Screening, education, and associated behavioral responses to reduce risk for falls among people over age 65 years attending a community health fair. (Research Report).

Approximately 30% of people over 65 years of age fall each year, (1-5) accounting for the majority of injury-related deaths among elderly people. (6) In addition, fear of falling is highly prevalent, particularly in women. (7) Because of the high risk of falling and the recognition that falling is a "geriatric syndrome," primary care practitioners and geriatricians often include fall-risk screening in their comprehensive geriatric assessments. (8,9)

Health fairs have been increasing in popularity and frequency over the last 2 decades. They are generally held in a community setting that is easily accessible and usually include health education and medical screening events staffed and coordinated by community organizations and volunteers. Health fairs are low-cost health promotion and prevention vehicles intended to reach a wide audience. (10-14) Various types of screening tests are frequently offered at health fairs for little or no charge to participants.

Some authors (15-17) have described the planning and implementation of screening practices, but few have evaluated the effect of screening. Evaluations that have been conducted have tended to focus on attendee satisfaction, (18-23) participation, (18-24) and knowledge gains, (18-24) rather than on changes in behavior. Little evidence exists that the educational materials presented at health fairs lead to changes in behavior. Several researchers (13,25,26) have examined specific health-risk behaviors. In 2 studies, one of children (25) and one of college students, (13) there was little change in health-risk behaviors related to sun exposure (among children) or chemical use (among college students) following attendance at a health fair that included educational components. In a third study, Francisco and colleagues (26) studied how to reduce serum cholesterol levels in adults who participated in a health fair. Participants in that study were screened twice, once before and once after an educational intervention. Their results demonstrated a reduction in serum cholesterol levels in subjects who participated in both the preintervention and postintervention screening. Although this finding may indicate a change in diet or exercise behaviors related to the educational intervention, 36% of the participants were lost to follow-up. Importantly, few researchers have examined whether health fair attendees follow up with their health care provider as the result of positive screening tests.

Screening at health fairs, like many other screening efforts, is often criticized. Inaccurate results due to poor screening tools or measurement error may lead to unnecessary costs when a false positive result is reported or may give an attendee a false sense of security when a false negative result is presented. Health-risk screening also is sometimes performed for impairments or pathologies that are not prevalent, which leads to poor predictive value for the screening tools used. (27) In addition, health-risk screening is sometimes conducted for impairments or pathologies for which there is no adequate treatment to improve the natural course of the disease. (28-30) Finally, screening activities are sometimes conducted without informing participants of the result. (28) Using a free screening event to draw participants to a health fair, but failing to present individuals with results, cannot improve knowledge or alter risk behavior.

Our study was designed to examine the effect of educational material delivered to participants in a fall-risk assessment station located in a community health fair setting. In addition to screening individuals for risk of falls, participants were offered verbal advice and written educational materials. The aims of this study were: (1) to determine if participants changed their fall-risk behavior following participation in the fall-risk assessment and educational sessions provided at a health fair, (2) to determine if the recommendation to seek further evaluation from a health care provider was followed by individuals who were determined from screening to be at risk for falling, and (3) to determine if individuals with scores of 45 or less on the Berg Balance Test were more likely to change their fall-risk behavior than those individuals with scores of 46 or more. We hypothesized that individuals identified as at high risk for falling would be more likely to follow recommendations than those who were not identified as being at high risk for falling.

Method

Subjects

Our subjects were 68 volunteers who participated in I of 2 fall screening and prevention seminars in April and August 2000 in Hennepin County, Minnesota, which includes the city of Minneapolis. Participants lived independently in the community, either in their own homes or in apartments. One of the sessions was incorporated into a health fair sponsored by the Center for Medicare and Medicaid Services (formerly the Health Care Financing Administration) and held in a large hotel on a weekday. The other session, also on a weekday, was held at a community center for elderly people. The session at the community center was requested by the center's education director and was offered as a free service to people who attended the center.

Participants in our study were self-selected; they either voluntarily approached the "Fall Screening and Education" station at the health fair or requested to be screened for risk of falls at the community center. Each participant was given an explanation of the study and was asked to provide informed consent for data collection. Individuals who chose not to participate were given the opportunity to participate in the screening process and the educational intervention with no collection of data. Participants were included only if they could transfer from one chair to another independently and stand for 10 seconds without the assistance of another person. Participants were allowed to use an assistive device (eg, cane or walker) to transfer and stand, if desired. Subjects were not allowed to use their assistive devices during the actual screening process because the assessment tool we used does not have a way to modify scores when there is use of an assistive device.

Subjects independently completed a demographic worksheet and questionnaire that asked them to provide information on birth date, sex, race, education, and 1-year fall history prior to completing the fall-risk screening with a physical therapist. On this worksheet, subjects wrote their telephone number and address for the follow-up portion of the investigation.

The subjects (18 male, 50 female) ranged in age from 57 to 89 years ([bar]X=74.62, SD=7.54). Two subjects declined to provide their age. Demographic information about the subjects is shown in Table 1. Two subjects indicated that their race was African American; 66 subjects indicated their race as Caucasian. Fifty-two of the participants were wearing glasses at the time of the screening. Eight subjects wore one or more hearing aids, and 12 subjects used either a cane (n=9) or a walker (n=3) to aid in ambulation. Twenty-one participants (3.1%) had experienced a fall in the year prior to data collection, most commonly in their homes (48%). Four of those participants who had experienced a fall in the year prior to data collection reported that the fall had resulted in a fracture.

Screening

Screening for risk of falls, educational intervention, and follow-up telephone calls to subjects were completed by 2 of the investigators (KKN, GHI). In an effort to increase consistency and reliability of the screening findings, subjects were assessed for risk of falls by a physical therapist (KKN) who is a certified geriatric clinical specialist and who regularly teaches other therapists how to administer the screening tool. The physical therapist used the Berg Balance Test (31) to determine each subject's risk for falling.

The Berg Balance Test (31) is a 14-item scale that is designed to measure 3 dimensions of balance (Fig. 1). Subjects performed activities to maintain a posture and activities that included movement, and they responded to external perturbations of a posture. Each of the items was scored on a 5-point (0-4) ordinal scale, with a score of 0 representing the inability to complete the task and a score of 4 indicating independence. Independence on all 14 items results in a total possible score of 56. The Berg Balance Test was originally examined for validity (32) and for reliability (33) in Canada with data from 113 residents of a home for elderly people (mean age = 71.6 years, SD = 10.1) and 70 people with strokes (mean age=83.5 years, SD=5.3) in 2 acute care hospitals. It was compared with clinical judgments of balance, laboratory measurements of pastural sway, future falls, and motor performance to determine validity. Multiple therapists evaluated each participant to determine reliability. The Berg Balance Test has high internal consistency (Cronbach alpha=.96), indicating that the individual test items all measure the same underlying construct (in this case, fall risk), and high interrater reliability (r = .98) and high intrarater reliability (r = .99). (30)

Figure 1.

Berg Balance Test sequence.

1. Sitting to standing without using hands while stabilizing independently.

2. Standing unsupported for 2 minutes.

3. Sitting with back unsupported, but feet supported on the floor.

4. Sitting down from a standing position with minimal use of hands.

5. Transferring from a chair with arms to a chair without arms and back with minor use of hands.

6. Standing unsupported with eyes closed for 10 seconds.

7. Standing unsupported with feet together for 1 minute.

8. Reaching forward 25.4 cm (10 in) with arm in 90[degrees] of shoulder flexion.

9. Pick up slipper from floor from standing position safely and easily.

10. Turn to look behind left and right shoulders while standing.

11. Turn 360[degrees] in 4 seconds or less.

12. Place alternating feet on stool while standing unsupported 8 steps in 20 seconds.

13. Standing unsupported with one foot in front of the other for 30 seconds.

14. Standing on one foot for 10 seconds.

Although originally developed to measure balance in people over age 65 years rather than to predict falls, (31,32) the Berg Balance Test, when combined with fall history, has been reported to have a sensitivity of 91% and a specificity of 82% when a cutoff score of 40 or less is used to classify those community-dwelling people over 65 years of age who fall and those who do not fall. (34) Tsang, (35) commenting on a study by Bogle Thorbahn and Newton, (36) reported that when the Berg Balance Test was used to predict future falls (within 6 months), it had a specificity of 85% and a sensitivity of 82%. Harada et al (37) also reported a high specificity (91%) and sensitivity (82%) when the Berg Balance Test was used to predict future falls. In both of these studies, (36,37) a score of 45 or lower was used to identify those at risk for falls; therefore, in our study, we used a score of 45 or less to identify those individuals we considered at high risk for falls.

The tools used to complete the Berg Balance Test in this study were a chair with arms and a 40.6-cm-high (16-in-high) seat, a chair without arms and a 40.6-cm-high seat, a yardstick, a stopwatch, a slipper, and a 25.4-cm-high (10-in-high) stool. Fourteen individuals (21%) scored at or below 45 on the Berg Balance Test and were identified by us to be at high risk for a fall. Of these, 9 individuals (64%) had experienced a fall during the year prior to data collection, and 2 of the falls resulted in upper-extremity fractures.

Educational Session

Once they had completed the Berg Balance Test, participants were directed by the examining physical therapist to one of the other examiners to review their results. Either a postdoctoral gerontology fellow or a physical therapist student, who had been previously trained by the primary author (KKN) to carefully review the results of the screening, provided the subjects with the results of the assessment. If an individual's score was equal to or less than 45 on the Berg Balance Test, the participant was told that he or she may be at risk for a fall. For these individuals, a letter was provided to take to their physician that explained the assessment procedure and the results of screening.

In addition to reviewing the results of the screening, the postdoctoral gerontology fellow or the physical therapist student provided a brief educational session for each participant. The educational session included a review of activities, behaviors, and environmental adaptations that may reduce the risk of falling. All participants were given the same list of tips to prevent falls (Fig. 2), which was based on a patient information excerpt from American Family Physician. (38) The list was printed on colored paper and was available in English, Spanish, and Hmong translations. Participants were encouraged to ask questions.

Figure 2.

Educational material--tips to reduce risk of falling. (38)

* Remove or secure throw rugs

* Cover wiring or lamp cords

* Exercise regularly

* Eat a calcium-rich diet

* Have your medications reviewed by your physician

* Have your hearing checked

* Have your vision checked

* Limit your alcohol intake

Follow-up

Thirty days after the initial screening and intervention, we conducted follow-up telephone interviews that included use of a structured questionnaire. Participants were asked if they had experienced a fall since the screening, and if so, if the fall had resulted in a fracture. Subjects were also asked if they had visited their physician, if they had seen a physical therapist, and if they had changed anything based on the handout given to them at the time of the initial screening. Additional screening questions included an inquiry about fear of falling ("Are you more afraid that you might fall following participation in fall risk screening than you were before?") and self-imposed activity limitations due to fear of falling ("Do you limit your activities because you are afraid you might fall?").

Data Analysis

Descriptive statistics (means, standard deviations, ranges, percentages, and frequency counts) were calculated for age, race, sex, use of assistive devices, and history of previous falls. To determine whether groups (high-risk and low-risk) differed in their adoption of one or more recommendations for risk reduction, a chi-square test was done.

Results

One of the initial subjects declined to participate at the time of the follow-up telephone call. Thus, the data obtained for the remaining 67 subjects were included in the follow-up data analysis. Six participants (9%) experienced a fall between the time of screening and the 30-day follow-up telephone call. Four of the participants who fell had been identified by the screening procedure to be at high risk for a fall. The other 2 people who fell before the 30-day follow-up telephone call had scores of 49 and 54 on the Berg Balance Test. Neither had a history of falling.

Based on self-report from the follow-up telephone interview, the mean number of recommendations implemented by participants was 1.75 (SD=1.7, range=0-8). At least one risk-reduction behavior was reported by 72% of the participants. The most common behavioral change was getting a vision examination (n=22). Adding an exercise program to the subject's daily routine was the second most common behavioral change made by participants in our study (n-21). Figure 3 presents the number of people making behavioral changes to reduce fall risk. The frequency of each behavioral change is reported in Table 2.

[FIGURE 3 OMITTED]

Among subjects considered as having a high risk for falling based on the screening, 46% followed our recommendation to visit their physician during the 30-day follow-up period, and 92% reported implementing at least one risk-reduction behavior. These findings contrast with the 36 individuals who were not identified to be at high risk for falling, 67% of whom reported at least one recommended change in behavior to reduce fall risk. Although the difference in risk reduction between the high-risk and low-risk groups was substantial, given our relatively small sample size in this project, the proportional difference was not statistically significant (P=.07).

Discussion

Although the 68 subjects in our study population represent a self-selected sample of community health fair attendees rather than a random sample, the reported fall incidence in the year previous to data collection (31%) was similar to that reported in other studies of community-dwelling people over age 65 years (30%). (1-5) The elderly subjects in our study were less likely to fall at home than the general population. In the general population, between 60% and 75% of fails are reported to occur at home. (39) Tripping over an object and slipping on a level surface are the most common ways elderly people fall. (39) Forty-eight percent of the subjects in our study fell at home. A large number of the participants in our study wore glasses, and some had hearing impairments. Although visual limitations are reported to increase fall risk, (1) in one study that examined the validity of the Berg Balance Test to predict fall risk, both severe visual and hearing impairments were exclusion criteria. (34) We do not believe that any of the participants in our study had visual or hearing impairments that interfered with their ability to perform the items on the Berg Balance Test.

Twelve of the subjects in our study used either a cane or a walker for ambulation. The percentage of people using assistive devices in our study (18%) was higher than in one study (5%) (34) and smaller than in 2 other studies (33% (32) and 50% (37)) that were designed to examine the validity of Berg Balance Test scores.

Almost three quarters of the participants in our study followed at least one of the risk-reduction strategies presented to them during the educational session at the health fair. Although we did not have a control group of subjects with whom to compare our study groups, we believe our high level of response provides evidence to support the efficacy of this low-cost intervention program.

Conclusion

Information on fall risk obtained in a community health fair setting appeared to promote risk-reduction efforts among fair attendees. People identified as and informed that they were at high risk for falls were more likely to make behavioral changes than those whose were not identified as and informed that they were at high risk for falls. A larger study with a control group that does not receive an educational intervention and with numbers that allow for sufficient power is needed. Such a study could determine whether or not the screening and educational intervention we used is efficacious in reducing the incidence of falls among elderly people.

Figure 3.
Number of participants making behavioral changes to reduce fall risk.

           High-Risk Group   Low-Risk Group

0                 1               18
1 or More        12               36
2 or More         7               25
3 or More         5               14
4 or More         3                8
5 or More         2                4
6 or More         1                2

Note: Table made from bar graph.

Table 1.
Demographic Information and Screening Results

                                 Number
Variable                         (Percentage)

Race
  Caucasian                      66 (97%)
  African American                2 (3%)
Wearing glasses
  Yes                            52 (76%)
  No                             16 (24%)
Wearing 1 or more hearing aids
  Yes                             8 (12%)
  No                             60 (88%)
Using assistive device
  Yes                            12 (18%)
  No                             56 (82%)
Fall in last year
  Yes                            21 (31%)
  No                             47 (69%)
Berg Balance Test score
  45 or less                     14 (21%)
  46 or greater                  54 (79%)

Table 2.
Individual Behavioral Changes

                                 No. of Subjects Who Changed This
                                 Behavior

                                 Berg Balance Test  Berg Balance Test
                                 Score of 45 or     Score of 46 or
Behavioral Change                Less (n = 13)      More (n = 54)

Going to primary physician       6                   4
Going to a physical therapist    4                   1
Removing or securing throw rugs  2                   4
Covering wiring or lamp cords    1                   4
Exercising regularly             8                  13
Eating a calcium-rich diet       3                  17
Having medications reviewed by
  your physician or pharmacist   4                   9
Having hearing checked           0                   8
Having vision checked            3                  19
Limiting alcohol intake          0                   9

References

(1) Tinetti ME, Speechley M, Ginter SF. Risk factors for falls among elderly persons living in the community. N Engl J Med. 1988;319: 1701-1707.

(2) Blake AJ, Morgan K, Bendall MJ, et al. Falls by elderly people at home: prevalence and associated factors. Age Ageing. 1988;17:365-372.

(3) Prudham D, Evans JG. Factors associated with falls in the elderly: a community study. Age Ageing. 1981;10:141-146.

(4) Campbell AJ, Reinken J, Allen BC, Martinez GC. Falls in old age: a study of frequency and related clinical factors. Age Ageing. 1981;10:264-270.

(5) Downton JH, Andrews K. Prevalence, characteristics and factors associated with falls among the elderly living at home. Aging. 1991;3: 219-228.

(6) Baker SP, Harvey AH. Fall injuries in the elderly. Clin Geriatr Med. 1985;1:501-512.

(7) Arfken CL, Lach HW, Birge SJ, Miller JP. The prevalence and correlates of fear of falling in elderly persons living in the community. Am J Public Health. 1994;84:565-570.

(8) Palmer RM. Geriatric assessment. Med Clin North Am. 1999;83: 1503-1523.

(9) Jacob GM, Palmer RM. Tools for assessing the frail elderly. Geriatric evaluation focuses on improving quality of life. Postgrad Med. 1998; 104: 135-153.

(10) Bogin R. 9 Health Fair. Colo Med. 1993;90:440-441.

(11) Gardner J, Nickolaus MJ, Steckbeck R. Health fair & CPR training: a successful community outreach, partnering, & marketing opportunity. J Cardiovasc Manag. 1999;10:18-27.

(12) Oase SJ, Tracy JP. Health fairs for older adults: the big day. Geriatr Nurs. 1982;3:174-175.

(13) Werch CE, Schroeder DA, Matthews LL. The health fair as a health promotion strategy: effects on health risk behaviors and the utility of specific health fair activities. J Am Coll Health. 1986;35:74-79.

(14) Serina M, Glove S. Health fairs bring wellness concepts to life. Nurse Educ. 1991;16:7-11.

(15) Begg E. Health promotion: planning a health fair. Community Outlook. October 11, 1989:19-22.

(16) Dillon DL, Sternas K. Designing a successful health fair to promote individual, family, and community health. J Community Health Nurs. 1997;14:1-14.

(17) Wilson LC. Implementation and evaluation of church based health fairs. J Community Health Nuts. 2000;17:39-48.

(18) Evans RG, Decker B. A smoking and health fair: implementation and evaluation. J Am Coll Health. 1987;36:172-174.

(19) Jones MS, Siegrist B. Health promotion for farmers at the country market. Nurse Educ. 1999;24:16-18.

(20) Bryan J, Deveraux JM, York ML, Schoh RJ. How effective are health fairs? Quantitative evaluation of a community health fair. Am J Health Promot. 1991;6:85-88.

(21) Heath JM, Lucic KS, Hollifield D, Kues JR. The health beliefs of health fair participants. J Community Health. 1991;16:197-203.

(22) Mess S, Reese PP, Lana DF, et al. Older, hypertensive and hypercholesterolemic fairgoers visit more booths and differ in their health concerns at a community health fair. J Community Health. 2000;25: 315-329.

(23) Macias EP, Morales LS. Utilization of health care services among adults attending a health fair in South Los Angeles County. J Community Health. 2000;25:35-46.

(24) Ness KK, Gurney JG, Ishani S. Screening for risk of falls in the elderly in a community health fair setting: a feasibility study. Issues on Aging. 2001;24:12-16.

(25) Buller MK, Goldberg G, Buller DB. Sun smart day: a pilot program for photoprotection education. Pediatr Dermatol. 1997;14:257-263.

(26) Francisco VT, Paine AL, Fawcett SB, et al. An experimental evaluation of an incentive program to reduce serum cholesterol levels among health fair participants. Arch Fam Med. 1994;3:246-251.

(27) Hennekens CH, Buring JE, Mayrent SL. Epidemiology in Medicine. Boston, Mass: Lippincott Williams & Wilkins Publishers; 1987:327-347.

(28) Allanson JF. The dubious values of community health fairs: a position paper. Health Val. 1982;6:17-19.

(29) Goldman KD, Kloss S, Manes T, Rojewski M. Following up on screening tests. Public Health Rep. 1998;113:100-101.

(30) Berwick DM. Screening in health fairs: a critical review of benefits, risks and costs. JAMA. 1985;254:1492-1498.

(31) Berg KO, Wood-Dauphinee S, Williams JI, Gayton D. Measuring balance in the elderly: preliminary development of an instrument. Physiotherapy Canada. 1989;41:304-311.

(32) Berg KO, Wood-Dauphinee SL, Williams JI, Maki B. Measuring balance in the elderly: validation of an instrument. Can J Public Health. 1992;83 (suppl 2):S7-S11.

(33) Berg KO, Wood-Dauphinee S, Williams JI. The balance scale: reliability assessment with elderly residents and patients with acute stroke. Scand J Rehabil Med. 1995;27:27-36.

(34) Shumway-Cook A, Baldwin M, Polissar NL, Gruber W. Predicting the probability for falls in community-dwelling older adults. Phys Ther. 1997;77:812-819.

(35) Tsang R. Berg Balance Test [letter to the editor]. Phys Ther. 1996;76:1126.

(36) Bogle Thorbahn LD, Newton RA. Use of the Berg Balance Test to predict falls in elderly persons. Phys Ther. 1996;76:576-585.

(37) Harada N, Chiu V, Damron-Rodriguez J, et al. Screening for balance and mobility impairments in elderly individuals living in residential care facilities. Phys Ther. 1995;75:462-469.

(38) Fuller GF. Patient information: "What causes falls in the elderly? How can I prevent a fall?" Am Fam Physician. 2000;61:2173-2174.

(39) Sorock GS. Falls among the elderly: epidemiology and prevention. Am J Prev Med. 1988;4:282-288.

KK Ness, PT, MPH, GCS, is Research Assistant, Department of Pediatrics, School of Medicine, University of Minnesota, Minneapolis, MN. Address all correspondence to Ms Ness at Division of Pediatric Epidemiology and Clinical Research, University of Minnesota, MMC 715, 420 Delaware St SE, Minneapolis, MN 55455 (USA) ([e-mail]).

JG Gurney, PhD, is Associate Professor, Division of Pediatric Epidemiology and Clinical Research, Department of Pediatrics, University of Minnesota.

GH Ice, PhD, is Assistant Professor, Department of Social Medicine, College of Osteopathic Medicine, Ohio University, Athens, Ohio.

Ms Ness and Dr Ice provided concept/idea/research design and data collection. All authors provided writing. Ms Ness and Dr Gurney provided data analysis. Ms Ness provided project management and clerical support. Dr Gurney provided consultation (including review of manuscript before submission),

This study was approved by the Institutional Review Board at the University of Minnesota.

This article was submitted September 23, 2002, and was accepted February 10, 2003.