marks a critical inflection point, shifting an acute illness into a trajectory of functional decline and loss of independence [1]. Even after resolution of the index condition, many patients experience persistent deterioration in mobility, cognition and self-care — a syndrome known as hospital-associated disability (HADS), which strongly predicts adverse outcomes [2, 3]. Previous studies suggest that exercise-based programmes and multicomponent approaches may help preserve physical function and reduce delirium [4–6]. In this context, multidomain interventions (MDI) may be an innovative strategy to prevent functional decline and promote recovery, though remaining largely unexplored in the acute care setting [7]. Meanwhile, healthcare systems are undergoing a digital transformation, with technological sensors and tele-monitoring progressively entering routine practice and creating the infrastructure to deliver and track multidomain care from ward to home [8]. However, the potential of information and communication technologies (ICTs) to extend reach, support adherence, and reduce operational burden in this setting is underexplored. Digital tools—including wearables, patient-facing cognitive applications, and telemedicine platforms—could also provide seamless post-discharge continuity, but their integration into hospital-based models has yet to be tested. The “Optimizing Prevention of hospital-acquired disability Through Integrated Multidomain interventions: the Age-IT project” (OPTIMAge-IT study) is designed to fill these gaps, within the Age-It program [9]. Grounded in the principles of WHO Integrated Care for Older People (ICOPE) guidelines [10] and inspired by the methodological pillars of WW-FINGERS [11], it is the first multicentre randomised clinical trial to evaluate the effects of a multidisciplinary, technology-supported MDI on multiple outcome measures in older patients hospitalised with mild to moderate frailty. The primary aim is to assess the efficacy of MDI on motor function at hospital discharge. Secondary aims include evaluating clinical, cognitive and psychological outcomes to capture the overall impact of the intervention, as well as assessing the feasibility of the technology-supported intervention in hospital and at home after discharge. Methods Study design and setting OPTIMAge-IT is a pragmatic, multicentre, parallel-group, cluster-randomised trial being conducted in eight Italian Acute Geriatric Units (AGUs). The trial compares a technology-supported MDI with usual care. After a 5-month preparatory phase to complete cluster randomisation and harmonise procedures, centres started recruitment in parallel. Four AGUs were randomly allocated to the intervention arm and four to usual care. Both enrolment and follow-up assessments (at 3 and 6 months after discharge) are still ongoing (Fig. 1).Fig. 1Flowchart of the study design Flowchart of the study design Participants Eligible participants are adults aged ≥ 70 years, consecutively admitted from Emergency Department (ED) to participating AGUs with mild-to-moderate frailty. Frailty is defined as a Primary Care–Frailty Index (PC-FI) [12] of 0.07–0.21 and a Clinical Frailty Scale (CFS) [13, 14] of 4–6. Additional inclusion criteria are the ability to walk with or without assistance (Cumulated Ambulation Score [CAS] ≥ 2) [15], the ability to communicate with the research team, and capacity to provide written informed consent. Exclusion criteria are an expected AGU length of stay < 3 days, life expectancy < 12 months, persistent clinical instability (National Early Warning Score 2 [16] > 5 within 48 h of admission and/or delirium), and any other contraindications to participation judged by the investigators (including, for example, a prolonged interval between hospital admission and eligibility assessment, due to transfer from another ward, that would prevent timely initiation of the intervention, or the inability to comply with study procedures due to sudden complications not captured by the other prespecified exclusion criteria). A full list of inclusion and exclusion criteria is reported in Table 1.Table 1Inclusion and exclusion criteriaInclusion criteria Age ≥ 70 years Mild to moderate frailty, defined as: PC-FI (Primary Care Frailty Index) between 0.07 and 0.21 and Clinical Frailty Scale (CFS) score 4–6 Ability to walk with or without assistance Ability to communicate and cooperate with the research team Ability to provide written informed consentExclusion criteria Expected hospital stay < 3 days Estimated life expectancy < 12 months Persistent clinical instability: NEWS2 score > 5 (within 48 hours of admission) and/or delirium History of schizophrenia, major depression, bipolar disorder, or psychosis Severe sensory deficits (visual and/or auditory) Presence of a nasogastric tube or percutaneous endoscopic gastrostomy (PEG) Severe cardiac conduction disorder (e.g. third-degree AV block), uncontrolled arrhythmia, new Q wave in the last 6 months, or ST-segment depression > 3 mm on ECG Terminal oncological or organic disease (e.g. Child-Pugh C cirrhosis, stage V renal disease, GOLD stage D COPD, end-stage heart failure) Residence in a nursing home prior to the index hospital admission Dissecting aortic aneurysm, severe aortic stenosis, acute endocarditis/pericarditis, acute thromboembolism, uncontrolled hypertension (>180/100 mmHg) or postural hypotension, uncontrolled hypoglycaemia, recent fracture within the last month, or any other medical condition that, in the clinician’s judgment, precludes safe physical activity Any condition preventing safe participation in the intervention and/or cooperation with the study procedures Concurrent participation in another clinical study Clinical conditions or situations which, in the investigator’s opinion, could interfere with the study or prevent optimal participation Refusal to participate in the study Table 1Inclusion and exclusion criteriaInclusion criteria Age ≥ 70 years Mild to moderate frailty, defined as: PC-FI (Primary Care Frailty Index) between 0.07 and 0.21 and Clinical Frailty Scale (CFS) score 4–6 Ability to walk with or without assistance Ability to communicate and cooperate with the research team Ability to provide written informed consentExclusion criteria Expected hospital stay < 3 days Estimated life expectancy < 12 months Persistent clinical instability: NEWS2 score > 5 (within 48 hours of admission) and/or delirium History of schizophrenia, major depression, bipolar disorder, or psychosis Severe sensory deficits (visual and/or auditory) Presence of a nasogastric tube or percutaneous endoscopic gastrostomy (PEG) Severe cardiac conduction disorder (e.g. third-degree AV block), uncontrolled arrhythmia, new Q wave in the last 6 months, or ST-segment depression > 3 mm on ECG Terminal oncological or organic disease (e.g. Child-Pugh C cirrhosis, stage V renal disease, GOLD stage D COPD, end-stage heart failure) Residence in a nursing home prior to the index hospital admission Dissecting aortic aneurysm, severe aortic stenosis, acute endocarditis/pericarditis, acute thromboembolism, uncontrolled hypertension (>180/100 mmHg) or postural hypotension, uncontrolled hypoglycaemia, recent fracture within the last month, or any other medical condition that, in the clinician’s judgment, precludes safe physical activity Any condition preventing safe participation in the intervention and/or cooperation with the study procedures Concurrent participation in another clinical study Clinical conditions or situations which, in the investigator’s opinion, could interfere with the study or prevent optimal participation Refusal to participate in the study Inclusion and exclusion criteria Outcomes Primary endpoint The primary endpoint is the change in motor performance at hospital discharge, assessed using the validated physical performance test Short Physical Performance Battery (SPPB). The SPPB will be administered at hospital discharge by trained study personnel at each participating centre, who will be blinded to the baseline assessment score and not directly involved in the study procedures of enrolment and delivery of the intervention. Secondary and exploratory endpoints Secondary endpoints include: i) feasibility of the technology-supported MDI (assessed by adherence, usability and retention) throughout hospitalisation and the 12-week post-discharge period; ii) effect of the MDI on key in-hospital and post-discharge outcomes, namely delirium incidence, length of stay and in-hospital complications, cognitive performance, nutritional status, frailty, falls, sleep quality, hospital readmissions and mortality at follow-up, health-related quality of life. Exploratory endpoints will examine the broader impact of the intervention on healthcare resource utilisation and costs, biomarkers of senescence, neurodegeneration, inflammaging and frailty, as well as on mood and affective status, social relationships, perceived social connectedness, and access to and use of digital resources. Table 2 summarises the study outcomes, the instruments used for their assessment, and the corresponding time points.Table 2Outcomes, assessment tools and time-pointsOutcome and assessment toolBaselineDuring hospitalisationDischarge3-month follow-up6-month follow-upSocio-demographic dataX--XXSomatic health (medical history, Charlson Comorbidity Index, current medications)X--XXMotor function (SPPB; Tinetti)X-XXXFeasibility home-based MDI (PREMs) with adherence/retention---XXFrailty (CFS, PC-FI)X--XXSenescence, neurodegeneration, inflammatory burden and frailty blood biomarkers[Albumin, Transferrin, IL-6 (Interleukin-6), FT3 (Free Triiodothyronine)/FT4 (Free Thyroxine) ratio, NF-L (Neurofilament Light Chain), s-RAGE (soluble Receptor for Advanced Glycation End-products), FGF-1 (Fibroblast Growth Factor 1), GDF-15 (Growth Differentiation Factor 15), Aβ (A-beta), p-tau181, p-tau217, p-tau231]--X-XIncident delirium (4AT, DSM-5)-X---In-hospital complications-X---Sleep quality (PSQI)XX-XXFunctional performance (ADL, IADL, Barthel)X--XXCognitive performance (CRIq, ACE-R, TMT A/B)X-XXXDietary intake & nutritional status (MDS, MNA-SF)X--XXMood (GDS)X--XXSocial relationships (Lubben Social Network Scale)X--XXAcceptability of technological solutions-XXXXHealthcare and social resource utilisation for cost analyses-XXXXLength of hospital stay--X--Falls (number)-X-XXReadmissions---XXQuality of life (EQ-5D-5L)XXMortalityXXXAbbreviations: SPPB Short Physical Performance Battery, PREMs Patient-Reported Experience Measures, CFS Clinical Frailty Scale, PC-FI Primary Care Frailty Index, 4AT 4 A’s Test, DSM-5 Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, PSQI Pittsburgh Sleep Quality Index, ADL Activities of Daily Living, IADL Instrumental Activities of Daily Living, CRIq Cognitive Reserve Index questionnaire, ACE-R Addenbrooke’s Cognitive Examination – Revised, TMT A/B Trail Making Test parts A and B, MDS Mediterranean Diet Scale, MNA-SF Mini Nutritional Assessment – Short Form, GDS Geriatric Depression Scale, EQ-5D-5L EuroQol 5-Dimension 5-Level questionnaire Table 2Outcomes, assessment tools and time-pointsOutcome and assessment toolBaselineDuring hospitalisationDischarge3-month follow-up6-month follow-upSocio-demographic dataX--XXSomatic health (medical history, Charlson Comorbidity Index, current medications)X--XXMotor function (SPPB; Tinetti)X-XXXFeasibility home-based MDI (PREMs) with adherence/retention---XXFrailty (CFS, PC-FI)X--XXSenescence, neurodegeneration, inflammatory burden and frailty blood biomarkers[Albumin, Transferrin, IL-6 (Interleukin-6), FT3 (Free Triiodothyronine)/FT4 (Free Thyroxine) ratio, NF-L (Neurofilament Light Chain), s-RAGE (soluble Receptor for Advanced Glycation End-products), FGF-1 (Fibroblast Growth Factor 1), GDF-15 (Growth Differentiation Factor 15), Aβ (A-beta), p-tau181, p-tau217, p-tau231]--X-XIncident delirium (4AT, DSM-5)-X---In-hospital complications-X---Sleep quality (PSQI)XX-XXFunctional performance (ADL, IADL, Barthel)X--XXCognitive performance (CRIq, ACE-R, TMT A/B)X-XXXDietary intake & nutritional status (MDS, MNA-SF)X--XXMood (GDS)X--XXSocial relationships (Lubben Social Network Scale)X--XXAcceptability of technological solutions-XXXXHealthcare and social resource utilisation for cost analyses-XXXXLength of hospital stay--X--Falls (number)-X-XXReadmissions---XXQuality of life (EQ-5D-5L)XXMortalityXXXAbbreviations: SPPB Short Physical Performance Battery, PREMs Patient-Reported Experience Measures, CFS Clinical Frailty Scale, PC-FI Primary Care Frailty Index, 4AT 4 A’s Test, DSM-5 Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, PSQI Pittsburgh Sleep Quality Index, ADL Activities of Daily Living, IADL Instrumental Activities of Daily Living, CRIq Cognitive Reserve Index questionnaire, ACE-R Addenbrooke’s Cognitive Examination – Revised, TMT A/B Trail Making Test parts A and B, MDS Mediterranean Diet Scale, MNA-SF Mini Nutritional Assessment – Short Form, GDS Geriatric Depression Scale, EQ-5D-5L EuroQol 5-Dimension 5-Level questionnaire Outcomes, assessment tools and time-points Senescence, neurodegeneration, inflammatory burden and frailty blood biomarkers [Albumin, Transferrin, IL-6 (Interleukin-6), FT3 (Free Triiodothyronine)/FT4 (Free Thyroxine) ratio, NF-L (Neurofilament Light Chain), s-RAGE (soluble Receptor for Advanced Glycation End-products), FGF-1 (Fibroblast Growth Factor 1), GDF-15 (Growth Differentiation Factor 15), Aβ (A-beta), p-tau181, p-tau217, p-tau231] Abbreviations: SPPB Short Physical Performance Battery, PREMs Patient-Reported Experience Measures, CFS Clinical Frailty Scale, PC-FI Primary Care Frailty Index, 4AT 4 A’s Test, DSM-5 Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, PSQI Pittsburgh Sleep Quality Index, ADL Activities of Daily Living, IADL Instrumental Activities of Daily Living, CRIq Cognitive Reserve Index questionnaire, ACE-R Addenbrooke’s Cognitive Examination – Revised, TMT A/B Trail Making Test parts A and B, MDS Mediterranean Diet Scale, MNA-SF Mini Nutritional Assessment – Short Form, GDS Geriatric Depression Scale, EQ-5D-5L EuroQol 5-Dimension 5-Level questionnaire Randomisation and allocation concealment An independent statistician generated the allocation sequence using permuted blocks with randomly varying block sizes (2 and 4) to maintain overall 1:1 balance while limiting predictability. No stratification or covariate constraints were applied. The allocation was stored in an access-restricted repository and released simultaneously to all principal investigators only after completion of joint training and formal site activation. Assignment notifications were issued in a single batched communication, and patient screening and enrolment commenced thereafter. Each ward maintains a consecutive screening log with timestamps and reasons for exclusion or refusal. Standardised eligibility checklists are used, and staff responsible for screening are operationally distinct from those delivering the MDI. Recruitment, comprehensive geriatric assessment and blood biomarkers Participants are recruited in the participating AGUs by trained site clinicians during the index hospitalisation. To ensure adequate enrolment, centres implement standardised pre-screening of eligible admissions and health records, provide comprehensive study information materials, and minimise participant burden by aligning assessments with routine clinical workflows whenever possible. At enrolment, clinicians perform a baseline Comprehensive Geriatric Assessment (CGA) for all participants in both arms, collecting the following variables (Table 2).i)Socio-demographic data: age, sex, marital status, living arrangements, social interactions;ii)Health status: medical history, Charlson Comorbidity Index (CCI, ranging from 0 to 37) [17], and usual chronic medications;iii)Functional status: Basic Activities of Daily Living (ADL, ranging from 0 to 6) [18]; Instrumental Activities of Daily Living (IADL, ranging from 0 to 8) [19]; Barthel Index (ranging from 0 to 100) mobility [20]; and handgrip strength [21, 22];iv)Motor function: Tinetti Scale (ranging from 0 to 28) [23] and SPPB, ranging from 0 to 12. The SPPB quantifies physical performance through standardised tests of standing balance, 4-m walking speed and repeated chair stands, yielding a composite score that reflects lower-extremity function and mobility. In older adults, SPPB scores ≥ 10 indicate preserved physical performance, whereas lower values identify increasing degrees of mobility impairment and vulnerability [24];v)Nutritional status: Mini Nutritional Assessment-Short Form (MNA-SF) screening tool, (ranging from 0 to 14) to assess the risk of malnutrition [25]; adherence to Mediterranean Diet Scale (MDS) (ranging from 0 to 9) [26, 27];vi)Cognition: Addenbrooke’s cognitive examination Revised (ACE-R, ranging from 0 to 100) [28]; Trail Making Test (TMT) parts A and B [29] and Cognitive Reserve Index questionnaire (CRIq) [30];vii)Delirium: screened daily with the 4AT [31], and confirmed according to the Diagnostic and Statistical Manual of mental disorders 5th edition (DSM-5) criteria [32];viii)Psychological status: 5- items Geriatric Depression Scale (GDS, ranging from 0 to 5) [33];ix)Social network: Lubben Social Network Scale (ranging from 0 to 30) [34];x)Quality of life: EuroQol (EQ-5D-5L) to measure the health profile on five dimensions, each scored 1 to 5, converted into a single utility value from 0 to 1 using country-specific normative value sets [35];xi)Sleep quality and disturbances: Pittsburgh Sleep Quality Index (PSQI; ranging from 0 to 21)[36]. i)Socio-demographic data: age, sex, marital status, living arrangements, social interactions; Socio-demographic data: age, sex, marital status, living arrangements, social interactions; ii)Health status: medical history, Charlson Comorbidity Index (CCI, ranging from 0 to 37) [17], and usual chronic medications; Health status: medical history, Charlson Comorbidity Index (CCI, ranging from 0 to 37) [17], and usual chronic medications; iii)Functional status: Basic Activities of Daily Living (ADL, ranging from 0 to 6) [18]; Instrumental Activities of Daily Living (IADL, ranging from 0 to 8) [19]; Barthel Index (ranging from 0 to 100) mobility [20]; and handgrip strength [21, 22]; Functional status: Basic Activities of Daily Living (ADL, ranging from 0 to 6) [18]; Instrumental Activities of Daily Living (IADL, ranging from 0 to 8) [19]; Barthel Index (ranging from 0 to 100) mobility [20]; and handgrip strength [21, 22]; iv)Motor function: Tinetti Scale (ranging from 0 to 28) [23] and SPPB, ranging from 0 to 12. The SPPB quantifies physical performance through standardised tests of standing balance, 4-m walking speed and repeated chair stands, yielding a composite score that reflects lower-extremity function and mobility. In older adults, SPPB scores ≥ 10 indicate preserved physical performance, whereas lower values identify increasing degrees of mobility impairment and vulnerability [24]; Motor function: Tinetti Scale (ranging from 0 to 28) [23] and SPPB, ranging from 0 to 12. The SPPB quantifies physical performance through standardised tests of standing balance, 4-m walking speed and repeated chair stands, yielding a composite score that reflects lower-extremity function and mobility. In older adults, SPPB scores ≥ 10 indicate preserved physical performance, whereas lower values identify increasing degrees of mobility impairment and vulnerability [24]; v)Nutritional status: Mini Nutritional Assessment-Short Form (MNA-SF) screening tool, (ranging from 0 to 14) to assess the risk of malnutrition [25]; adherence to Mediterranean Diet Scale (MDS) (ranging from 0 to 9) [26, 27]; Nutritional status: Mini Nutritional Assessment-Short Form (MNA-SF) screening tool, (ranging from 0 to 14) to assess the risk of malnutrition [25]; adherence to Mediterranean Diet Scale (MDS) (ranging from 0 to 9) [26, 27]; vi)Cognition: Addenbrooke’s cognitive examination Revised (ACE-R, ranging from 0 to 100) [28]; Trail Making Test (TMT) parts A and B [29] and Cognitive Reserve Index questionnaire (CRIq) [30]; Cognition: Addenbrooke’s cognitive examination Revised (ACE-R, ranging from 0 to 100) [28]; Trail Making Test (TMT) parts A and B [29] and Cognitive Reserve Index questionnaire (CRIq) [30]; vii)Delirium: screened daily with the 4AT [31], and confirmed according to the Diagnostic and Statistical Manual of mental disorders 5th edition (DSM-5) criteria [32]; Delirium: screened daily with the 4AT [31], and confirmed according to the Diagnostic and Statistical Manual of mental disorders 5th edition (DSM-5) criteria [32]; viii)Psychological status: 5- items Geriatric Depression Scale (GDS, ranging from 0 to 5) [33]; Psychological status: 5- items Geriatric Depression Scale (GDS, ranging from 0 to 5) [33]; ix)Social network: Lubben Social Network Scale (ranging from 0 to 30) [34]; Social network: Lubben Social Network Scale (ranging from 0 to 30) [34]; x)Quality of life: EuroQol (EQ-5D-5L) to measure the health profile on five dimensions, each scored 1 to 5, converted into a single utility value from 0 to 1 using country-specific normative value sets [35]; Quality of life: EuroQol (EQ-5D-5L) to measure the health profile on five dimensions, each scored 1 to 5, converted into a single utility value from 0 to 1 using country-specific normative value sets [35]; xi)Sleep quality and disturbances: Pittsburgh Sleep Quality Index (PSQI; ranging from 0 to 21)[36]. Sleep quality and disturbances: Pittsburgh Sleep Quality Index (PSQI; ranging from 0 to 21)[36]. Blood samples are also collected from all participants at discharge and at 6-month follow-up to quantify a broad panel of biomarkers related to senescence, neurodegeneration, inflammation, and frailty. Multidomain intervention The intervention group receive personalised MDI delivered by a multidisciplinary team. The MDI begins soon after clinical stabilisation and continues for 12 weeks after discharge. During their AGU stay, participants are equipped with ICTs, including a tracker for physical activity monitoring and a tablet for computerised cognitive training. As illustrated in Fig. 2, the MDI is tailored to the needs of each patient (and their study partner, if present):Multicomponent physical exercise: after having signed informed consent, the physiotherapist assesses patient’s physical performance using the SPPB. Based on the SPPB score, the appropriate Vivifrail training level (A–D) is assigned and a personalised exercise plan (“Vivifrail Passport”) is provided [37, 38]. The programme includes two 30-min sessions per day: one supervised session in the morning and one self-administered session in the afternoon, targeting strength, endurance, balance, and coordination. Physical activity is monitored via accelerometers (Garmin Vivosmart 5 tracker) and supported with a physical activity kit (e.g., fitball, elastic bands).Cognitive intervention and digital training: between the first and second day after enrolment, participants in the intervention arm receive a tablet (Samsung Galaxy Tab). Basic tablet functions (e.g. video calls, camera, internet access) are demonstrated before training in BrainHQ [39], a computerised cognitive training platform selected by the neuropsychology team and validated in older adults with cognitive impairment. BrainHQ provides interactive exercises across six domains—Attention, Brain Speed, Memory, People Skills, Intelligence, and Navigation — targeting processing speed, working and episodic memory, social cognition, reasoning, and spatial orientation. The platform automatically adapts task difficulty to individual performance. Training sessions are personalised and supervised by a neuropsychologist, with continuous support from the digital coach, who assists with device use, monitors adherence, and provides technical support.Nutritional intervention: a dietitian evaluates nutritional status, identifies risk of malnutrition, and develops an individualised dietary plan based on Mediterranean diet principles. Structured nutritional education sessions involve both participants and study partners, focusing on balanced meal composition, hydration, and strategies to maintain adequate energy and protein intake during recovery [40, 41].Health education and environmental interventions: a nurse case manager assesses participants’ needs and implements environmental modifications and behavioural strategies to promote healthy habits, fall prevention, and sleep hygiene, including guidance on maintaining these behaviours after discharge [42]. Moreover, a geriatrician performs medication reconciliation using STOPP/START criteria [43] and optimises therapies according to individual risk–benefit profiles. Any changes are communicated to general practitioner.Social engagement: to foster socialisation and an age-friendly environment, group activities and environmental changes (e.g. extended visiting hours) are implemented throughout the intervention tailored to the patient’s individual needs. Any changes are communicated to general practitioner.Fig. 2Focus on multidomain intervention arm during hospitalisation Multicomponent physical exercise: after having signed informed consent, the physiotherapist assesses patient’s physical performance using the SPPB. Based on the SPPB score, the appropriate Vivifrail training level (A–D) is assigned and a personalised exercise plan (“Vivifrail Passport”) is provided [37, 38]. The programme includes two 30-min sessions per day: one supervised session in the morning and one self-administered session in the afternoon, targeting strength, endurance, balance, and coordination. Physical activity is monitored via accelerometers (Garmin Vivosmart 5 tracker) and supported with a physical activity kit (e.g., fitball, elastic bands). Multicomponent physical exercise: after having signed informed consent, the physiotherapist assesses patient’s physical performance using the SPPB. Based on the SPPB score, the appropriate Vivifrail training level (A–D) is assigned and a personalised exercise plan (“Vivifrail Passport”) is provided [37, 38]. The programme includes two 30-min sessions per day: one supervised session in the morning and one self-administered session in the afternoon, targeting strength, endurance, balance, and coordination. Physical activity is monitored via accelerometers (Garmin Vivosmart 5 tracker) and supported with a physical activity kit (e.g., fitball, elastic bands). Cognitive intervention and digital training: between the first and second day after enrolment, participants in the intervention arm receive a tablet (Samsung Galaxy Tab). Basic tablet functions (e.g. video calls, camera, internet access) are demonstrated before training in BrainHQ [39], a computerised cognitive training platform selected by the neuropsychology team and validated in older adults with cognitive impairment. BrainHQ provides interactive exercises across six domains—Attention, Brain Speed, Memory, People Skills, Intelligence, and Navigation — targeting processing speed, working and episodic memory, social cognition, reasoning, and spatial orientation. The platform automatically adapts task difficulty to individual performance. Training sessions are personalised and supervised by a neuropsychologist, with continuous support from the digital coach, who assists with device use, monitors adherence, and provides technical support. Cognitive intervention and digital training: between the first and second day after enrolment, participants in the intervention arm receive a tablet (Samsung Galaxy Tab). Basic tablet functions (e.g. video calls, camera, internet access) are demonstrated before training in BrainHQ [39], a computerised cognitive training platform selected by the neuropsychology team and validated in older adults with cognitive impairment. BrainHQ provides interactive exercises across six domains—Attention, Brain Speed, Memory, People Skills, Intelligence, and Navigation — targeting processing speed, working and episodic memory, social cognition, reasoning, and spatial orientation. The platform automatically adapts task difficulty to individual performance. Training sessions are personalised and supervised by a neuropsychologist, with continuous support from the digital coach, who assists with device use, monitors adherence, and provides technical support. Nutritional intervention: a dietitian evaluates nutritional status, identifies risk of malnutrition, and develops an individualised dietary plan based on Mediterranean diet principles. Structured nutritional education sessions involve both participants and study partners, focusing on balanced meal composition, hydration, and strategies to maintain adequate energy and protein intake during recovery [40, 41]. Nutritional intervention: a dietitian evaluates nutritional status, identifies risk of malnutrition, and develops an individualised dietary plan based on Mediterranean diet principles. Structured nutritional education sessions involve both participants and study partners, focusing on balanced meal composition, hydration, and strategies to maintain adequate energy and protein intake during recovery [40, 41]. Health education and environmental interventions: a nurse case manager assesses participants’ needs and implements environmental modifications and behavioural strategies to promote healthy habits, fall prevention, and sleep hygiene, including guidance on maintaining these behaviours after discharge [42]. Moreover, a geriatrician performs medication reconciliation using STOPP/START criteria [43] and optimises therapies according to individual risk–benefit profiles. Any changes are communicated to general practitioner. Health education and environmental interventions: a nurse case manager assesses participants’ needs and implements environmental modifications and behavioural strategies to promote healthy habits, fall prevention, and sleep hygiene, including guidance on maintaining these behaviours after discharge [42]. Moreover, a geriatrician performs medication reconciliation using STOPP/START criteria [43] and optimises therapies according to individual risk–benefit profiles. Any changes are communicated to general practitioner. Social engagement: to foster socialisation and an age-friendly environment, group activities and environmental changes (e.g. extended visiting hours) are implemented throughout the intervention tailored to the patient’s individual needs. Any changes are communicated to general practitioner. Social engagement: to foster socialisation and an age-friendly environment, group activities and environmental changes (e.g. extended visiting hours) are implemented throughout the intervention tailored to the patient’s individual needs. Any changes are communicated to general practitioner. Focus on multidomain intervention arm during hospitalisation After hospital discharge, participants in the intervention group continue the MDI at home for 12 weeks through remote support. Under the supervision of the multidisciplinary team (together with the study partner, whether available, being encouraged but not mandatory for study participation), the programme includes: i) 12-week remote physiotherapy sessions, along with self-administered physical activity according to the assigned Vivifrail Passport; ii) three self-administered BrainHQ cognitive stimulation sessions per week; iv) adherence to the personalised diet and lifestyle recommendations. Potential harms are expected to be minimal and mainly related to physical exercise (fatigue, soreness, falls) and use of digital tools (stress, frustration), and will be monitored through systematic adverse-event reporting. Usual care (control group) Participants randomised to the control arm receive usual care. This includes medical management of acute and chronic conditions, routine nursing care, and rehabilitation, nutritional or psychological consultations only when requested by the team. No structured multidomain program and no systematic use of ICTs are provided to support cognitive function. At discharge, participants receive standard written information along with a leaflet containing generic physical, nutritional, and behavioural recommendations. No organised home-based follow-up or remote coaching is offered. Control participants are re-evaluated only at the 6-month follow-up, using the same outcome measures as the intervention group to allow comparison of clinical, functional, cognitive, and economic endpoints. Feasibility, retention, adherence, and acceptability Feasibility and acceptability are evaluated at both patient and site level during hospitalisation and throughout the home-based phase, using predefined domains captured in the electronic case-report form. The “reach” proportion is defined as the proportion of eligible patients who provide informed consent and are enrolled, while the “uptake” is defined as the proportion of participants in the intervention arm who receive at least one core MDI component (physical, nutritional or cognitive) within 48 h of enrolment. These indicators quantify the initial penetration and operational implementation of the intervention across wards. “Retention” is defined as the proportion of enrolled participants who complete the intervention and/or all planned follow-up assessments. Retention ≥ 60% is considered acceptable and ≥ 70% optimal; values < 60% indicate substantial implementation challenges or excessive burden. “Adherence” during hospitalisation is monitored daily through structured logs that record whether each scheduled component (two physical sessions and one cognitive session per day) is completed, partially completed or missed, together with reasons for non-completion (e.g. clinical instability, logistical issues, refusal). For each participant, adherence is expressed as the proportion of eligible days on which all mandatory components are delivered, and site-level adherence is defined as the median of patient-level adherence within each cluster. Across the whole programme, adherence is calculated as the proportion of prescribed sessions that are completed for all components of the MDI (physical exercise, cognitive stimulation, nutritional counselling and digital engagement). Participation in ≥ 60% of scheduled activities, or completion of at least two-thirds of the protocol, are considered acceptable, while adherence ≥ 75% optimal. Domain-specific variability in adherence is anticipated (e.g. higher adherence to physical exercise, moderate adherence to cognitive modules, slightly lower adherence to nutritional components). Fidelity of intervention delivery is monitored using standardised checklists documenting timing, duration and protocol compliance. At 3 and 6 months, structured interviews with participants and study partners are conducted to collect information on attendance at online physiotherapy sessions, frequency of home exercises, adherence to the nutritional plan and reasons for reduced participation (e.g. complications, rehospitalisation, logistical barriers). “Acceptability and usability” of the technological components are assessed with the System Usability Scale (SUS), with scores ≥ 68 indicating acceptable usability and scores ≥ 80 excellent usability and strong perceived benefit [44]. Additional acceptability data are obtained from brief satisfaction questionnaires (5-point Likert scales) administered at discharge and during follow-up visits. “Technical feasibility” is defined as the proportion of participants able to use the assigned digital devices (tablet, app, tracker) autonomously or with minimal support. A proportion ≥ 65% is considered acceptable, given the advanced age and frailty of the population, and ≥ 80% optimal. At the end of the trial, an “integrated feasibility benchmark” will be calculated as the proportion of predefined feasibility domains meeting their respective thresholds. The intervention will be considered feasible if (i) > 50% of participants complete ≥ 80% of prescribed sessions and (ii) at least 70% of key process indicators (recruitment, adherence, retention, acceptability/usability and technical feasibility) reach their acceptable cut-offs. This composite assessment will provide a pragmatic evaluation of the implementability of the OPTIMAge-IT MDI. No independent Data Monitoring Committee (DMC) was convened because this pragmatic study evaluates a low-risk, non-pharmacological intervention and does not include formal stopping rules or interim analyses. Adverse events, discontinuations and protocol deviations are recorded at each site and reviewed regularly by the coordinating centre. Follow-up procedures Participants in the intervention arm are followed up at 3 and 6 months after discharge; those in the control arm are assessed at 6 months only. This schedule reflects the 12-week duration of the Vivifrail programme, which constitutes the active phase of the home-MDI. The 3-month visit therefore focuses mainly on feasibility, adherence, and user experience with digital and home-based components. At each follow-up, a CGA is repeated to track changes in motor, cognitive, functional, and nutritional domains. In the intervention group, feasibility and acceptability of the ICT-supported approach are assessed using Patient-Reported Experience Measures (PREMs) and the SUS at 3 months. Home-based adherence and overall retention are also evaluated. For all participants, follow-up data collection includes drop-out, falls, emergency visits, re-hospitalisations, and medical consultations. Global health-care utilisation and direct medical costs are recorded. Social and economic costs are also analysed to estimate broader resource use in the study population. The trial will conclude when the last enrolled participant completes the 6-month follow-up. Training, harmonisation of procedures, and multidisciplinary team Training activities are delivered via webinars and in-person meetings to ensure consistency in clinical assessments, intervention delivery, and data entry across centres. Within the intervention arm, a multidisciplinary team –comprising a technology expert (“digital coach”), physiotherapist, dietitian, neuropsychologist, geriatrician, and nurse specialised in geriatric care–collaboratively delivers the intervention and monitors adherence during hospitalisation. To enhance public engagement and raise awareness of the project, information leaflets are disseminated, newspaper articles written, and press conferences organised. A dedicated publicly accessible project website (https://lisa.unimib.it/) provides comprehensive information about the study and includes a participant-oriented section. Economic evaluation and cost-effectiveness analysis An exploratory economic evaluation will be conducted to describe resource utilisation, implementation-related costs, and their relationship with clinical and quality-of-life outcomes. Resource-use information is collected prospectively over 6 months after discharge, using structured forms both in the intervention and control arm. Healthcare resource utilisation includes hospital readmissions, emergency department visits, outpatient and specialist consultations, nursing visits, physiotherapy or rehabilitation sessions, and other relevant healthcare contacts during follow-up. In the intervention arm, detailed monitoring forms are also completed by the healthcare professionals to document intervention delivery, including professional time, digital support, physiotherapy supervision, nutritional counselling, cognitive training support, technical assistance, and other resources required to support the home-based phase of the intervention. Informal care needs are explored through dedicated form sections, recording the presence and approximate intensity of caregiver support related to personal care, mobility, medication management, nutrition, transportation, and instrumental activities of daily living during the follow-up period. During the home-based phase, the study partner (usually a family member or another trusted person involved in the patient’s daily care), where available, supports reporting of healthcare contacts and care needs. In participants without a study partner, information is collected directly from the participant and complemented by other clinical sources. As for the economic analyses, total costs are estimated as the sum of direct medical, direct non-medical, and indirect costs related to informal caregiving. Incremental costs will also be related to changes in physical function and health-related quality of life: the incremental cost-effectiveness ratio will be estimated as the ratio between mean cost differences and mean effect differences between intervention and control arms, with effectiveness expressed as change in physical function and quality-adjusted life years (QALYs) derived from EQ-5D scores. Sensitivity analyses will examine the robustness of results under alternative cost assumptions and missing-data scenarios. Statistical analysis Sample size calculation We estimated that a total sample size of 240 participants, recruited across eight Italian AGUs, would provide 80% power to detect as statistically significant a between-group difference of 1.8 points in the mean change of SPPB score at hospital discharge. Assuming an average cluster size of 30 participants recruited over a 3-month period, the total sample size was inflated to 288 to account for an anticipated 20% attrition. The assumed effect size of 1.8 points was based on the results proposed by Martínez-Velilla et al., who observed a larger between-group difference (2.2 points; 95% CI 1.7 to 2.6) [45]. The calculation assumed an intraclass correlation coefficient of 0.10, a standard deviation of SPPB of 2.5, and a two-sided type I error of 0.05. Sample size computations were performed using PASS 2021 (NCSS, Kaysville, UT, USA). Statistical analysis plan Clinical research staff at each participating centre will enter all pseudonymised study data into a secure REDCap Cloud platform in compliance with applicable privacy and data protection laws and regulations, and statistical analyses will be performed centrally at the coordinating institution. Data monitoring will be performed by the coordinating centre using risk-based procedures, including central checks for completeness and consistency, validation rules, and resolution of data queries with participating centres. Descriptive statistics will be used to summarise baseline characteristics and outcomes in both study arms. All primary analyses will follow the intention-to-treat (ITT) principle. The main clinical endpoint (change in SPPB score from admission to discharge) will be analysed using mixed linear models, including a random effect for cluster, a fixed effect for treatment group, and, where appropriate, fixed effects for prespecified covariates at both cluster and patient level. Baseline SPPB will be included as a covariate to account for regression to the mean. The same modelling framework will be applied to other continuous longitudinal outcomes, with treatment group, time, and their interaction specified as fixed effects. Both clusters and individual patients will be treated as random effects. Among alternative variance–covariance structures for repeated measures within subjects, the one yielding the lowest Akaike Information Criterion (AIC) will be selected. Missing data will be addressed using maximum likelihood estimation in the mixed-effects models under a missing-at-random assumption, with sensitivity analyses using multiple imputation if warranted. A p-value < 0.05 will be considered statistically significant. All analyses will be performed using R version 4.3.2 (R Foundation for Statistical Computing, Vienna, Austria), Stata 18 (StataCorp LLC, College Station, TX, USA), or SAS software (SAS Institute, Cary, NC, USA). Discussion The OPTIMAge-IT study evaluates a novel model of hospital-based multidomain care that combines CGA-driven interventions with digital technologies to support continuity, adherence and monitoring across the transition from hospital to home. By extending the intervention beyond the inpatient setting, the trial addresses a key limitation of previous hospital-based programmes, which have largely been confined to the ward despite evidence that functional trajectories are shaped across the entire early recovery phase. OPTIMAge-IT builds on a robust body of evidence showing that multidomain interventions can preserve physical performance and reduce hospital-associated disability[5, 45], but it introduces several important innovations. First, it is the first multicentre trial to enrol acutely hospitalised older adults on the basis of objective frailty criteria, operationalised through validated instruments such as the CFS and the PC-FI. By focusing on patients with mild-to-moderate frailty, who are at increased risk of hospital-acquired disability yet remain responsive to rehabilitative interventions, the study addresses a population in which multidomain strategies are most likely to be effective. In this context, OPTIMAge-IT is designed not only to test efficacy but also to characterise how frailty-modifying strategies perform across diverse acute-care environments. The enrolment of patients from eight geriatric units with different organisational and socio-economic profiles further enhances the relevance of the findings to hospital practice. In contrast to many previous programmes, which were monocentric [46] or reliant on specialised gym-based equipment [45], OPTIMAge-IT adopts a low-cost, scalable model using simple, portable materials (ankle and wrist weights, elastic bands, stress balls) embedded within the multicomponent Vivifrail exercise protocol. This design facilitates integration into routine ward activities and supports safe continuation at home under study partner supervision, enhancing the potential for dissemination across heterogeneous healthcare settings. Conceptually, the trial embraces a truly MDI. Whereas earlier interventions often focused on one or two domains [5, 47], this study is aligned with the multidimensional philosophy of the ICOPE and FINGER models, addressing physical, cognitive, nutritional, social and medical domains simultaneously. The cognitive component exemplifies this integration: participants engage in digital cognitive training via the BrainHQ platform, guided by a neuropsychologist and supported by a digital coach who promotes both adherence and digital literacy. This structure allows the study not only to test clinical efficacy but also to examine acceptability and user experience of digital tools in a complex hospital environment, aspects that are seldom explored in rehabilitation research. The nutritional and physical components are synergistically designed to counteract sarcopenia and malnutrition by reinforcing adherence to Mediterranean diet principles and delivering structured reconditioning based on Vivifrail. Social engagement is fostered through supervised in-hospital sessions and remote contacts after discharge, tailored to individual levels of fitness. These strategies aim to sustain motivation, preserve autonomy and strengthen social connectedness, key determinants of recovery and resilience in older adults. The 6-month follow-up will provide information on feasibility, adherence and short-term clinical outcomes, and will help identify barriers and facilitators to implementing MDIs within routine geriatric pathways. Another distinctive feature of OPTIMAge-IT is the collection of circulating biomarkers, enabling a deeper understanding of the biological processes underlying frailty and its interaction with acute illness and functional trajectories. Rather than serving purely for risk stratification, these measures are designed to help elucidate mechanisms linking systemic stress, resilience and recovery in hospitalised older adults. Combined biological, clinical and digital phenotyping—including continuous monitoring via trackers—will support clustering and exploratory AI-based analyses aimed at refining patient profiles and anticipating trajectories of recovery. Such findings may contribute to the development of precision geriatric medicine, in which personalised rehabilitation strategies are guided by integrated biological and functional data [48]. The planned economic evaluation will quantify the cost and potential savings associated with the intervention, informing a shift in resource allocation from standard hospital care towards approaches aimed at preventing functional decline and disability. These data will be highly relevant for policymakers seeking scalable, cost-effective models of care. Beyond clinical and economic outcomes, OPTIMAge-IT may help inform the broader agenda of harmonising digital and geriatric care across European health systems. The integration of telemonitoring and AI-assisted tools requires shared technical standards, regulatory frameworks and equitable access. Evidence generated by this trial could therefore support national and European strategies to develop sustainable, person-centred geriatric models that bridge hospital and community services, and will foster further comparative research within collaborative networks such as Age-It. Limitations Some limitations must be acknowledged. First, although a patient-level randomised controlled trial would have been methodologically preferable, it was not feasible in the present setting because hospital rooms in Italy are commonly double-occupancy. Individual randomisation within the same room could have resulted in one patient receiving the intervention while the roommate did not, increasing the risk of contamination and potentially generating dissatisfaction or misunderstandings between participants. Second, recruiting older adults with mild-to-moderate frailty in the absence of overt dementia may be challenging in acute-care settings, where many patients present with advanced clinical complexity and cognitive impairment. Third, implementing digital tools within busy hospital workflows requires ongoing staff training, technical support, and sustained institutional commitment; variability in these factors may affect intervention fidelity across centres. Fourth, the randomized design with specific eligibility criteria may limit extrapolation to broader populations of older adults. Although procedures are standardised, generalisability beyond the Italian context may be limited by differences in infrastructure, care organisation, and digital literacy among patients and professionals. In addition, the predominantly Italian and Caucasian study population may reduce applicability to more diverse settings. Adherence during the home-based phase, including sustained engagement with digital technologies, may also be influenced by study partner availability, socio-economic circumstances, and participants’ comfort with technology. The exploratory economic evaluation should be interpreted with caution because resource-use data may be affected by recall bias and incomplete capture of healthcare or social care contacts, particularly outside the hospital network, potentially leading to under- or overestimation of total resource use in both study arms. Another potential source of bias may derive from incomplete blinding of outcome assessors to group allocation (although outcome assessments will be performed by trained personnel not directly involved in participant enrolment or intervention delivery, and blinded to baseline assessment scores). Finally, while a 6-month follow-up is appropriate for feasibility and short-term outcomes, it may be insufficient to capture longer-term effects on disability, survival, and quality of life. Conclusion OPTIMAge-IT has the potential to redefine the continuum of geriatric care across acute hospital settings. By testing the feasibility and efficacy of a technology-supported MDI, the study aims to shift the hospital stay from a passive episode of care to an active opportunity for recovery. Its findings may inform strategies for preventing hospital-acquired disability, reducing hospital readmissions and embedding digital health and telemonitoring within standard geriatric pathways. Beyond its immediate clinical implications, OPTIMAge-IT offers a translational model for integrating multidimensional assessment, personalised rehabilitation and biological monitoring into everyday practice, thereby contributing to a more biologically informed, multidomain and person-centred vision of ageing well in an ageing society.