The Winchester Falls Project

BACKGROUND INFORMATION

At least one in three people aged sixty five and over are reported to fall in the community each year and in the very elderly this figure is nearer 40 % (1,2). The morbidity and mortality resulting from falls in the elderly is substantial and the costs associated with such falls are likely to rise with the predicted growth in the elderly population. The Health of the Nation document recognises that falls are an important cause of disability and use of the health service in older people (3). More than 50% of accidental deaths in those aged sixty five years and over are due to falls, and survivors of falls have a tow-fold increased chance of being hospitalised (4). The target of the Health of the Nation is to reduce the death rate for accidents among people aged sixty five and over by at least 33% by the year 2005 (3). Prevention of falls and fall-related injuries will be a major factor in achieving such targets. It is therefore wise that approaches which are appropriate and likely to be effective are adequately researched and implemented if indicated. There have been a number of approaches to studying the prevention of falls, and their sequelae, in the elderly and these are explored in the accompanying literature review.

AIM

To determine whether multidisciplinary assessment of community-based patients, who have had at least one previous fall, reduces the rate of further falls and fall-related problems. The study also aims to determine whether such an intervention, if effective, is best carried out in a primary or secondary care setting.

STUDY DESIGN

A secondary prevention randomised controlled trial of two different settings for the multidisciplinary intervention versus "usual care".

Randomisation will be done prior to allocation on a cluster basis. This means that the principal liaising General Practitioner in each practice will sign consent to volunteering their "cluster" involvement in the study (this will also set out their duties to their patients). The reason for this type of randomisation is to avoid "contamination" of usual care patients which would be likely if a General Practitioner were receiving regular information about patients who were in an intervention arm of the trial. Assent to study participation would also be sought from all individuals, without their first knowing the precise nature of the intervention (5) so as to allow follow up of the control group. Subjects recruited to the intervention groups would receive an information sheet outlining the study and what the intervention would involve and assent would be sought.

STUDY SIZE

It is thought that the trial will involve approximately 516 patients in total. A recent local pilot study of a smaller falls related project identified 70 patients in 9 months in a practice with a list of 17,000 patients. It is expected that the total population base from which patients will be identified will be 60,000. It is likely that the number required to participate will be identified over a one year period. Power calculations indicate that to provide a good chance of achieving a statistically significant result approximately 150 individuals will be needed in each of the three arms of the study.

INTERVENTION

There are two intervention groups.

The first intervention group will undergo a multidisciplinary assessment by a physician, occupational therapist and physiotherapist (see enclosed assessment form) based in a falls clinic at the Royal Hampshire County Hospital (RHCH). All patients will have full blood count, urea and electrolytes, liver function tests, thyroid function tests, Vitamin D level, bone profile, electrocardiogram and urinalysis. Further investigation, treatment modification and onward referral will be made as appropriate and relevant to each individual. The patient's General Practitioner will receive written communication regarding the visit, outcome and further recommendations.

The second intervention group will under go a Community-based process. All identified fallers will undergo an assessment by a health worker in each participating practice. This individual, identified within the practice, and trained in the assessment of fallers by staff at the RHCH Falls Clinic, will administer a questionnaire, and make referrals to physiotherapy, Occupational Therapy and the patients own General Practitioner as deemed necessary. Individuals who score particularly highly or are considered appropriate by the Primary Care team may still be referred onward to the Hospital Falls Clinic.

SUBJECTS

Patients aged over 65 years of age who are identified in the community as having had at least one fall in the preceding month and who live in one of the participating areas. Subject identification will be opportunistic and also through the ambulance service notifying the key worker in a practice when a patient has fallen at home but not been transferred elsewhere.

A fall is defined as "inadvertently coming to rest on the ground or other lower level with or without loss of consciousness and other than as a consequence of sudden onset of paralysis, epileptic seizure, excess alcohol intake or overwhelming external force (as per Close et al (6)).

EXCLUSION CRITERIA

• Abbreviated mental test score of less than 7 out of 10
• Not living in the participating areas
• Planning to move from the area within the next year
• Not expected to survive for the follow up period of one year
• Non-English speakers who do not have a relative/carer who could interpret

METHODOLOGY

Patients will be identified by a key worker in each participating General Practice Surgery. A baseline assessment will then be done either in the patient's own home or the local surgery by the key worker (district nurse/practice nurse/health visitor). The assessment will be mostly questionnaire based and will seek information on basic demographic data, living circumstances, Barthel score, abbreviated mental test score, circumstances of most recent (index) fall and injuries sustained, information regarding previous falls in the last year, risk factors for osteoporosis, 5 point Geriatric Depression Score, and Elderly Falls Test Score score. A timed "get up and go" will also be done. All assessors will be trained in the administration of the assessment to reduce inter-assessor variability.

The patients in the intervention groups will attend for assessment in the Falls Clinic or in the primary care setting as described above.

Those in the conventional management group will receive usual care from their primary care team.

All subjects will be sent a paid reply postcard monthly on which to record any falls and fall injuries. If this is not returned the subject will be telephoned (when possible). If a fall has been recorded the patient will be contacted by telephone for further details.

After one year the baseline assessment will be repeated. Data will be analysed on an intention to treat basis.

PRIMARY ENDPOINT

• Proportion of participants in each group to have at least one further fall during follow up year

SECONDARY ENDPOINTS

• Death
• Move to institutional care
• Change in Barthel score
• Change in "Get up and go" score
• Fall related admissions during follow-up period
• Fall related fractures

DISSEMINATION

It is intended that this trial will be subject to peer review and publication will be sought. Dissemination of information to local practices and interested parties via presentations.

REFERENCES:

1. Tinetti ME et al. Risk Factors for Falls Among Elderly Persons Living in the Community. NEJM, Vol 319:No26;1701-1707.
2. Prudham D et al. Factors Associated with Falls in the Elderly: A Community Study. Age \& Ageing 1981:10, 141-6.
3. Health of the Nation Document. DOH
4. Askham J et al. Home and Leisure Accident Research: A Review of Research on Falls Among Elderly People. DTI Consumer Safety Unit/Age Concern Institute of Gerontology, London, 1990.
5. Edwards SJL et al. Ethical issues in the design and conduct of cluster randomised controlled trials. BMJ 1999;318:1407 -1409 (22 May).
6. Close J et al. Prevention of falls in the elderly tial (PROFET): a randomised controlled trial. Lancet 1999;353;93 - 9