People over the age of 65 are involved in three-fourths of all the fatal falls in the United States and have a greater than average number of falls that produce injury'.Even so, little agespecific research has been conducted to isolate causes of accidents among the elderly, especially falls.- Smith did find that 34% of !falls among the aged are accidental and many are due to environmental hazards, which implies that better control could reduce these falls and resultant injury.
A survey study was implemented to solicit additional information regarding factors that are related to falls by elderly people residing in long-term care facilities. The purpose of the study was identification of factors/ variables related to falls by the residents. The two long-term facilities also offered a unique opportunity to make comparisons based on the skill level of care provided (skilled, intermediate, self-care), as well as architectural design (high rise versus ground floor, and shape of the nursing care unit).
Unintentional injuries among people aged 65 and older are less likely to occur than injuries among people in younger age groups. However, older people are more likely to have fatal outcomes when injuries are sustained.4 After a five-year study of accidents resulting in either injury or hospitalization among older people living in care facilities, Margulec, Librach, and Schädel found that 95% of the injuries were due to falls.5 These authors reported that accidents resulted in severe injury, fracture, and/ or hospitalization at a rate of 3.4 per 100 residents and that overall accidents with or without injury occur at a rate of 5.1 per 100 residents.
Factors related to falls can be categorized into host and environmental factors. The host (internal) factors include age, sex, and functional level. Environmental (external) factors include the presence of confusion-potentiating drugs, design of facility, and presence of assistive devices.
Sex - In relation to the sex of the individual, Margulec et al. found that men had a fall rate of 2.1 per 100 residents while women had a rate of 4.0 falls per 100 residents.5 Sheldon also found that women were involved in about twice as many falls as men.6 Miller and Elliott, however, found that men were involved in about 25% of the total number of falls, although Feist found that they suffered a little over one-third of the injuries due to falls and significantly more deaths.8 Feist's conclusion suggests that men have more than their proportionate share of injuries, if not more falls.
Age - Studies show that the number of people falling and sustaining an injury from the fall increases with age. The degree of injury also can be correlated to increasing age, with the rate of resulting fractures increasing as the resident's age increases.5'9'10
Gryfe, Amies, and Ashley found that the rate of severe falls in a care facility was 1 17 per 1,000 persons per year. Another study showed an even higher percentage of falls resulting in fractures (80% or 415 fractures in 519 falls).5 Margulec et al also found that 60.9% of individuals with fractures died shortly thereafter. Using the death rate of that particular facility, 33.7% or 175 of the 519 clients would have been expected to die in the same time period. Agate found that people over 80 years of age have a mortality from falls eight times greater than those who are 60 years of age.12 These findings suggest that negative results of a fall increase as one grows older, especially after the seventh decade.
Level of physical activity- Both Margulec et al and Feist8 reported similar conclusions. First, both authors found that more falls occurred among people experiencing more limited activity of daily living than those who were more mobile. Secondly, the death rate following falls is significantly higher among more debilitated individuals. Kalchthaler, Bascon, and Quintos found those using assistive devices, implying greater debility, had the least number of falls, suggesting that there are ways to reduce falls even among the more debilitated.13
It is interesting to note that the blind aged were found to have a significantly lower accident rate than their sighted cohorts.5 The length of time they had been blind was not identified and could have influenced the finding.
Clustering of falls - Miller and Elliott noted that many falls occurred in a series (cluster) over a short period of time.7 It also has been found that recurring accidents are experienced by a small percentage of those who fall.5'10 Miller and Elliott found that 58% of the people who fell were involved in multiple falls.7 In their study, the rate of falls was identified as 6.5 per 1,000 patient days. They also found that 75 persons were involved in 203 accidents, with 88 or 43.3% resulting in bodily injury and an injury rate of 2.85 per 1 ,000 patient days.
Drugs - Many studies support the position that the use of drugs contributes to falls, especially with older persons. Margulec et al, however, did not find tranquilizers to be related to accidents.5 Feist, in contrast, did find that only 21.6% who fell were not taking tranquilizers.8 Of those who were taking tranquilizers, 57% were on regular doses and 43% were on doses as needed. Since many drugs have side effects of confusion, some nursing home staff believe that falls occur only among the confused. However, Feist reported that 26% of the falls occurred among welloriented residents.8 Kalchthaler et al found that many of the well-oriented who fell had poor judgment skill, which could account for the finding in both studies.13
Physical environment- The physical environment has been identified as a factor related to falls. Margulec et al found that most accidents occurred in the living quarters of institutions, rather than on the residence grounds.5 Miller and Elliott further identified 55% of the falls as occurring in the residents' bedrooms and 19% in the recreation and dining areas.7 Interestingly, bathrooms were involved minimally (8.3%). Equipment was involved in nearly 50% of the falls, with chairs and wheelchairs accounting for 30%. Kalchthaler et al found that residents using assistive devices such as canes, walkers, or crutches had the least number of falls.13 They suggested that these devices stabilized gait disorders by increasing proprioception and coordination, and strengthened a feeble gait by redistributing weight.
Time of day - Time of day has been identified as a factor in most studies; however, there is little agreement as to when a fall is most likely to occur. Margulec et al found that most accidents occurred between 7 am and 9 PM. as the residents were arising and preparing for the day's activities.5 Feist, however, found 6 am to 9 am to have the lowest rate of falls.8 The period with the highest rate of falls (24%) was 6 PM to 9 PM. She suggested that this was the time the residents were most fatigued and involved in bedtime activity.
Boucher found the poor light periods of early morning and early evening to be the peak periods for falls, another way to interpret the findings of the last two studies.9 Feist also noted the second-highest period for falls to be 9 AM to 12 noon (18.9%), while the period between 9 PM to 6 AM was relatively free from falls.8 In contrast, Miller and Elliott found that 27% of the falls occurred during night hours. They found the one-hour period of highest incidence to be 1 1 AM to 1 2 noon, with 3 PM to 4 PM to be the period of next highest incidence. Miller and Elliott suggest that the incidence of falls increases during decreased staff supervisory periods (staff changes and breaks).
Seasonal variation - Most studies have not reported a statistical significance for time of year. Gould did note that over half of the incidents occurred during the first quarter of the year (January to March).14 Miller and Elliott found June (12.3%) and October (1 1.8%) to have the highest percentages of falls.7 Boucher suggests from his study of falls in Scotland and Southern England that there is a climate factor.9 He proposed that the death rate in Scotland, which is twice as great, was due to the icy, snowy, and cold conditions that promote a greater number of falls.
Many of the findings concerning environmental and host factors are inconsistent or unsupported. Consequently, a survey-type study was conducted in which the incident reports of two separate care facilities for the elderly were analyzed in an attempt to add to the information available regarding falls by elderly persons.
At two care facilities, data were collected from incident reports completed by the charge nurse at the time a fall occurred. Incident reports were available for a two-year period at Agency A and a three-month period at Agency B. It should be noted that Agency B had been open less than a year, the reason for the short study period.
The completeness and accuracy of the data were dependent upon the charge nurse's ability and time to complete the form. However, because the forms are considered legal documents, the data were assumed to be accurate, if not always complete. The records were reviewed by the investigator and data were collected on predesigned forms.
One of the problems in comparing findings across reported studies is the lack of consistency in analysis. Use of bed capacity on 100-patient units does not allow for pooling of findings because of occupancy rate variation and difference in duration of studies reported. Consequently, the data in this study were analyzed by converting the falls, injury, etc., to rate of injury per 1000 resident days as done by Miller and Elliott.7 This method offers opportunity for other researchers and practitioners to assess and compare findings more accurately.
FINDINGS AND DISCUSSION
The population studied included residents aged 60 or older of two care facilities for the elderly. Agency A was a single-level structure with one nursing unit of a circular design and the other of a more traditional rectangular design. Agency B was a multilevel, high-rise structure with one skilled care unit on the groundfloor level.
A total of 101,280 resident days were included in the period surveyed. The period studied included approximately 84,000 resident days at the 115-bed, intermediate care Agency A. At Agency B, a total of 17,280 resident days were studied. This included 13,230 self-care and 4,050 skilled care resident days. The bed capacity of 190 at Agency B included 44 skilled care and 145 self-care.
During the survey period, Agency A had 253 reported falls involving 113 residents and Agency B had 36 reported falls by 29 residents (see Table 1 ). These figures are in agreement with the findings that there are some people who are "fallers." It has been suggested that people who are identified as accident-prone when they are young remain accidentprone as they age.
Sex - The findings of this study indicate that men had more than their proportionate share of falls at both agencies. At Agency A, men made up 10% of the total population but they suffered 28.5% of the falls there. At Agency B, 13% of the falls were incurred by male residents, who constituted only 5% of the agency population. In both agencies, the men had significantly (p< 0.05) more falls than their female counterparts. This finding agrees with the more recent studies.7'
Age - The residents who fell at Agency A ranged in age from 61 to 91, with the mean age being 79.4 years. The mean age of those who fell at Agency B was 87.3; ages ranged from 72 to 93. The overall mean age of all residents at Agency B was 83. 1 , with ages ranging from 62 to 98 years. The mean age difference at Agency B was significant at p<.001. This supports previous findings that falls are more likely to occur as one increases in age.
FALLS AND THE ELDERLY- SUMMARY
Drugs - Data regarding use of drugs were collected (see Table 2). Those falling at Agency A took from one to 16 different drugs per day, with a mean of 6.36 different drugs per resident who fell. There were two residents who were taking no drugs. At Agency B, those who fell were taking from one to ten different drugs with a mean of 5.11. Three persons who fell were taking no drugs.
A comparison of the total number of drugs taken by those who fell with those who did not fall did not show any significant difference. However, a significant difference was identified between the number of drugs taken at the two agencies (x2 = 14.87, ? < 0.01). A significant difference also was found when a comparison of the number of drugs was made between the two groups of subjects who fell at both Agency A and Agency B (x2 =10.69, ? < 0.0 1)
Month/season - There was a significant difference in the number of falls occurring by month and season, with the most falls occurring in the fall and winter months. The months of November, December, January, and March had the largest number of falls, while May and June had the fewest. These findings are supportive of Boucher's9 and Gould's14 findings.
Day of week- At Agency A, 18.6% of the falls occurred on Thursday, while Monday and Friday had the fewest. Monday was the least likely day for a fall at Agency B; Friday and Saturday were the most likely days for a fall to occur.
Time of day - The time of day at which the falls occurred at Agency A varied, with most falls occurring during the hour of 2 PM to 3 PM (9%). The next most frequent times were 6 AM to 7 AM and 6 PM to 7 PM, followed by 7 AM to 8 am, 1 2 noon to 1 PM, and 7 PM to 8 PM. The fewest falls occurred during the night hours of 8 PM to 5 AM. The records indicated that the falls occurring during the night hours usually were related to the victims' slipping in urine (usually their own) while on their way to or from the bathroom.
At Agency B, the multilevel care facility, the greatest number of falls (16.7%) occurred from 6 AM to 7 am, with 5 PM to 6 PM being the next most frequent period for falls (8.4%). The remainder of falls occurred on a relatively equal basis throughout the day. Most of these hours at both agencies correspond closely with meal and bed times and are in relative agreement with other studies' findings.5'8'9 When time of day and month of fall were compared, a p<.01 of correlation was found. This finding suggests that time of year, which affects daylight hours, should be considered in future study. Boucher also suggested that falls were more likely to occur at sunrise and /or dusk, when light level is changing.9
Site of fall- At both agencies, most of the falls occurred in the residents' rooms, the next most frequent site being the hallway outside the residents' rooms (p<.00l). The remaining 14.6% of the falls occurred in the recreation area or outside on the agency grounds (see Table 3).
At Agency B, the greatest number of falls again occurred in the skilled care residents' rooms (47.2%), while 13.9% were in the hallway. Self-care area residents incurred 27.8% of the falls. No one resident floor (upper five of a six-floor building) could be identified as having more falls than another floor. Sixty-one percent of the falls occurred in the skilled care unit and involved 23% of the residents living in that unit. These residents were much more debilitated and, on the average, several years older than the self-care residents. This finding supports previous findings that increase in age and/ or debility increases the chance for falls.5'8
The level of care required by a resident also shows differences. At Agency B, a significant difference in the rate of falls was noted between the two levels of care. The fall rate in the skilled care unit was 7.32 falls per 1,000 resident days, but only 0.53 falls per 1,000 resident days in the self-care areas. Overall, Agency B's fall rate was 2.10 per 1,000 resident days, comparable to Agency A's rate during the second year of data collection.
The floor plan at Agency A allowed comparison of the number of falls occurring in a circular versus a rectangular unit. It can be assumed the patients living on each unit were from the same population, as no distinction was made by the agency in assigning a resident to a particular unit. The circular unit, with the nurse's station equidistant to all rooms, had 36.5% of the falls. The rectangular unit, with some rooms much farther from the nurse's station than others, had 63.5% of the falls that occurred in patient rooms or adjacent hallways. The difference in number of falls occurring between the two types of floor plans was significant at p < 0.01 level, with 34.8% of the variance being accounted for. This finding suggests that a circular unit design that offers closer observation of all residents is preferred to a rectangular unit design that does not afford such easy observation.
LOCATION OF THE FALLS
RATE OF FALLS
Rate of falls - The rate of falls offers interesting information (see Table 4). At Agency A, the rate of falls was 4.05 per 1,000 resident days for the first year of data collection. The rate dropped to 2.59 the second year. This difference in rate of falls was statistically significant at p< 0.025. Changes made at Agency A suggest the reasons for the drop in the rate of falls. There was an increase in number of staff who were able to provide more attention to each resident. A full-time inservice person was added to the staff and was holding programs designed to increase the knowledge of the staff regarding care of the elderly. A third change was the addition of a closed hallway between the two wings and dining area. Residents no longer needed to go through doors and leave a building for meals or to visit in the other wing. It was not possible to identify one factor as being more influential than another.
Degree of injury - The degree of injury resulting from the falls also seems predictive (see Table 5). Of the individuals who fell at Agency A, no treatment was required by 59.8% (148) but 28.5% (68) needed treatment at the agency. The treatment at the facility usually included bandage of a cut or abrasion. Hospitalization was required for injuries incurred in 1 1.7% (28) of the falls. Of residents requiring hospitalization, 16 (6%) were transferred to a skilled care agency, indicating an increase in their degree of debility as a result of the fall.
LEVEL OF TREATMENT REQUIRED FOR RESIDENT INJURY
In Agency B, 52.0% of the falls resulted in no identified injury. Moderate injury needing treatment at the agency occurred in 34.1% of the residents. Hospital treatment was required by 13.9% of those who fell, while three (8%) of the self-care residents who fell were moved to the skilled care unit as a result of the falls.
These findings of increased debility after a fall and greater injury with increasing age are supportive of previous findings.
This study involved two care facilities for the elderly with a combined bed capacity of 300. Data were collected over a total of 27 months (101,280 resident care days) and analysis was made of 289 falls by 142 residents. The mean age of the residents at Agency A (intermediate care) who fell was 79.4 years, and 87.3 years at Agency B (skilled and sheltered) for those who fell. Specific findings of this study were:
1. A significant difference in rate of falls per 1,000 resident days was identified based on level of care of the unit, and age of the resident. The greater the age or the greater the debility of the resident, the greater the potential for a fall.
2. The degree of injury was correlated positively to increasing age.
3. The male residents in this study had a higher rate of falls and resultant injury than their female counterparts at both agencies; no reason for this is readily apparent.
4. Residents who required hospitalization after a fall experienced a post-hospitalization change in health status. After recovery from such a fall, 6% and 8%, respectively, were identified as being more incapacitated (that is, they were moved to an agency or unit that offered more nursing care).
5. No significant difference in drug usage was found between those residents who fell and those who did not fall, except that the latter group took significantly more antihypertensive drugs.
6. Environmental factors related to frequency of falls include:
a. The time of day. This varies somewhat between the agencies but does seem to correlate directly with peak activity periods at the given agency.
b. The day of week. Interestingly, Monday was a low fall rate day at both agencies, with weekends being the highest rate days. Explanation is not conclusive, but staffing on the weekend days was less than on weekdays and a survey of staffing patterns did indicate that more staff members were scheduled on Mondays.
c. The time of year. Falls occurred more frequently in January, March, November, and December. This could be attributed to inclement fall and winter weather, except that Agency A was located in a mild-weather area, and 85% of these falls occurred indoors.
d. Design of agency. The floor plan of Agency A offered a unique opportunity for a relatively controlled study comparing round versus rectangular nursing units. The significant finding of fewer falls and injuries in the round design unit suggests that this should be considered by those planning and designing care facilities for the elderly.
Some of the findings of the study add support to the findings of previous studies. A problem with most of the studies is that they were retrospective and, consequently, scientifically weak. Predictive studies are needed to validate the solutions and actions that have been proposed to prevent and reduce falls by elders in care facilities.
Nursing practice actions that can be proposed based on the various findings include:
1. Increasing staff ratio when the care facility residents include (a) those over 75, (b) highly physically dependent persons, and (c) men. The data from this study also support the need for greater staff ratio when the floor plan is rectangular, with the nursing station not equidistant to all rooms.
2. Close monitoring of residents receiving drugs that could affect balance and gait. The residents at Agency B who receive antihypertensives are monitored regularly by the nurses. The monitoring includes assessment of vital signs, history of dizziness, weakness, etc., and monthly blood tests. This close observation by the nurses may explain why residents taking greater numbers of antihypertensives had fewer falls. These data suggest that optimal use of the drug was occurring.
Some of these actions can be facilitated by nurses using their professional status and nursing organizations to get regulations modified to require consideration of these factors in determining minimum staffing requirements. In addition, nurses can share their knowledge and expertise with legislative bodies to keep them informed of the need for providing sufficient monies to meet these staffing requirements. Keeping these elders as healthy as possible (that is, preventing falls and reducing resultant injury) should save money in the long run by preventing situations that require more costly medical and hospital care. Prevention of falls also should reduce the level of debility of the elderly and, in turn, reduce the need for more costly care.
Some agreement in study findings can be shown in factors that are present when elders in care facilities fall. It now seems appropriate to conduct predictive studies to validate the effectiveness in reducing falls and the resultant injuries of elders in care facililties through implementation of the various actions proposed.
- 1. Waller JA: Injury in aged, clinical and epidemiological implications. NY State J Med 1974; I2(I2):2200-2207.
- 2. Cooper S: Common concern: Accidents and older adults. Geriatric Nursing 1981; 2(4)^89-290.
- 3. Smith C: Accidents and the elderly. Nursing Times 1976; 72(48): 1872-1874.
- 4. National Safety Council. Accident Facts, 1980 Edition. Chicago. National Safety Council, 1980.
- 5. Margulec 1, Librach G, Schädel M: Epidemiological study of accidents among residents of homes for the aged. J Gerontol 1970; 25(4):342-346.
- 6. Sheldon JH: On the natural history of falls in old age. Br Med J 1960; 2:16851690.
- 7. Miller M, Elliott D: Accidents in nursing homes: Implications for patients and administrators, in Current Issues in Clinical Geriatrics. New York, The Tiresias Press. 1979.
- 8. Feist R R: A survey of accidental falls in a small home for the aged. Journal of Gerontological Nursing 1978; 4(6): 15-17.
- 9. Boucher CA: Accidents among old persons. Geriatrics 1959; 14(5):293-300.
- 10. Rodstein M: Accidents among the aged: Incidence, causes and prevention. Journal of Chronic Disability 1964; 17:515526.
- 11. Gryfe CI, Amies A, Ashley MS: A longitudinal study of falls in an elderly population: Incidence and morbidity. Age Ageing 1977; 6:201.
- 12. Agate J: Accidents to old people in their homes. Br Med J 1966; 2:785-788.
- 13. Kalchthaler T, Bascon RA, Quintos V: Falls in the institutionalized elderly. J Am Geriatr Soc 1978; 26(9):424-428.
- 14. Gould G: A survey of incident reports. Journal of Gerontological Nursing 1975; 1(4):23-26.
- Iskrant AP, Joliet PV: Accidents and Homicide. Cambridge, MA, Harvard University Press. 1968.
- Jacobson SB: Accidents in aged: Psychological and psychiatric viewpoint. NY State J Med 1974; 13( I ):24 17-2420.
- Mortality for leading types of accidents. Statistical Bulletin Metropolitan Life Insurance Company 1978; 59(10).
- Overstall PW: Prevention of falls in the elderly. J Am Geriatr Soc 1980; 28(1 1):481483.
- Recht JL: Accident Facts- 1973. Chicago, National Safety Council, 1973.
- Rodstein M: Interrelations of the aging process and accidents. J Am Geriatr Soc I972;20(3):97-I01.
- Witte NS: Why the elderly fall. Am J Nur s 1979;79:1950-1952.
FALLS AND THE ELDERLY- SUMMARY
LOCATION OF THE FALLS
LEVEL OF TREATMENT REQUIRED FOR RESIDENT INJURY