Improving Adherence With the Use of Hip Protectors Among Older People Living in Nursing Care Facilities: A Cluster Randomized Trial

Article Outline

• Abstract
• Methods
  • Participants
  • Randomization
  • Intervention
  • Outcomes
  • Statistical Analysis
  • Sample Size Calculation
• Results
  • Educational Sessions
  • Adherence
  • Prediction of Adherence
  • Falls and Injuries
  • Reasons for Noncompliance
• Discussion
• Conclusion
• Acknowledgment
• References
• Copyright

Objectives

To test different adherent strategies aimed at improving hip protector use among nursing care facility residents.

Design, Setting and Participants

A cluster randomized controlled trial with 234 residents older than 75 years from 9 units of 7 nursing care facilities in the Northern Sydney region, Australia.

Intervention

Residents were cluster randomized in 3 groups. The first group received hard shell hip protectors without cost (no cost group). The second group received an educational session, a demonstration of the use of hip protectors, and free choice of type of hip protectors without cost (combined group). The third group was the control group who received a brochure about hip protectors.

Measurements

Primary outcome was adherence with the use of hip protectors at 3 and 6 months after recruitment. Secondary outcomes were falls, injuries, and fractures.

Results

No participants in the control group purchased hip protectors at any stage. At 3 months, 33% of participants in the no cost group and 27% in the combined group wore a hip protector at the time of visit. This declined to 25% and 24% respectively at 6 months. No significant difference was seen in any of the 3 adherence outcomes between the 2 intervention groups. The number of falls or hospitalizations did not differ between groups, with 5 hip fractures reported during the intervention period. Residents were more likely to be adherent if they were female and had greater restriction in daily activities.

Conclusion

Providing free hip protectors to older people living in nursing care facilities was necessary to increase initial acceptance and adherence. Nevertheless, after 6 months the achieved level of adherence was not high enough to be associated with a reduction of hip fractures. The provision of educational sessions and demonstrations to nursing staff and participants had no added value in this trial.

Keywords: Adherence, hip protector, nursing care facilities, falls prevention

The effectiveness of external hip protectors in older people at high risk of hip fracture was first reported in 1988.¹ Since then, 14 randomized trials of hip protectors in residential aged care facilities have been published with varying results on their efficacy.², ³, ⁴ The current Cochrane Collaboration systematic review of hip protectors included 11 of these 14 randomized trials carried out in nursing homes and concluded “accumulating evidence casts some doubt on the effectiveness of the provision of hip protectors in reducing the incidence of hip fractures in older people.”⁵ However, for frail older people in institutional care, pooled data do indicate a significant positive effect of hip protectors. Acceptance and adherence by users of hip protectors remains poor primarily because of discomfort and practicality.⁵, ⁶ Hip protectors can be effective only when worn as instructed, and correct adherence with their use has been reported as a problem. Across a range of studies, hip protector adherence ranged from 20% to 92% with a median of 56%.⁷ If hip protectors are to be effective, adherence with their use needs to be improved.⁵

In studies reported thus far, attention has mainly been paid to the features of the hip protector garment with concerns being raised about comfort, appearance, fit, and manageability. But it is also important to examine the characteristics of the person being asked to wear the hip protectors, as in many cases it is the older person's motivation to wear the hip protectors that mostly affects adherence.⁸ The social environment and external support are also important factors, particularly for older people living in residential care. Knowledge and attitudes of nursing staff (particularly senior nursing staff) in these facilities may have a substantial effect on whether an older person wears hip protectors.⁹, ¹⁰, ¹¹ This view was substantiated by a recent prospective cohort study suggesting that caregivers' understanding of osteoporosis and hip protectors does influence adherence, and consequently educational programs should be designed to improve adherence.¹²

The cost of purchasing hip protectors has been identified as a barrier to their use in a number of studies and also in routine practice.¹³, ¹⁴ In Australia the costs of hip protectors vary from approximately $A120 to $A270 for 3 pairs. Generally it is accepted that 3 pairs are the minimum necessary for full-time use given the need to wash the underwear. The supply of hip protectors at no cost and the delivery of an adherence-improving strategy is an intervention that is potentially feasible for routine use. This could occur if subsidies for purchase of hip protectors were available, and if Aged Care Services provided a limited program of education and support for their use as part of a falls and injury prevention program.

The trial that is reported here investigated the effects of strategies to increase adherence with use of hip protectors. Two hypotheses were tested: (1) the supply of hip protectors without direct financial cost (no cost group) increases adherence with their use, (2) the use of an individualized adherence strategy and supply of hip protectors without cost to the user (combined group) further increases adherence with their use. A control group received a currently available brochure about hip protectors (control group).

Back to Article Outline

Methods 

Participants 

A sample of 7 of the larger nursing care facilities in the Northern Sydney area agreed to participate in the trial. Residents were invited to participate after approval of nursing staff of each facility. Inclusion criteria were the following: aged 75 years or older; likely to survive for more than 12 months as assessed by the Illness Severity Rating¹⁵; at least 1 hip without previous fracture or arthroplasty. In addition, participants were assessed as being at high risk of hip fracture using the modified FREE study algorithm (low body weight (<70 kg for women and <90 kg for men) and able to stand without assistance for at least 5 seconds.¹⁶ Consent was sought from individual residents, and for those unable to give consent, from family members with knowledge of the intervention to be undertaken. Baseline measures in this study therefore occurred after randomization.

Randomization 

A random numbers table and concealed opaque numbered envelopes of clusters were used for the randomization procedure. A cluster was defined as a residential aged care facility by itself or an independently working unit within a large facility. The outcome assessors were masked to the allocation.

Intervention 

The control group was provided with a brochure about hip protectors that included a contact number for suppliers of hip protectors. This method has been used in a comparable study by Meyer and colleagues.¹³ The no cost group was supplied with 3 pairs of correctly sized, hard shell hip protectors (Hornsby Healthy Hips) and a brochure providing information on their use. The research nurses supplying the hip protectors liaised with nursing and other staff to check that hip protectors were fitted correctly and to explain how they should be worn. Replacement of worn out or lost hip protectors was not offered. The combined group also received 3 pairs of hip protectors but they could choose between energy-diverting protectors (hard shell protectors, Hornsby Healthy Hips) or energy-absorbing hip protectors (soft protectors by Hip Saver). Next, they also received an educational program designed to enhance adherence based on the study of Meyer et al.¹³ Two registered nurses who had substantial experience with hip protectors provided informal educational sessions, supported by written materials and computer presentations based on social learning theory, for the nursing staff providing direct care to the residents using hip protectors. These aimed to encourage recommended use of hip protectors and development of an adherence plan for each participant. One nursing member was appointed in each facility to monitor adherence. The number of educational sessions varied depending on the size of the facility. All participants were also individually educated and investigators followed up at least twice.

All participants in the study received health and medical care as usual. No education or training about falls prevention was part of the study, and participating residential aged care facilities were asked not to implement new falls or injury prevention strategies during the study.

Outcomes 

Baseline measures included demographic details, history of falls and fractures, functional status using the Barthel Index,¹⁷ cognitive function (Short Portable Mental Status Questionnaire [SPMSQ]),¹⁸ and quality of life using the EQ-5D.¹⁹

The primary outcome (adherence with use of hip protectors) was assessed at visits 3 months and 6 months after enrollment. We used 3 measurements to assess adherence in line with the recommendations from the International Hip Protector Research Group²⁰: (1) actual participant adherence at each visit was recorded as wearing or not wearing hip protectors at the time of visit; (2) mean adherence during the previous month was also noted by research staff and measured as a percentage of the waking hours that hip protectors were worn; (3) adherence was measured by attempting to ascertain whether the hip protectors were worn at the time that falls occurred.

Secondary outcomes were also recorded at 3 and 6 months. The occurrence of falls, injuries, and fractures was assessed from the records in the residential aged care facility and confirmed by other sources (general practitioner or hospital records). Nursing staff and participants were asked open-ended questions about treatment complications. The follow-up assessments were performed, as far as possible, on all randomized subjects regardless of their adherence with the experimental interventions.

Statistical Analysis 

Data were coded to permit blinding to group allocation in the statistical analysis. Differences in the primary outcome measure were analyzed using the continuity corrected χ² test and adjusted for cluster randomization.²¹ Separate analyses were performed on 3-month and 6-month follow-up data. Secondary outcomes were examined using analyses of variance (or using nonparametric techniques where appropriate) on the 6-month follow-up data. Pretest variables were examined, and possible associations with adherence were tested using a univariate analysis. Analysis was by “intention-to-treat.”

Sample Size Calculation 

Calculations were based on statistical power of 80% (with α = 0.05, 2-sided test). Anticipated adherence at 6-month follow-up was 8% in the control group, 30% in the no cost group, and 60% in the combined group. For this study, 9 clusters with an average size of 28 participants were recruited. The sample size for the cluster randomized studies included an inflation factor of 0.01 that assumed small variance between clusters, and medium to large variance within clusters.²² Power was calculated to establish significant differences between any of the 3 intervention arms.

Back to Article Outline

Results 

Over the course of 13 months we located 638 eligible participants from the 7 nursing care facilities. From this group 148 participants did not meet our inclusion criteria and a further 255 participants declined to participate. The 7 facilities were divided into 9 clusters, with 1 facility being divided into 3 clusters based on the units of the nursing care facility. After randomization, 235 participants were divided into (1) the no cost group with 55 participants, (2) the combined group with 84 participants, and (3) the control group with 96 participants. The numbers of participants in each group varied substantially because of the varying size of the cluster or facility (Figure 1).

• 
  • View Large Image
  • Download to PowerPoint

• Fig. 1 

  Flow of participants.

Baseline characteristics of participants are provided in Table 1. Participants had mean age of 86 years and 82% were female. Health status was considered good to excellent by 53% of all participants. Residents generally were able to walk without the assistance of another person. The control group had the least number of errors in the SPMSQ test and scored better on both the EQ-5D regarding the usual activities category and the EQ-VAS. The no cost group scored the most problems for all categories of the EQ-5D including the EQ-VAS except for anxiety.

Table 1. Baseline Characteristics of (n = 235) Participants by Study Group

Educational Sessions 

All planned educational sessions were attended by at least one member of the nursing staff. However, overall attendance by the nurses was incomplete despite provision of the sessions at varying times to coincide with each of the 3 daily shifts. Other issues influencing provision of education reported by the field investigators were the following: limited English language skills, limited level of education of some nursing staff, and often very limited time for the education so that only a shortened version was actually given. In some instances the incentive for hip protector use was low because nursing staff felt there were already too many existing standards and regulations that they perceived as burdensome, or because prejudice against hip protectors existed. This resulted in resistance to change and improved existing models of care. All participants in the combined group received education and follow-up as planned.

Adherence 

Ninety-four percent of the combined group chose to wear a soft shell hip protector. No participants in the control group purchased hip protectors at any stage. Table 2 provides details of adherence with the use of hip protectors. At 3 months, 33% in the no cost group actually wore a hip protector at the time of the visit compared with 27% in the combined group. At 6 months these percentages had declined to 25% and 24% respectively. The estimated mean adherence in the previous month at the 3-month assessment was 51% for the no cost group, and 36% for the combined group. At 6 months this had declined to 36% and 34% respectively. Using a report of the hip protector being in place at the time of a fall as a measure of adherence this was recorded in only 21.0% of falls in the no cost group over the 6 months of the study, and declined from 26.7% at 3 months to 6.5% at 6 months for the combined group. In none of the 3 measurements for adherence was there any statistical significant difference measured between the 2 intervention groups.

Table 2. Adherence^∗ With Use of Hip Protectors, at 3- and 6-Month Follow-up Assessments (Individuals and Institutions)

HP, hip protectors.

Prediction of Adherence 

As adherence did not significantly differ between the 2 intervention groups we analyzed independent predictors for adherence at 6 months for both groups together. For this analysis we computed a dichotomous outcome “adherent to hip protector use” when there was a report of more than 75% adherence in the previous month. Only 2 significant predictors were detected: females were more adherent to hip protectors than males; and those who reported serious problems with usual activities at baseline were more adherent at 6 months than those reporting no problems with usual activities at baseline. No other baseline variables showed any association with adherence, and in particular the allocated cluster had no predictive value (see Table 3).

Table 3. Predictors of Adherence to Hip Protectors at 6 Months (Defined as Mean Adherence ≥ 75% in Previous Month)

SPMSQ, Short Portable Mental State Questionnaire.

Falls and Injuries 

There was no difference in the number of falls between the control and intervention groups (see Table 4). Similarly, there was no difference in the percentage of fallers among the 3 groups (F = 0.234, P = .792). Falls with recorded injury were relatively common (about 0.3 per participant) and did not vary among groups. Five hip fractures occurred during the study: 1 in the control group, 2 the no cost group (1 person while not wearing a hip protector because of illness and hospitalization, and 1 person while wearing the hip protectors), and 2 occurred in the combined group while the hip protectors were not being worn because of nonadherence.

Table 4. Falls Outcomes at 6 Months^∗

Reasons for Noncompliance 

Problems with the hip protector because of lack of comfort at 6 months was strongly associated with low adherence (Spearman's rho 0.357, P = .00). Concerns with appearance, or extra effort in using hip protectors, showed no association with adherence. After 6 months, 4 participants in the no cost group and 13 in the combined group reported they had to take the hip protector off because it felt too hot or was too uncomfortable to wear. Six participants in the no cost group and 7 in the combined group mentioned issues with appearance or style when wearing hip protectors. Other possible reasons for nonadherence such as bulkiness, problems with continence, and needing help to put them either on or off was reported by 20 participants in the no cost group and 29 in the combined group. Forty-nine percent of the no cost group and 67% of the combined group agreed that 3 pairs of hip protectors was considered enough. Reasons for wanting more than 3 pairs were related to participant continence or issues surrounding laundering the garments, including the time the laundry took to clean and return the hip protectors to participants.

Back to Article Outline

Discussion 

The study has shown that providing hip protectors at no cost to older people in nursing care facilities is associated with higher adherence than the provision of a brochure about hip protectors. This suggests that if hip protectors are recommended to be worn by older people in these facilities they should be provided at no or limited cost. No substantial difference was found between providing correctly sized hip protectors with little or no follow-up contact, and a somewhat more detailed program of hip protector provision with additional education to the user and nursing staff of the residential aged care facilities, which also included ongoing contact with a nurse to encourage hip protector use.

As expected, adherence in the 2 intervention groups was higher at 3 months than 6 months after supply of the hip protector (mean adherence in previous month at 3 months was 42% and at 6 months was 35%). Nonetheless, the level of adherence was overall modest in this study and lower than the median of 56% from most reports on adherence.⁷ Although still more research is needed to ascertain what level of adherence would be sufficient to reduce hip fractures, this result is unlikely to be sufficient enough to be associated with clinically useful reductions in hip fractures.

The strength of this research is that it is the first study that compares 2 types of strategies to enhance adherence. The study by Meyer et al.¹³ evaluated the effect of a similar educational and enhancement strategy of nursing staff with the provision of free hip protectors to residents of nursing care facilities. That study showed that protectors were used during 54% of falls compared with 8% in the control group, resulting in a relative reduction of 40% in occurrence of hip fractures. With the addition of the strategy in our study to provide older people with hip protectors at no cost, it now seems unclear whether the education to nursing staff has any added value. Nevertheless it should be remembered that the education received was less than optimal in many facilities owing to a variety of factors. In this study, the investigators' experience was that the actual education differed in time and intensity depending on the time made available in the facility or unit of the facility. The interest and understanding of the nursing staff differed between the facilities, sometimes because of language issues or high turnover in staff. There were facilities that were well organized with 1 nurse appointed as coordinator, but there were also facilities where it was a struggle to get all nursing staff involved in the educational sessions. Despite this difference between facilities, no effect of cluster was found in the adherence at 6 months for the 2 intervention groups. Apparently these differences in educational sessions had no effect on adherence.

A potential weakness of this study is the use of a cluster design, which may increase the risk of bias. However, so far mostly individual randomized controlled trials failed to show strong positive effect of hip protector use, whereas cluster randomized trials have had more positive results.⁶ We think that the way data were analyzed in this study, taking into account the possible cluster effect, diminished possible bias to a minimum.

It has been suggested that residents who are perceived by nursing staff to be at greater risk of falling and are more dependent on help are more likely to wear hip protectors.²³ Education on risk of falling may therefore be most effective if directed at nursing staff, who are in the best position to advise and influence residents and their relatives. This is substantiated in this study as we found that serious problems with usual activities (according to the EQ-5D) were associated with a higher adherence at 6 months than no problems with usual activities. However, we did not find any association between serious problems with self-care or mobility and high adherence. The experience of our investigators was that in care facilities where residents had cognitive deficits, multiple medical comorbidities, or there was a low level of nursing staff, the focus of care becomes more basic such as showering, feeding, and toileting. This is at the expense of other aspects of care such as strategies that may encourage independence, eg, use of hip protectors, falls education, exercise, and so forth.

The measured correlation between nonadherence and lack of comfort is in line with earlier reports on reasons for not wearing hip protectors such as lack of comfort (too tight or poor fit), lack of style, the extra effort needed to put them on when getting dressed, urinary incontinence and physical difficulties/illnesses.⁵, ⁷, ¹⁰ Lack of comfort was mainly reported by the combined group (20%), but it was precisely this group that had free choice in the type of hip protector (94% chose a soft shell hip protector). The no cost group, who received hard shell hip protectors, reported lack of comfort only in 10.5% of the cases. Literature suggests there is no difference in adherence between soft and hard shell hip protectors.², ²³ Our study confirms this finding as no difference in adherence was measured between the 2 groups.

Earlier studies indicated that high adherence rates for hip protectors in long-term care facilities are feasible and success depends in part on whether there is broad-based acceptance by support staff.²⁴ The process of making hip protectors an integral part of the daily routine for each resident requires a person or team to assume accountability, not only for measuring adherence, but also for attending to small details such as measuring, ordering, and laundering the hip protectors.²⁴ In this trial, the educational sessions to the nursing staff did not accomplish this. Future trials measuring the effect of education on adherence with hip protectors should consider involving more senior nursing staff members, repeated sessions of education, and tactics aimed at overcoming preexisting prejudice against hip protector use. As no participant in the control group actually purchased a hip protector, it could be argued that supplying free hip protectors only helps to increase initial acceptance. For actual adherence more is necessary. Another issue is the high number of initial refusals (40%) to participate in the study. This high number of refusals and the low achieved adherence demonstrates the difficulty of testing the potential effectiveness of hip protectors.

The generalizability of this study could be limited as there may be issues peculiar to Australian residential aged care facilities that limited adherence with use of hip protectors. In particular, regulations do not support the use of significant numbers of skilled nurses in Australian facilities. However, the evidence from overseas studies also indicates that effectiveness of hip protectors for older people living in nursing care facilities is uncertain and adherence remains a problem.⁶

Back to Article Outline

Conclusion 

This randomized controlled trial has demonstrated that providing hip protectors at no cost was necessary to increase initial acceptance and adherence with their use. However, the achieved level of adherence at 3 and 6 months was modest and not considered high enough to be associated with a reduction in the incidence of hip fractures. The provision of educational sessions and demonstrations showed no additional effect on adherence with wearing hip protectors.

Back to Article Outline

Acknowledgment 

We thank the older people, their families, and the residential aged care facilities that participated in this study.

Back to Article Outline

References 

1. Wortberg WE. Huft-fraktur-bandage zur Verhinderung von Oberschenkelhalsbruch bei alteren Menschen [Hip-fracture-bandage to prevent hip fractures in elderly people.] Zeitschrift fur Gerontologie 1988;21:169–173.
   • View In Article
2. Bentzen H, Bergland A, Forsén L. Risk of hip fractures in soft protected, hard protected, and unprotected falls. Inj Prev. 2008;14:306–310
   • View In Article
   • CrossRef
3. Koike T, Orito Y, Toyoda H, et al. External hip protectors are effective for the elderly with higher-than-average risk factors for hip fractures. Osteoporos Int. 2009;20:1613–1620
   • View In Article
   • CrossRef
4. Kiel DP, Magaziner J, Zimmerman S, et al. Efficacy of a hip protector to prevent hip fracture in nursing home residents. The HIP PRO Randomized Controlled Trial. JAMA. 2007;298:413–422
   • View In Article
5. Parker MJ, Gillespie WJ, Gillespie LD. Hip protectors for preventing hip fractures in older people. Cochrane Database Syst Rev 2005;(3): CD001255. Available at: http://www.mrw.interscience.wiley.com/cochrane/clsysrev/articles/CD001255/frame.html. Accessed March 3, 2010.
   • View In Article
6. Parker MJ, Gillespie WJ, Gillespie LD. Effectiveness of hip protectors for preventing hip fractures in elderly people: Systematic review. BMJ. 2006;11(332):571–574
   • View In Article
7. Van Schoor NM, Deville WL, Bouter LM, Lips P. Acceptance and compliance with external hip protectors: A systematic review of the literature. Osteoporos Int. 2002;13:917–924
   • View In Article
   • CrossRef
8. Kurrle SE, Cameron ID, Quine S. Predictors of adherence with the recommended use of hip protectors. J Gerontol A Biol Sci Med Sci. 2004;59:958–961
   • View In Article
9. Hubacher M, Wettstein A. Acceptance of hip protectors for hip fracture prevention in nursing homes. Osteoporos Int. 2001;12:794–799
   • View In Article
   • CrossRef
10. Cryer C, Knox A, Stevenson E. Factors associated with hip protector adherence among older people in residential care. Inj Prev. 2008;14:24–29
    • View In Article
    • CrossRef
11. Sawka AM, Nixon M, Giangregorio L, et al. The use of hip protectors in long-term care facilities: A survey of nursing home staff. J Am Med Dir Assoc. 2007;8:229–232
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (83 KB)
    • CrossRef
12. Honkanen LA, Monaghan N, Reid MC, et al. Can hip protector use in the nursing home be predicted?. J Am Geriatr Soc. 2007;55:350–356
    • View In Article
    • MEDLINE
    • CrossRef
13. Meyer G, Warnke A, Bender R, Muhlhauser I. Effect on hip fractures of increased use of hip protectors in nursing homes: Cluster randomised controlled trial. BMJ. 2003;326:76–80
    • View In Article
    • CrossRef
14. Cameron ID, Quine S. External hip protectors: Likely non-compliance among high risk elderly people living in the community. Arch Gerontol Geriatr. 1994;19:273–281
    • View In Article
    • MEDLINE
    • CrossRef
15. Holtzman J, Lurie N. Causes of increasing mortality in a nursing home population. J Am Geriatr Soc. 1996;44:258–264
    • View In Article
    • MEDLINE
16. Lord SR, March LM, Cameron ID, et al. Differing risk factors for falls in nursing home and intermediate-care residents who can and cannot stand unaided. J Am Geriatr Soc. 2003;51:1645–1650
    • View In Article
    • MEDLINE
    • CrossRef
17. Granger CV, Albrecht GL, Hamilton BB. Outcome of comprehensive medical rehabilitation: Measurement by PULSES Profile and the Barthel Index. Arch Phys Med Rehabil. 1979;60:145–154
    • View In Article
    • MEDLINE
18. Pfeiffer E. A short portable mental status questionnaire for the assessment of organic brain deficit in elderly patients. J Am Geriatr Soc. 1975;23:433–441
    • View In Article
    • MEDLINE
19. Euroquol The EuroQol Group. EuroQol—a new facility for the measurement of health-related quality of life. Health Policy. 1990;16:199–208
    • View In Article
    • MEDLINE
    • CrossRef
20. Cameron ID, Robinovitch S, Birge S, et al. Hip protectors: Recommendations for conducting clinical trials—an international consensus statement (part II). Osteoporos Int. 2010;21:1–10
    • View In Article
21. Donner A, Klar N. Methods for comparing event rates in intervention studies when the unit of allocation is a cluster. Am J Epidemiol. 1994;140:279–289
    • View In Article
    • MEDLINE
22. Donner A, Birkett N, Buck C. Randomization by cluster. Sample size requirements and analysis. Am J Epidemiol. 1981;114:906–914
    • View In Article
    • MEDLINE
23. O'Halloran PD, Murray LJ, Cran GW, et al. The effect of type of hip protector and resident characteristics on adherence to use of hip protectors in nursing and residential homes—an exploratory study. Int J Nurs Stud. 2005;42:387–397
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (228 KB)
    • CrossRef
24. Burl JB, Centola J, Bonner A, Burque C. Hip protector compliance: A 13-month study on factors and cost in a long-term care facility. J Am Med Dir Assoc. 2003;4:245–250
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (205 KB)
    • CrossRef