A multi-centre, parallel-group, randomised controlled trial to assess the efficacy and safety of eurythmy therapy and tai chi in comparison with standard care in chronically ill elderly patients with increased risk of falling (ENTAiER): a trial protocol

The primary objective is to determine whether EYT and Tai Chi can reduce the risk of falling (i.e. experiencing at least one fall), which usually marks the beginning of dependency and the decline of mobility. This is assessed by comparing EYT plus standard care, Tai Chi plus standard care and standard care alone with respect to the risk of falling over a time period of 6 months. The secondary objective is to determine whether EYT and Tai Chi can improve mobility and balance and reduce injuries through falls and fear of falling, which are usually the primary threat to mobility. The follow-up question is to determine whether this leads to the better management of daily tasks and a better health-related quality of life, cognition and mood and whether it affects healthcare utilisation and costs. Finally, the inner correspondence of the participants with the practices and how well they adhere to them will be assessed.

Multi-centre, randomised controlled, open-label, confirmative trial with two active and one control parallel groups.

Eight qualified trial sites are used in this trial: Medical Center – University of Freiburg (Centre for Complementary Medicine), Tübingen, (Institute of General Practice and Interprofessional Care), Witten Herdecke (Institute of Integrative Medicine), Duisburg-Essen (Department of Naturopathy and Integrative Medicine), Ulm (Integrative Medicine), Charité, Berlin (Immanuel Hospital), Havelhöhe Hospital, Berlin and Filderklinik, Filderstadt. The University Centre for Complementary Medicine and the Clinical Trial Unit (CTU) of the Medical Centre–University of Freiburg organise and manage the trial. Patient recruitment is planned from the third quartile of 2019 to the second quartile of 2021. The follow-up will end in the second quartile of 2022.

Included will be adults who are at least 65 years, have a chronic disease (e.g. musculoskeletal, neurological or cardiovascular), have an increased risk of falling [Berg Balance Scale (BBS) score 49 or less and a self-reported history of imbalance], are able to participate in 1 h group-based sessions and are able to leave home on their own. There is no restriction with respect to their domicile.

Excluded will be patients with medical conditions limiting participation (e.g., terminal illness, heart failure NYHA III-IV, unstable angina pectoris, uncontrolled seizure disorder, decompensated lung disease, cancer in the advanced stage, chemo- or radiotherapy ongoing or during the last 3 months or amputation of one or both legs); complete dependence on a rollator (wheeled walker); major visual, hearing or language problems affecting the understanding of trial documents and procedures; major cognitive impairment (MoCA score 18 or less); conditions affecting the participation of group classes (e.g. high risk for individual support, urge incontinence, personality disorder and substance use disorders); life expectancy less than 1 year; permanent confinement to bed to be expected in less than a year; participation in regular Tai Chi or EYT exercises within the last 6 months and participation in vigorous sports for exercise within the last month.

A specific recruiting strategy will strengthen the enrolment of elderly people who are economically disadvantaged and living a reclusive life.

Participants will be randomly allocated to receive standard care plus EYT or plus Tai Chi or standard care only (control).¹

Primary outcome Risk of falling is the incidence of experiencing at least one fall within 6 months. A fall is defined as ‘an unexpected event in which the participants come to rest on the ground, floor or lower level’ [42]. The frequency of falls is self-reported daily in the fall diaries and ascertained during the monthly phone calls. Its assessments follow international guidelines [42].

Secondary outcomes are the risk of falling at month 12; the number of falls and injurious falls categorised into ‘moderate’ (sprains, bruises, scrapes, joint injuries and needing medical care) and ‘serious’ (fracture, head injury, tendon rupture, inpatient care and requiring stitches) [43]; balance and mobility (BBS [44]); cognition (MoCA [45]); fear of falling (short FES-I [46]); mood (GDS-15 [47]); health status (SF-12 [48]); instrumental activities of daily living (IADL [49]); use of medical and non-medical services (FIMA [50]); medications (total number and classes of fall-risk-increasing drugs [51]); adherence to practicing in groups and at home (participants’ lists and training diary); inner correspondence/peaceful harmony with practices (ICPH [52]), use of standard fall prevention measures and subjective estimation of motivational support; and contamination with Tai Chi or EYT in the non-Tai Chi and non-EYT groups (self-designed questionnaires, see Supplementary file 1, 2, 3 and 4)).

The complications that can arise from EYT, Tai Chi or other exercises will be collected monthly. A complication is any untoward medical occurrence such as any unfavourable and unintended sign, symptom, or disease that is possibly related to the study intervention. Not regarded as complications are conditions that were already present at the time of the informed consent or aching muscle, which is to be expected. Starting with the open questions, participants are asked about any untoward medical occurrences or unplanned medical visits. In case of a possible relation to the intervention, these occurrences are documented, including severity and outcome. Moreover, therapists or teachers can inform the trial sites about the complications. Falls and injurious falls are documented as primary outcomes, and injuries are reported as complications in case they occurred during training. Further information on safety will be derived from the FIMA, which assesses healthcare utilisation such as physician contacts and hospitalisations.

The time schedule of enrolment, interventions, assessments and visits for participants is shown in Table 1.

Sample size calculation is based on the assumptions on the proportion of participants with at least one fall during a six-month period. The proportion of participants with at least one fall is assumed as 50% in the standard care group [44, 53]. A decrease of this proportion to 35% is considered clinically relevant and achievable by EYT and Tai Chi [54]. The study is planned with 80% power to show the difference between treatments at the two-sided level alpha of 5% when the proportions of fallers (i.e. participants experiencing at least one fall within 6 months) is 50% with standard care and 35% with both EYT and Tai Chi. According to the chi-square distribution, 155 participants per group (calculated with NQuery Advisor 7.0) are required. The sample size calculation is performed for the three-group comparison. To account for multiplicity issues in this three-group comparison, a closed test procedure at the two-sided level alpha of 5% will be used. That is, the global null hypothesis of no differences between any of the three intervention types will be tested first, and if it can be rejected, pairwise group comparisons will be conducted each at the two-sided level alpha of 5%. The two-group comparison of EYT versus standard care has a power of 80% when the proportion of fallers is reduced from 50 to 34.3%. The two-group comparison of EYT versus Tai Chi has a power of 80% when the difference in the proportion of fallers is 15%. To account for a certain amount of non-compliance and incomplete observations, 550 participants will be randomised at a ratio of 1:1:1.

The trial will be announced through local media, a web page (https://​www.​uniklinik-freiburg.​de/​entaier-studie.​html), flyers, posters and personally to elderly people and to key people in contact with the elderly (senior citizens, physicians, hospitals, pharmacies, ambulances, nursing services, household help, food banks, social workers, social welfare office).

Potential participants contacting the local trial site will be pre-screened through a phone call. Elderly people that appear to be eligible will enter a waiting time until 15 or 30 participants could be enrolled. Only then will they receive detailed information. After providing their written consent, they are screened, enrolled and assessed for the baseline by the local trial sites. Within four subsequent weeks, they are randomly allocated to the three groups and receive their study book containing the dairies and all relevant information. This step initiates the observation period and is immediately followed by the start of the EYT and Tai Chi sessions.

Central randomisation by fax will be performed in this trial (CTU). Randomisation will be stratified by clinical site and performed in blocks of length 12, 15 or 18 at a ratio of 1:1:1. The randomisation code will be produced by validated programmes based on the Statistical Analysis System. For each batch size (12, 15 and 18), a separate randomisation fax form and randomisation lists will be prepared.

The blinding of the Tai Chi and EYT group and the standard care group to their allocation is not possible because it is obvious in which group the participants are randomised, as all three trial interventions differ substantially. The outcomes are either directly reported by the participants (falls, questionnaires) or in close communication with the participants (falls, complications, BBS, MoCA, IADL). Therefore, also a reliable blinding of the outcome assessment to participant allocation is not possible. Outcome assessors are instructed by a written guide to start with open and general questions and to avoid mentioning the group allocation as long as possible. Since the electronic case report forms differ depending on randomized treatment, the statistical analysis will not be blinded, but all confirmatory analyses are prespecified in the clinical trial protocol and decisions on analysis data sets will be finalized without access to the unblinded data. Concealment of the randomisation will be guaranteed to minimise selection bias. All assessments will be conducted in a comparable setting and time. The primary outcome parameter is clearly defined and described in detail [42]. The combination of prospective, daily, participant-completed falls diary and monthly calls is regarded to be reliable and captures most of the falls among seniors [55]. To reduce incomplete outcome data, travel costs for public transportation are refunded, and the trial sites may conduct the follow-up visits at the participants’ home. Two of the groups receive interventions in a comparable setting, support and frequency. The third (control) group, although not offered regular training sessions, can still avail of such opportunities through the addresses in the brochure, the general courses in their city or their primary care physician. They receive human attention through an appreciating communication with the trained study personnel and their primary care doctors, other consultants and medical referrals at their discretion and informed by evidence. The patient informed consent (form) makes no remarks on the potential efficacy of EYT or Tai Chi to reduce expectancy bias. Publication bias will be reduced by study registration in the trial registers, the publication of the trial protocol and the results and the support of open data.

BBS, MoCA, IADL and the complications are assessed by the study personnel.

The FES-I, GDS-15, SF-12, ICPH and FIMA questionnaires, the falls and training diaries and the self-designed questionnaires (see Supplementary file 1, 2, 3 and 4) are filled in by the participants in the same environment, which is quiet and respectful of the subject’s privacy. As much as possible, the outcome measures are assessed at a consistent time of the day and consistent with the medication cycles, particularly in participants with conditions like Parkinson’s and multiple sclerosis. The data obtained by the investigators are kept in source documents and directly entered in a trial-specific electronic case report form (eCRF). The patient-reported questionnaires are stored as the participants’ source documents, and a duplicate is sent to the CTU in Freiburg to be entered in the eCRF.

Participants may withdraw from the group classes or from the study for any reason at any time. The investigator may withdraw a participant from the intervention if the risks outweigh the benefit or if the inclusion/exclusion criteria are severely violated. Nevertheless, further follow-up visits and the assessment of the trial endpoints will be attempted to conduct an analysis according to the intention-to-treat principle.

Following the recommendation of public and patient representatives (PPR), EYT therapists, Tai Chi teachers and study personnel will receive a 3 h personal and written refresher training in appreciating, encouraging, non-judgmental communication and managing challenges of the group situation provided by a communication trainer. The EYT therapists and Tai Chi teachers will also receive a motivation concept on elderly people to comply with regular exercises, with a specific focus on the elderly living a reclusive life. The motivation concept is based on published research [21, 56, 57‐61] and expert consultation.

Extensive procedures ensure the quality control of data collection, data entry and subject confidentiality. The study data are collected and managed using REDCap™ Version 8.6.5, a fully web-based remote data entry system based on web forms developed and maintained by the REDCap Consortium (redcap@vanderbilt.​edu).

Data entry and data corrections on the e-forms are automatically tracked in an audit trail created by the electronic data capture system. Data will be regularly reviewed for completeness, consistency and plausibility. During the trial, the clinical research associate will visit the trial sites regularly. The clinical research associate will verify if the trial is conducted according to the trial protocol (signed informed consents, eligibility of participants, primary endpoint and safety data), the trial-specific procedures, the International Conference on Harmonisation-Good Clinical Practice guideline (ICH-GCP, R2) and the national/local regulatory requirements and will compare the entries in the eCRF with the source data.

Data will be exported from the database into the statistical software for analysis.

The primary efficacy analysis will be performed according to the intention-to-treat principle. All randomised participants will be analysed as belonging to their randomised arm, regardless of whether they refused therapy or whether other protocol deviations are known.

For the analysis of the primary endpoint fall (yes, no) within 6 months, a time-to-event methodology will be applied using the time to the first fall to account for the potentially incomplete observation times of the participants. This is considered to be more efficient than an analysis of the binary endpoint fall (yes, no), therefore leading to a gain in power. For participants who do not fall within the six-month observation period, the time to last contact will be used as censored observation. Death without a previous fall will be regarded as a competing event in the analysis. A Cox regression model will be used to test the efficacy of EYT versus Tai Chi versus standard care in reducing the risk of falling within 6 months after handing out the study book. The model will include the intervention type, trial site, BBS score and age as the independent variables. To account for multiplicity issues in this three-group comparison, a closed test procedure at the two-sided level alpha of 5% will be used. That is, the global null hypothesis of no differences between any of the three intervention types will be tested first, and if it can be rejected, pairwise group comparisons will be conducted each at the two-sided level alpha of 5%. Hazard ratios between the interventions will be calculated from the model with a two-sided 95% CI. In addition to the overall comparison of the treatment arms, a comparison of treatments in the key participant subgroups will be of interest. The treatment effect will be estimated separately in the participants defined by different diagnoses (present vs. absent) from the regression models, including the interaction between treatment and diagnosis.

For the analysis of the secondary endpoint total number of falls per participant within 6 months after handing out the study book, a negative binomial regression model will be used to test the efficacy of EYT versus Tai Chi versus standard care. A negative binomial is preferred over a Poisson regression because of the potential overdispersion. Fall rates for participants with follow-up times of less than 6 months will be adjusted by accounting for their exposure time. The model will include intervention type, trial site, BBS score and age as the independent variables. The incidence rate ratios between interventions will be calculated from the model with a two-sided 95% CI. To consider the potentially present excess of zeroes, a zero-inflated negative binomial model will be used as a sensitivity analysis. Other secondary endpoints will be analysed with the appropriate regression models.

Safety analyses will be performed for all participants for whom one of the interventions is started, and participants will be analysed according to the intervention received. The rates of adverse events and serious adverse events will be calculated with a two-sided 95% CI. No formal interim analyses of the efficacy endpoints will be performed.

The utilisation of healthcare cost will be calculated by applying a standardised unit cost allowing the evaluation of cost development for each group over time. The costs of healthcare utilisation (months 6 and 12) will be analysed using appropriate regression models. Intervention type, trial site and utilisation costs at baseline will be included as the independent variables.

In the second step, cost-effectiveness analyses of the intervention will be conducted from the healthcare perspective. Therefore, the development of the different endpoints will be analysed in relation to the costs of healthcare utilisation and the costs of the intervention. Incremental cost effectiveness ratios and the corresponding confidence intervals will be estimated.

A DMC comprises four specialists with high expertise in conducting clinical trials in Tai Chi and EYT and in safety assessment and falls in the elderly. These specialists are independent from the study organisers and all trial sites. They will regularly review the complications of the trial treatments and advise the principal investigator (the details are described in the DMC charter).

The PPR are involved during the planning, conduction and analysis of the trial following the recommendations by INVOLVE (http://​www.​invo.​org.​uk). Through this, the views, values, needs, strengths and limitations of the older people are integrated. In the personal semi-standard interviews, the details of the trial design are explained and discussed. Their comments and recommendations centre on motivation; how to recruit older people, elderly leading a reclusive life and people from a lower socioeconomic background; and the importance of an appreciating, encouraging and calm atmosphere, the giving of sufficient time and avoiding the sense of being in need of help. Audio-visual material to support practicing at home is also recommended. Therefore, communication training, the provision of audio-visual material and the details of the recruitment strategy are added, and more time is planned for the visits. The PPR are invited to all relevant trial meetings and receive relevant trial documents, analyses and trial reports. Particularly, the documents handed out to participants (i.e. brochure, audio-visual material, information sheets, flyer and web page information) and their adequacy, comprehensibility and readability are discussed in detail. The PPR will be informed about the progress of the trials and unexpected problems and will be consulted regarding participant recruitment.

Before enrolment, the potential participants will be informed about the goals, procedures and possible risks of the trial; data protection and confidentiality; and the voluntary nature of the participation throughout the study. A written informed consent must be obtained for all potential participants before they are registered. Before the study begins, approval from the local ethic committees will be obtained. (From the Ethics Committee of the University of Freiburg, the ethical approval has been obtained: 183/19; 23.07.2019, as well as from the Ethics Committees of Tübingen for the sites Tübingen and Filderstadt, 561/2019BO1, 21.10.2019, Witten/Herdecke 150/2019, 31.07.2019 and Ulm 284/19, 01.08.2019; Berlin relies on the vote from Freiburg). The study will be carried out in compliance with the Declaration of Helsinki and the ICH-GCP.

Data protection follows the data protection concept of the CTU and must be fully and permanently complied with. The names of participants are only available in the source documents, which are kept in a secure and lock-protected location at the trial sites and in the participation lists of the EYT therapists and Tai Chi teachers. The findings obtained in the course of the study will be stored in electronic media and treated in strict confidence. For the protection of these data, organisational measures are taken to prevent disclosure to unauthorised third parties. For example, the subject data will be captured in pseudonymised form (i.e. subject ID number for the particular study and year of birth) throughout the documentation and evaluation phases.

The participants enrolled in the trial are included in an insurance that covers the expenses in case a participant suffers harm from trial participation.

The principle investigator has full access to the final trial dataset. The trial results, treatment manuals and videos will be published and presented to healthcare professionals, the public and other relevant interest groups. The full trial protocol is accessible by the public (see Supplementary file 5). The principle of open data is supported (https://​www.​bihealth.​org/​de/​quest-center/​mission-ansaetze/​open-science). Anonymised data may be shared with cooperating scientists or scientists who have other medically or scientifically well-founded reasons (goals referring to transparency, replicability, different analyses, combinations and meta-analyses of data). Their research must aim at improving the care of elderly people, and the researchers have to ensure data protection.​.