Evaluation of a novel technology-supported fall prevention intervention – study protocol of a multi-centre randomised controlled trial in older adults at increased risk of falls

The study design is a multi-site randomised controlled longitudinal intervention study which includes four comprehensive assessments: T₀, T₁, T₂, and T₃ for all participants. All participants are randomly assigned to either the intervention group (INT) or the control group (CTR; see randomisation procedures). Figure 1 shows the SPIRIT-figure, including the schedule of enrolment, interventions, and assessments.

For the intervention group, T₀ is a baseline assessment; T₁ an assessment after finishing the technology-supported part of the intervention; T₂ an assessment after finishing the home-based exercise programme; and T₃ a final follow-up assessment 12 months after T₂. In total, the training period of the intervention group comprises 2*24 units (48 sessions in total) with a training frequency of two planned training sessions of 30 min each per week. The intervention period of the technology-supported exercise programme is individually dependent on the time progress of the completed 24 sessions. Given the defined training frequency of two training sessions per week, the technology-supported exercise programme has a minimum intervention period of 12 weeks. If individual training sessions are cancelled, the intervention period may be extended, however, it is limited to a maximum of 16 weeks. Identical to the intervention period of the technology-supported exercise programme, the period of the home-based programme is then determined individually for each participant. Participants in the control group receive all four measurements. Participants in the control group are not offered any actions; they are instructed to maintain their activities as usual. The time periods between the T₀, T₁ and T₂ measurements are defined as 16 weeks for the control group. Both groups end at T₃ with a final follow-up assessment 12 months after T₂. Figure 2 shows the detailed timeline for study participants. Participants in the control group will have the opportunity to participate in an optimized training programme using the hunova robot after the end of the study.

All assessments, the technology-supported exercise programme (TSEP), as well as two sessions in which participants receive instructions for the home-based exercise programme (HBEP) take place in participating study centres in Cologne, Bonn, Dusseldorf, and Bremen, respectively (addresses are included at the end of this protocol). From each centre, 2-4 sport or physio-therapists take part in the research study. Before the start of the intervention, all participating therapists are trained. This training includes several e-learning (online) modules that must be completed before on-site training begins. All training content of this examination is developed by the German Sport University in cooperation with Movendo Technology. The on-site sessions will be implemented by a researcher from the German Sport University together with a Movendo Technology specialist before the study starts to have a high standardisation and fulfil quality criteria for each training and exercising person.

The target group for the intervention consists of community dwelling older persons aged at least 65 years at an increased risk of falls. Recruitment of potential participants will be through Generali Health Solutions (GHS), who will address insured persons of “Generali Deutschland Krankenversicherung AG” (Generali Germany Health Insurance AG) and its subsidiary. Insured persons receive a personalised invitation letter in which they are asked to participate in the study. If they do not contact any of the study partners within the first two weeks after the initial invitation letter, they will receive a reminder letter. If an insufficient number of appointments at T₀ is made, invited insured persons will receive phone calls. Inclusion criteria: aged at least 65 years; an increased fall risk (defined via a composite risk score (see further text)), ability to walk more than 6 m without assistance, residence within a maximal distance of 20 kms (as the crow flies) to the closest study centre. Exclusion criteria are: a) body height < 1.50 m; body weight > 150 kg (the hunova robot was designed and sized taking into consideration these anthropometric measures for the target population of the device. These choices impacted both the mechanical design and the sizing design of the device); level of care in Germany between 2 and 5 (In Germany, a distinction is made between five levels of care. For classification into a care level, the independence and abilities in six areas of life (mobility, mental and communicative abilities, behaviour and psychological problems, self-care, independent handling of illness- or therapy-related demands and burdens, as well as coping with them, organisation of everyday life and social contacts) are assessed using an instrument based on nursing expertise. Points are awarded for these six areas, which are weighted differently and lead to an overall score (up to a maximum of 100). This score is used for classification into one of the five care levels, which range from care level 1 (minor impairments of independence or abilities) to care level 5 (most severe impairments that are accompanied by special requirements for nursing care). The care level assigned determines the care insurance benefits that the person in need of care receives [39]); incapacity to be physically active over at least 30 min; medical surgery within the last 12 months at the spine, hip, legs, knees, feet (e.g. medical surgery after femoral neck fracture, medical surgery for the purpose of joint replacement or joint stiffening); acute cancer disease including chemotherapy, radiotherapy, biological or palliative therapy; dementia; Parkinson’s disease; multiple sclerosis; epilepsy; cardiovascular disease (e.g. cardiac arrhythmias, heart failure, poorly controlled hypertension, chest pain at rest or on exertion); b) depression, anxiety disorder, schizophrenia, that is not treated properly; consequences of a stroke that impair motor and cognitive functions and/or limitations within speaking substantially; serious respiratory disease (e.g. COPD - chronic obstructive pulmonary disease); Menière’s disease or another disease associated with vertigo; injury or condition that limits the sensitivity and/or motor function of the legs or feet (e.g. spinal stenosis, back pain that radiates to the legs, acute disc prolapse); problems reading information on a screen about 1 m away (e.g. on TV or on a computer screen) -with or without glasses; acute inflammation of the musculoskeletal system (e.g. arthritis, rheumatism in acute attack, ankylosing spondylitis (Bechterew’s disease) in acute attack, acute attack of gout); deep vein thrombosis or "open" leg (e.g. ulcer); the need to wear custom-made orthopaedic footwear (this does not mean shoe inserts) due to sensitivity disorders in the feet; bone fracture due to osteoporosis or noticeable physical changes (sharp decrease in height, increasing development of a hunchback) (list from the health questionnaire). Diseases from a) are “hard” exclusion criteria. If one (or more) of the other b) diseases is present, interested persons are advised to call the Generali health phone hotline to check if a study participation is still possible. Together with the invitation letter, all identified insured persons receive all study materials (detailed study information brochure, health questionnaire, informed consent) by mail. Upon signing the informed consent, they can participate in the study. They will then be tested and, if eligibility is confirmed, then randomly assigned to either INT or CTR. Randomisation procedures, contents of the intervention, as well as the assessments for determining eligibility and evaluating the intervention are described in next subsections.

An equal number of participants in the intervention and control group will be determined for each study centre. For this purpose, all eligible participants draw a lot with either INT or CTR after the baseline assessment. The number of lots for INT and CTR is equal and limited. Drawn lots are not returned to the lottery pot until all lots have been drawn. The examiners of the centres have no influence on the allocation to the groups, a change between groups is not possible.

As part of screening for eligibility, date of birth, gender, body height and weight, data from a health questionnaire, as well as a fall risk profile and current medications, will be collected. The health questionnaire aims to identify the presence of medical conditions (e.g., cardiovascular or neurological disease, high blood pressure, osteoporosis, acute inflammatory conditions, herniated disk, hip fractures, vision deficits). The fall risk profile is based on a composite risk score defined as the “Silver Index” (SI) which is calculated using specific parameters of the hunova robot [35]. The model is presented and validated by Cella et al. [35].

To evaluate intervention effects, a comprehensive assessment strategy which includes the SI, and a mobility test will be repeated at T₀, T₁, T₂, and T₃ (see Fig. 1 and 2). In addition to these repeated assessments, participants in the study will be instructed to keep a fall diary throughout the complete study (i.e., between T₀ and T₃).

Important predictors of fall risk, such as muscle function, balance control and gait quality, can be improved by appropriate fall prevention interventions [10, 11]. Effective fall prevention strategies should simultaneously address different risk factors [10, 12, 13]. Exercise-based strategies are most effective when they combine balance exercises and functional exercises optionally supplemented with strength training [10].

Although several studies in recent years used miscellaneous new technology within novel fall prevention interventions [23‐28], so far, the potential contribution of the hunova robot (Movendo Technology, Genoa, Italy) to reducing fall risk has not been investigated. Based on clinical studies that demonstrated positive results in the rehabilitation of stroke patients as well as in patients with Parkinson’s disease (PD) [37, 38]. Cella et al. [35] demonstrated that the hunova robot can discriminate and evaluate older patients and predict falls. Furthermore, the hunova assessment can discriminate older persons with different grades of impairment of physical performance [50]. With this in mind, the hunova robot seems to be suitable for fall prevention. With its specific assessment strategy which results in an individual fall risk score (SI) and the possibility to create an individually tailored specific exercise programme based on the deficits in the dimensions, the hunova robot provides a new approach to fall prevention. To ensure training success, basic principles of exercise / training science were considered within the training programme.

Based on this planning, we expect positive changes regarding the overall SI but also in the seven subsections (dimensions) addressing different risk factors for falls and for the timed up-and-go test. First positive results on risk factors can be expected after completing the first 24 training sessions using the hunova robot. After this period, an additional 24 training sessions lasting period of home-based training should sustain or even expand the positive changes in the following time. With 48 training sessions over at least 6 months, fall risk factors should positively be influenced in the participants. However, as primary outcomes, the number of falls and fallers within the training period and in a one-year follow-up period are the most relevant parameters. We believe that by applying our training programme, older adults having an increased fall risk (SI > 40%) can experience relevant improvements in functional performance, tested within different dimensions of fall risk, and a reduced number of falls and fallers as primary outcomes, compared to the control group.

Participants who take part in the fall prevention training programme using the hunova robot and the subsequent home-based training programme have potential benefits from an individually tailored training programme that takes basic principles of exercise / training science into account. No adverse events are expected during the measurements or the intervention using the hunova robot and the first two training sessions of the home-based programme because of the supervision of trained sport or physio-therapists who are trained with all details of the training programme in advance. In addition, all training content is developed based on current knowledge of training science, e.g., having a progression over time, first balance exercises and strengthening exercises at the end. During the supervised training sessions, participants are always made aware of potential hazards and learn how to assist themselves, e.g., during balance exercises, or having problems within an exercise. In addition, different potential adjustments, and modifications to simplify and to complicate the exercises are given by the therapists. With this preparation over the first 26 training sessions (supervised by the therapists in the study centres: 24 sessions TSEP + 2 sessions HBEP = 26 sessions on-site), having the manual and the accompanying videos, participants should be optimally prepared for the home-based training programme in the second period. By implementing the health questionnaire and carefully selected inclusion and exclusion criteria, all participants are expected to fulfil the prerequisites for study participation. Measurements and training sessions are instructed and supervised by well-trained personnel, so that potential risks are minimised. Proper safety precautions additionally minimize the risk of uncomfortable and adverse events. In summary, the expected benefits exceed the potentially occurring risks.

The recruitment of potential participants will be via Generali Health Solutions (GHS), who will address insured older persons of “Generali Deutschland Krankenversicherung AG” (Generali Germany Health Insurance AG) and its subsidiary who live in the close area of the participating study centres. They receive detailed study information. The recruitment process could potentially lead to a selective recruitment of those older adults who are intrinsically motivated to exercise and who are already physically active. The fact that our study included older adults with a higher fall risk (according to the Silver Index, > 40%), while still being able to do 30 min of physical activity, could potentially bias our results. This aspect should be kept in mind while interpreting the results of the study. All interested persons had to state that they were willing to be randomised in one of the two groups to minimise dropouts, particularly in the control group. Within the randomisation procedure, a random allocation of all available places for the study is made by drawing lots. The examiners of the centres have no influence on the allocation to the groups, a change between groups is not possible and group preferences are strictly not allowed. Since the therapists do both measurements and the supervision of the training sessions, blinding of assessors is not possible which yields an acceptable risk for bias. The measurements at T₀, T₁, T₂ and T₃ are standardised and performed in the same room in each study centre by the same assessor. Measurements should be scheduled at the same day of the week and at the same time of the day for the four measurements if possible. The control group does not receive any training treatment, but they obtain the opportunity to use an optimised training programme after the study is finished (including 1-year follow-up). The participants in the control group are instructed to maintain their usual physical activity during the study. Due to the potential bias and influence of different levels of physical activity, the amount is documented through the German PAQ-50 + during the study duration for both groups. With having a multi-centre study, a bias of regional differences should be minimised. Furthermore, health status of older adults can change over the study duration and influence participation, drop-out rate and results of the study. To gain insights here, the EQ-5D-5L is implemented.

This study is expected to lead to new insights which may help establish a new approach to fall prevention training for older adults at risk for falls. Based on the results of the first part of the intervention, adjustments of load criteria, training content, and exercises can be realised for further using of this approach. Together with the SI measurement, the approach includes a systematic and individually tailored innovative fall prevention programme that can be implemented afterwards in a broad fall-prone target group of older adults. The results will be published in a peer-reviewed scientific journal to make them accessible to the scientific community. After the study completion, approaches to examine the cost-effectiveness and develop an implementation plan are relevant aspects for further steps.