Fall and Injury Prevention in Older People Living in Residential Care Facilities: A Cluster Randomized Trial

Context In frail older people living in residential care facilities, hip protectors reduce fall-related femoral and pelvic injuries. In older people living in the community, prevention programs that target exercise and fall-related risk factors reduce falls and injuries. In older people living in residential care facilities, fall prevention programs, other than the use of hip protectors, have had mixed results. Contribution This randomized trial shows that a multidisciplinary fall prevention program reduces falls and femoral fractures in older people living in residential care facilities. The program included general as well as resident-specific, tailored strategies. Editors' Note The Cochrane Library (Issue 3, 2001) has two systematic reviews that summarize randomized trial evidence about interventions to prevent falls. The Editors Falls are a major problem in older people (1, 2). Hip fractures resulting from falls are particularly common in elderly persons living in residential care facilities (3, 4); they account for a substantial proportion of dependency and mortality (5). During the past decade, randomized, controlled trials studying fall prevention have shown both positive and negative results (6-8). Differences in target groups, interventions, and outcome measures may explain the inconsistent results. In residential care facilities and nursing homes, only one trial demonstrated that intervention programs may help prevent falls; however, this trial studied only people who fell repeatedly (9). Other trials have not shown reductions in falls or injuries (10-13) but have shown fewer hospital admissions (11) and improvements in mobility (12), visual acuity, and hypotension (13). The use of hip protectors has consistently proved effective in preventing hip fractures in selected high-risk populations (14). Some trials that have included cognitively healthier older people in the community have shown reduced falls and injuries when specific risk factors are targeted (15-18). However, evidence is lacking for older people living in institutions (19). We hypothesized that an intervention program that targeted multiple risk factors for falls in older people living in residential care facilities, in particular those with a high risk for falling, would reduce falls and fall-related injuries. We therefore conducted a cluster randomized trial for preventing falls in nine residential care facilities. Methods Design Study participants were older people living in residential care facilities located in Ume, a city in northern Sweden. The selected facilities had to have more than 25 residents. The nine that met this criterion were divided into group A and group B. The distribution was based on the age and number of residents, type of setting (care and service offered as well as corridor or private home design), and record of previous falls as routinely reported to the local authority. To keep the groups distinct from one another, the physicians, registered nurses, physical therapists, and occupational therapists who were responsible for working with the residents in group A could not also work with group B residents. Group A consisted of four facilities that accommodated 224 residents; the facilities had 29 to 74 residents each, and the median age by facility ranged from 82 to 85 years. Group B consisted of five facilities with a total of 215 residents; there were 31 to 66 residents per facility, and the median age by facility ranged from 79 to 85 years. The number of falls reported to the local authorities in the 2.5 years preceding the trial was similar for both groups: 1.26 per resident per year for group A and 1.29 for group B. After baseline assessment of all residents, groups A and B were randomly assigned by lots to an intervention or a control group (Figure 1). The random allocation was conducted by a person with no knowledge of the study. Two sealed, dark envelopes were used. In each envelope, a letter specified one of the groups. Before the lot was drawn, the first envelope drawn was designated as the intervention group. The local authorities, residents, staff of the nine facilities, and the research group were then informed of the results of the randomization. Figure 1. Study design. All residents in the study received written and oral information. All participants (or the relatives and guardians of participants with severe cognitive dysfunction) gave informed consent. The administrators and staff of the nine facilities involved also received information about the study and agreed to participate. The Ethics Committee of the Medical Faculty of Ume University approved the study. Definition of a Fall and an Injury A fall was defined as an event in which the resident unintentionally came to rest on the ground or floor, regardless of whether an injury was sustained. Thus, this definition also includes falls that resulted from acute illness or epileptic seizure and incidents that resulted in a resident's falling and being found on the floor by staff or another resident. An injury was defined according to the Abbreviated Injury Scale (20). Classifications were minor for injuries limited to superficial wounds and bruises; moderate for intermediate-level injuries, such as vertebral and wrist fractures; and serious for major fractures, such as hip fractures and other femoral fractures. Participants and Settings All residents in the nine facilities who were 65 years of age or older were selected in a cross-sectional manner. Thirty-seven of these residents declined to participate, were admitted to hospitals, or died before randomization. Sex and age of the 37 nonparticipants were similar to those of the remaining 402 residents. In Sweden, older people living in residential care facilities are disabled by cognitive or physical impairment and thus require supervision, functional support, or nursing care. In this study, some residents lived in private apartments and others had private rooms but shared dining and living rooms. In all facilities, residents had 24-hour daily access to assistance with activities of daily living, household issues, and medical care. The median age was 83 years (range, 65 to 100 years), and most residents were female (72%). Few residents could walk outdoors without a walking aid (14%) or shower without assistance (18%); few were nonambulatory (19%) or entirely dependent when eating (8%). Additional baseline characteristics are presented in the Table. Table. Baseline Characteristics of the 402 Residents Participating in the Study All members of the permanent staff, regardless of profession, participated in the study. In addition, eight physiotherapists were employed part-time (a total of 200 h/wk) until the end of the intervention period, and three physiotherapists were employed part-time (a total of 10 h/wk) during the follow-up period. A total of 273 nurses' aides or licensed practical nurses and 20 registered nurses worked at the nine facilities. Baseline Assessment Each resident's physician completed a questionnaire regarding clinical characteristics and drugs prescribed. A registered nurse reported delirium episodes (Table). Physiotherapists interviewed and assessed all residents. Hearing was rated as impaired if the resident could not hear normal speech from a distance of 1 meter or used a hearing aid. Vision was rated as impaired when the resident, with or without glasses, could not read a word written in 5-mm capital letters at reading distance. Global cognitive function was screened by using the Mini-Mental State Examination (MMSE) (21). Licensed practical nurses or nurses' aides were interviewed to determine the number of falls that had occurred during the 6 months preceding the study and the extent of use of physical restraints. Activities of daily living were assessed according to the Barthel index (22, 23). All residents were screened for the risk for falling. First, by using the Mobility Interaction Fall Chart (24), a resident was classified to be at higher risk for falling if he or she stopped walking when talking to an accompanying person (25), walked more slowly when carrying a glass of water (26), or had impaired vision or difficulty concentrating. Second, a physiotherapist globally rated the fall risk as higher if the resident showed risk-taking behavior considered to jeopardize balance. If the residents were not classified to be at higher risk for falling by any of these described measures, they were considered to be at lower risk for falling. Residents at higher risk were likely to be older than those at lower risk (median age, 84 years [range, 65 to 98 years] vs. 83 years [range, 65 to 100 years]), to have lower MMSE scores (median score, 17 [range, 0 to 30] vs. 21 [range, 0 to 30]), and to have more medical diagnoses (median, 4 vs. 3). The main areas of each facility were also screened according to a checklist for environmental hazards, such as lighting, flooring, obstacles inside the facility, and dangerous areas outside the facility (for example, icy areas). Twelve residents in the control group and six in the intervention group died or moved during the 11-week intervention period (Figure 1). Intervention Program The intervention program comprised strategies that targeted both general and resident-specific risk factors for falling. The strategies were designed to be meaningful to the residents without compromising mobility. The 89 residents screened as being at higher risk as well as the 19 residents at lower risk who fell during the 11-week intervention period were the focus of the individualized intervention program. Increasing the staff's knowledge about fall prevention was believed to be the starting point of a process that would produce long-term results. Staff Education All staff were invited to a 4-hour educational session, and more than half attended. The sessions were led by a physician and a physiotherapist and covered risk factors for falls and intervention strategie