Can a tailored exercise and home hazard reduction program reduce the rate of falls in community dwelling older people with cognitive impairment: protocol paper for the i-FOCIS randomised controlled trial

A randomised controlled trial will be conducted with 360 older people with cognitive impairment. Figure 1 gives an overview of the study design.

Trial reporting will be guided by the CONSORT Statement extensions for trials of non-pharmacological interventions and pragmatic intervention trials.

Participants will be recruited via two centres in Sydney, Australia. To be eligible for the study, participants must be a) community dwelling, b) aged 65 years or older, c) have CI as defined by having a Mini-Mental State Examination (MMSE) score or Mini-Addenbrookes Cognitive Examination Australian Version (Mini-ACE) <24, an Addenbrookes Cognitive Examination – III, Australian Version (ACE-III) <83 or a specialist clinical diagnosis of cognitive impairment, d) have an identifiable and consenting “person responsible” and e) have a carer who has a minimum of 3.5 hours of face to face contact with the participant each week and is willing to assist with reporting falls and supervising the exercise intervention (3 or more times per week).

People living in Residential Aged Care Facilities will be excluded as will those with an MMSE or Mini-ACE < 12/30 as it is considered unlikely such people will be able to engage with the intervention. Participants need to have sufficient English language skills to understand the assessment and intervention procedures. Other exclusion criteria include inability to walk more than one metre despite assistance with a walking aid and/or another person, blindness, severe psychiatric disease, a progressive neurological disease other than dementia or any medical condition precluding exercise (e.g. unstable cardiac disease).

Ethics approval has been obtained from the relevant human research ethics committees (South Eastern Sydney Local Health District Ethics committee (HREC 14/046), Northern Sydney Area Health Service and Neuroscience Research Australia).

All eligible participants will be visited at home by a research physiotherapist. The study will be explained in detail to the participant and their carer and written consent will be obtained prior to the baseline assessment. Capacity to consent will be determined by whether the potential participant demonstrates an ability to comprehend, retain and recall information about the study and any risks involved. Where there is doubt about a potential participant’s ability to process, retain and recall information about the study, assent to participate will be obtained from the identified “person responsible”.

All participants will be assessed in their home at three time points: on entry to the study (baseline assessment, prior to randomisation); at 6 months post randomisation and at 12 months post randomisation. Table 1 highlights the measures to be undertaken at each time point. Post randomisation assessments will be undertaken by assessors blinded to group allocation.

Participants will be randomised after completion of the baseline assessment. Randomisation will be stratified by hospital recruitment site using computer generated random numbers with variable block sizes of 6–8. The randomisation will be performed centrally using an independent web-based program by an investigator not involved in assessment or intervention.

Following baseline assessment, the control group will receive usual care from their medical practitioner and community services. The control group will not receive any additional intervention as part of the study.

Prior to commencing the intervention, each participant’s medical practitioner will be contacted to ensure support for the participant’s involvement in the study. All participants in the intervention group will undergo an assessment of functional cognition and a home safety assessment to inform the approach to the exercise and home safety program. The intervention group will receive a combined total of 11 visits over a 12 month period from the occupational therapist and physiotherapist, with the ratio of visits determined by the identified needs of the person at the time of assessment and their home environment. Factors influencing the ratio of visits will include physical performance, functional cognition, participant and carer willingness to engage, identified intervention priorities and issues of safety. Additional phone calls in between visits will be used to encourage continued engagement in the intervention and identify any problems with either the exercise or environmental recommendations.

The exercise intervention will consist of exercises commonly used in fall prevention research and clinical practice. The program will be delivered by experienced physiotherapists to minimise the risk of adverse events and taught to the carer who will supervise practice sessions as necessary. The visits will be more frequent at the beginning of the program to ensure safety and enable tailoring of the program to the cognitive and physical abilities of each participant. Each scheduled visit will last 40–60 minutes and participants will then be requested to undertake a 30 minute exercise program up to 6 times per week at home for 12 months. The exercises will be primarily conducted in the standing position and emphasise balance training and muscle strengthening.

The optimal intensity and type of exercises for each participant will be assessed and adjusted by the study physiotherapists to ensure that the intervention remains safe but challenging.

Uptake and adherence to the home safety and exercise recommendations will be recorded.

As this study involves the prescription of home-based exercise to a high risk population, safety while exercising will be a prime consideration. Participants will be shown how to perform exercises with stable supports if necessary and will be provided with a tailored exercise booklet. This material will be adapted to suit the participants’ cognitive abilities including large photographs of exercises, simple instructions and safety precautions. Carers will be considered intervention partners and will be taught how best to supervise exercise sessions including the type of cueing required.

A logbook for recording exercises completed and effects of exercise (e.g. muscle soreness) will be provided. Participants (or their carers) will be advised to telephone study staff if they experience any major adverse effects, e.g. muscle soreness lasting for more than 48 hours that interferes with daily activities or requires medical attention. If a participant becomes unwell or has an admission to hospital, the program will be resumed when the participant and the relevant professionals deem him/her well enough to participate again, and the exercise program may be modified by the therapist if the participant’s status has changed from pre-hospital.

The primary and secondary outcome measures are outlined in Table 1. The primary outcome measure will be the rate of falls over the 12 month follow up period. A fall will be defined using the internationally derived consensus definition of “an unexpected event in which the participant comes to rest on the ground, floor, or lower level” as a result of a loss of balance [28]. Fall rate will be calculated from monthly calendars. All participants will receive 12 calendars at the time of the baseline assessment. Participants/carers will be asked to record falls and use of health and community services on the calendars and return the information to the co-ordinating centre each month using prepaid addressed envelopes. Participants/carers who do not return calendars will be telephoned to ask for the information. Participants who report having fallen will be telephoned to seek additional information about the circumstances and consequences of the fall. Staff who receive calendars, make follow up phone calls and enter data will be unaware of participants’ group allocation.

The secondary outcome measures listed in Table 1 will be collected at the identified time points by an assessor unaware of group allocation. The order of measurements will be standardised. Participants/carers will be instructed not to inform the assessor of their intervention status and all home exercise equipment will be removed or concealed prior to the assessment.

The cost of implementing the exercise and home hazard program will be obtained from trial records and NeuRA financial records. The use of public and private healthcare resources will be recorded as part of the monthly calendars over the 12-month study period. Resources will be costed using published government sources where available (Pharmaceutical Benefits Scheme unit costs; Medicare Benefits Schedule, and the National Hospital Costs Data collection). We will also utilise our own previously collected data [32] to provide estimates of out of pocket costs to patient and family.

Power analysis (with 5% significance level, 80% power) has been undertaken using data from our previous work and similar studies. A total of 360 participants (180 per group) will provide 80% power to detect a significant 30% lower rate of falls for intervention participants than control participants (i.e. IRR 0.70). We assumed alpha (measure of over-dispersion in negative binomial regression model) to be 0.8 and the control group rate of falls would be 1.8 falls/person year over the 12-month follow-up period - based on our cohort study [1]. An average follow-up period of 11 months (rather than the planned 12 months) was used in the sample size calculation to account for loss to follow-up.

An intention-to-treat approach will be used for all analyses. The number of falls per person-year in the intervention and control groups will be compared with incidence rate ratios using negative binomial regression. This provides a more powerful analysis than a simple comparison of the proportion of fallers in the follow-up period, as it takes into account all falls during the trial, and also the distribution of falls, which is Poisson-like but has a longer tail. General linear models will be used to assess the effect of group allocation on the continuously scored secondary outcome measures. Modified Poisson regression models will be used to compare groups on dichotomous outcome measures. Predictors of adoption and adherence will be analysed using multivariate modelling techniques such as multiple linear and logistic regression.

The economic evaluation will be conducted from the perspective of the health and community service provider. Data will be collected regarding costs of the exercise and home hazard reduction program delivery (including staff costs, training, capital costs and consumables). Inpatient hospital admissions and duration, emergency department presentations and other health and community service contact (including GP and specialist visits) will be recorded via the monthly calendars. Pharmaceutical use will also be collected. Incremental cost-effectiveness ratios will be calculated using multiple health outcomes, i.e. the incremental cost per a) fall prevented, b) per fall requiring medical attention avoided, c) ED presentation avoided, d) hospital admission avoided, and e) QALY gained (based on the EQ5D-5 L). Using the mean costs in each trial arm, and the mean health outcomes in each arm, the incremental cost per health outcome (a-e, above) of the intervention group compared to the control group will be calculated and results will be plotted on a cost-effectiveness plane. Bootstrapping will be used to estimate a distribution around costs and health outcomes, and to calculate the confidence intervals around the incremental cost-effectiveness ratios taking account of joint uncertainty in costs and benefits. One way sensitivity analysis will be conducted around key variables; a cost-effectiveness acceptability curve will be plotted. A cost-effectiveness acceptability curve provides information about the probability that an intervention is cost-effective, given a decision maker’s willingness to pay for each additional health outcome gained.