Distribution and Characteristics of Fall Risk Index-21 Scores in Patients with Fall-Related Hip Fractures

Abstract

Background

Fall risk assessment and preventive interventions are important for preventing secondary fractures after hip fracture. The aim of this study was to investigate the distribution and characteristics of Fall Risk Index-21 (FRI-21) scores based on prefracture living conditions in patients with fall-related unilateral hip fracture.

Methods

This study included 95 people who had suffered a hip fracture after a fall and had undergone surgery. FRI-21, dual energy X-ray absorptiometry (DXA), and hip structure analysis/ trabecular bone score based on DXA were performed within 1 week after surgery. We compared bone parameters between the two groups, which were divided according to the cutoff value of the FRI-21. We also performed cluster analysis on the components of the FRI-21 and compared the characteristics of the subgroups via analysis of variance and multiple comparison tests.

Results

All participants had experienced a fall-related fracture, but 47 (49.5%) of the patients were judged to have low scores (<10) on the FRI-21. Moreover, cluster analysis revealed that the subjects were classified into three subgroups, and the characteristics of the items that constitute fall risk, namely, physical function, disease or geriatric syndrome, and environmental factors, were diverse.

Conclusions

These results suggest that caution is needed when interpreting the FRI-21 total score based on preoperative living conditions in post-fracture patients, as it may not adequately reflect fall risk. Considering the characteristics of the FRI-21 score in patients with hip fractures, identifying appropriate fall prevention strategies and healthcare professionals involved may be important for preventing future re-fractures.

GRAPHICAL ABSTRACT

INTRODUCTION

Hip fractures due to osteoporosis are associated with a high mortality rate of approximately 30% within a year after fracture.[1] The number of hip fractures worldwide is projected to reach 6.3 million by 2050.[2] Importantly, one in three patients who experience a first hip fracture will suffer a second,[3] most often occurring soon after the initial event.[4] Therefore, refracture prevention programs should be initiated promptly following the first hip fracture.

The occurrence of a fracture is related to bone strength and the magnitude of the external force. Therefore, to prevent fractures, two approaches are necessary: improving bone strength and preventing falls.[5] Currently, a variety of tests exist to assess the risk of falls in elderly individuals. The Fall Risk Index-21 (FRI-21) is a questionnaire that can be easily used to assess fall risk [6]; it is recommended by the Japanese Fracture Liaison Service (FLS) clinical standards. In Japan, medical fee revisions aimed at preventing secondary fractures following hip fractures were implemented in 2022, and the FRI-21 is now being used in many healthcare facilities. The FRI-21 consists of 21 items related to physical function, disease or geriatric syndrome, and environmental factors, with a total score of 10 or more indicating a high risk of falling.[7] This questionnaire has been reported to be useful for predicting future falls [6] and is associated with physical function tests such as the timed up and go (TUG) test, the one-legged standing assessment, and the functional reach test.[8] In addition, unlike physical function tests, which are used to evaluate balance ability related to falls, the FRI-21 may be particularly beneficial for postoperative patients in that the evaluation results are not affected by post fracture pain.

In general, fall assessments such as the FRI-21 focus on identifying individuals at high risk of falling and who are more likely to benefit from interventions, and it is typically used to screen individuals in the general population. The Japanese FLS clinical standards for the prevention of secondary fractures state that a fall and fracture prevention program should be considered if the FRI-21 score is 10 or higher; however, almost the target population in the secondary fracture prevention program has already experienced at least a fall-related fracture. There is limited evidence on the distribution of the FRI-21 score in patients with preexisting fractures, and it is unclear how meaningful it is to use the FRI-21 score to determine whether to intervene to prevent falls in patients with fractures.

Most individuals who suffer a hip fracture as a result of a fall need to be provided with appropriate fall prevention program instructions or interventions. However, there is no evidence regarding the distribution and characteristics of FRI-21 scores based on pre-fracture living conditions in patients with fall-related fractures, and it is unclear whether patients with unilateral hip fractures have higher scores. Therefore, the primary aim of this study was to evaluate the distribution and characteristics of FRI-21 scores based on pre-fracture living conditions in patients with fall-related unilateral hip fracture. A secondary objective of the study was to examine fall prevention strategies on the basis of the characteristics identified by the FRI-21. Clarification of these findings might provide strategies to prevent secondary fractures following a hip fracture.

METHODS

1. Subjects

The study subjects were 95 patients (80 women, 81.1 ± 10.1 years) who underwent hip fracture surgery at our hospital between May 2022 and March 2023 and did not meet the exclusion criteria. The exclusion criteria were as follows. Injury from high-energy trauma, a history of contralateral proximal femur fracture surgery, no bone mineral density (BMD) examination, and those who scored less than 20 on the Mini-Mental State Examination and had difficulty answering the FRI-21 were excluded. In our hospital, FRI-21, blood tests, BMD tests, and osteoporosis treatment after hip fracture, as described below, are routinely performed.

This study was approved by the Ethics Committee of our hospital (Approval No. 2022-2). The study was a retrospective, cross-sectional study; the subjects were informed and provided consent via the opt-out method.

2. FRI-21

Within one week after surgery, a physical therapist performed an FRI-21 assessment on the basis of the patient’s prehospitalization living conditions. The questions in the FRI-21 are shown in the Supplementary Appendix 1. We classified the subjects into a high-score group (≥10 points) and a low-score group (<10 points) on the basis of the cut-off score of the FRI-21.[7]

3. Blood tests and bone metabolism markers

In this study, we assessed the levels of 25-hydroxy-vitamin D, which has been reported to be associated with falls. [9] The level of procollagen type N-terminal propeptide of type I collagen (P1NP; via electrochemiluminescence immunoassay; BML, Inc., Tokyo, Japan) was evaluated as an osteogenic marker. The level of tartrate-resistant acid phosphatase 5b (TRACP-5b; BML, Inc.) was evaluated as a bone resorption marker. P1NP and TRACP-5b levels were measured by BML, Inc. Blood samples were collected at the time of admission.

4. Bone-related parameters

Dual energy X-ray absorptiometry (DXA; Hologic Inc., Bedford, MA, USA) was used to evaluate the BMD and T-scores of the proximal femur on the nonfractured side (femoral neck and total hip) and the lumbar spine (L1–L4). For patients with preexisting vertebral fractures, the BMD and T-score of the vertebrae excluding the preexisting fractured vertebra was evaluated. DXA was performed within 1 week after surgery.

Additionally, we used hip structure analysis (HSA) to subperiosteal width, estimated endosteal width, cortical bone thickness, calculate cross-sectional area, cross-sectional moment of inertia, section modulus, and buckling ratio. HSA parameters were calculated separately for the narrow neck and intertrochanteric regions.

In addition, we calculated the trabecular bone score (TBS) for L1 to L4. As with the T-score assessment, the TBS was evaluated after excluding preexisting vertebral fractures.

5. Statistical analysis

First, a Shapiro–Wilk test was performed to check the normality of the data. Patients were then divided into two groups on the basis of the cutoff value of the FRI-21: high score (FRI-21 score; ≥10) and low score (FRI-21 score; <10), and demographic data, blood data, and bone-related parameters were compared via two-sample t-tests or Mann–Whitney tests. In addition, χ² tests were performed to examine the distributions of FRI-21 scores and femoral neck T-scores.

Next, hierarchical cluster analysis was performed via Ward’s method, with the scores of the three items that make up the FRI-21 (physical function; Q1–Q8, disease or geriatric syndrome; Q9–Q16, environmental factors; and Q17–Q21) as variables. Dissimilarity between subjects was calculated via the Euclidean distance. To clarify the characteristics of the subgroups, one-way analysis of variance (ANOVA) was subsequently performed for each parameter in each group, followed by multiple comparison tests.

Finally, the distribution of responses to each FRI-21 item among the subgroups identified through cluster analysis was examined via the χ² test.

All significance levels were set at P value less than 0.05. R (version 4.2.1; The R Foundation for Statistical Computing, Vienna, Austria; http://www.R-project.org/) was used for statistical analysis.

RESULTS

1. Characteristics of all the subjects

The characteristics of the subjects included in this study are shown in Table 1. All 95 subjects had suffered hip fractures due to falls. The subjects’ FRI-21 score was 9.4±4.1 points (range, 0–18 points). There were 48 participants (41 women) with FRI-21 scores of 10 or more, which is considered to indicate a high risk of falling. The Shapiro–Wilk test indicated that the FRI-21 scores of the subjects were normally distributed (P=0.114).

Table 1.

Subject characteristics

Variable	Value (N=95)	Range
Age (yr)	81.1±10.1	50–98
Height (cm)	152.0±8.7	138.7–184.0
Weight (kg)	50.16±10.2	28.5–76.6
BMI (kg/m2)	21.8±4.3	12.2–35.4
TRACP-5b (mU/dL)	415.6±158.9	111–775
P1NP (μg/L)	52.6±27.6	14.5–135.8
25(OH)D (ng/mL)	15.7±5.5	5.2–32.5
FRI-21 score	9.4±4.1	0–18
FRI-21 score
≥10	48 (50.5)
<10	47 (49.5)
T-score (SD)
Femoral neck	−3.3±1.1	−5.2–0.7
Total hip	−2.8±1.1	−4.8–1.0
Lumbar	−1.8±1.6	−4.5–2.7
BMD (g/cm2)
Femoral neck	0.49±0.10	0.33–0.85
Total hip	0.60±0.11	0.40–0.98
Lumbar	0.79±0.18	0.49–1.30
Walking condition before injury
Walking alone	56 (58.9)
Use of T-cane	25 (26.3)
Use of a walking vehicle	14 (14.7)
Types of hip fractures
Femoral neck	62 (65.3)
Intertrochanteric	33 (34.7)
The surgical procedure performed
Total hip arthroplasty	2 (2.1)
Bipolar hip arthroplasty	49 (51.6)
Intramedullary nail	33 (34.7)
Femoral neck system	11 (11.6)
Personal history
Vertebral fracture
1	12 (12.6)
≥2	4 (4.2)
Diabetes	19 (20.0)
Stroke	9 (9.5)
Rheumatoid arthritis	2 (2.1)
Parkinson’s disease	1 (1.1)
Hypothyroidism	3 (3.2)
Previous surgery for any cancer	14 (14.7)

The data is presented as mean±standard deviation (SD) or N (%).

BMI, body mass index; TRACP-5b, tartrate-resistant acid phosphatase 5b; P1NP, procollagen type N-terminal propeptide of type I collagen; 25(OH)D, 25-hydroxy-vitamin D; FRI-21, Fall Risk Index-21; BMD, bone mineral density.

Responses to each question on the FRI-21 are shown in Figure 1. All the subjects were able to answer all 21 questions. Questions Q2 (Can you climb stairs without help?), and Q3 (Do you feel your walking speed has declined recently?), Q14 (Do you think you are forgetful?), and Q19 (Is there any difference in level within your home?) were scored by more than 60% of the subjects.

Fig. 1.

Response status for each question on the Fall Risk Index-21. The black bars indicate the number of people who answered “Yes,” and the gray bars indicate the number of people who answered “No.” For some questions (Q2, Q4, Q5, Q6, and Q8), a score of 1 was awarded if the answer was “No.”

2. Comparison between the high- and low-score groups

The results of the two-sample t-test, the Mann–Whitney test and the χ² test are shown in Table 2. On the basis of the FRI-21 score, 48 participants (41 women) with a score of 10 or more were classified into the high score group, and 47 participants (39 women) with a score of less than 10 were classified into the low score group. The high score group was significantly older (P<0.001) and shorter (P=0.048) than the low score group was. Additionally, the high score group had higher lumbar spine T-scores (P=0.013). However, no significant differences were observed in the femoral neck or total hip T-scores.

Table 2.

Comparison of two groups (high-score group vs. low-score group)

	FRI-21 score ≥10 (N=48)	FRI-21 score <10 (N=47)	P-value
Female	41 (85.4)	39 (83.0)
FRI-21 score	12 (10 to 14)	6 (4 to 8)	<0.001
Age (yr)	85 (69 to 90)	79 (72 to 85)	<0.001
Height (cm)	149.25 (138.70 to 153.25)	152.00 (147.00 to 157.90)	0.048
Weight (kg)	50.45±13.14	49.89±9.19	0.784
BMI (kg/m2)	22.35±5.75	21.14±3.71	0.176
25(OH)D (ng/mL)	15.73±5.71	15.61±5.75	0.920
Bone turnover marker
TRACP-5b	417.58±178.92	413.60±148.07	0.903
P1NP	46.95 (14.50 to 65.78)	47.90 (35.85 to 64.95)	0.433
T-score
Femoral neck	−3.50 (−5.20 to −2.85)	−3.40 (−4.15 to −2.75)	0.639
Total hip	−2.80±1.12	−2.77±1.01	0.869
Lumbar	−1.41±1.75	−2.21±1.34	0.013
BMD (g/cm2)
Femoral neck	0.48 (0.42 to 0.53)	0.48 (0.42 to 0.54)	0.634
Total hip	0.60±0.11	0.60±0.10	0.977
Lumbar	0.79 (0.69 to 0.95)	0.74 (0.64 to 0.84)	0.018
Femoral neck T-score			0.960
≤ −2.5 SD	39 (81.3)	38 (80.9)
> −2.5 SD	9 (18.8)	9 (19.1)
Past history of vertebral fracture	10 (20.8)	6 (12.8)	0.294

Two-sample t-test results are listed as the mean±standard deviation (SD), and Mann-Whitney test results are shown as the median (interquartile range). The cross-tabulation results of the Fall Risk Index-21 (FRI-21) score, T-score (femoral neck), and history of vertebral fracture are also shown. P<0.05 is statically significant. Bold values indicate statistical significance.

BMI, body mass index; 25(OH)D, 25-hydroxy-vitamin D; TRACP-5b, tartrate-resistant acid phosphatase 5b; P1NP, procollagen type N-terminal propeptide of type I collagen; BMD, bone mineral density.

A χ² test was performed on the basis of the distribution of femoral neck T-scores and FRI-21 scores. There were 39 patients (81.3%) with FRI-21 scores of 10 or more and femoral neck T-scores of −2.5 standard deviation or less, but no statistically significant differences were found (χ² = 0.002, P=0.960).

Furthermore, there were no significant differences in any of the HSA and TBS parameters between the two groups (Supplementary Table 1).

3. Characteristics of the subgroups classified via hierarchical cluster analysis

To investigate the characteristics of FRI-21 in more detail, we performed hierarchical cluster analysis. As a result, it was possible to classify the 95 people into three groups (Group A [N=38], Group B [N=28], and Group C [N=29]) (Table 3, Fig. 2).

Table 3.

Comparison between subgroups calculated via hierarchical cluster analysis

Variables	Mean±standard deviation			ANOVA P-value	P-value for multiple
	Group A (N=38)	Group B (N=28)	Group C (N=29)		A vs. B	A vs. C	B vs. C
FRI-21	5.74±2.87	14.00±1.91	9.62±1.86	<0.001	<0.001	<0.001	<0.001
Age (yr)	76.53±12.10	85.00±6.50	83.31±7.69	0.003	0.006	0.034	0.807
Height (cm)	154.33±9.90	148.75±6.74	152.08±7.96	0.043	0.035	0.657	0.271
Weight (kg)	49.52±9.04	50.21±12.91	50.97±8.72	0.806	0.968	0.786	0.964
BMI	20.83±3.54	22.71±5.76	22.05±3.47	0.204	0.287	0.341	0.859
TRACP-5b	425.39±139.89	409.07±149.51	409.10±192.83	0.667	0.912	0.911	1.000
P1NP	54.13±25.43	52.86±28.73	50.18±30.05	0.583	0.954	0.537	0.817
25(OH)D	15.15±5.30	15.00±4.29	16.99±6.63	0.982	0.993	0.369	0.364
DXA
T-score
Neck	−3.31±1.2	−3.54±1.05	−3.17±1.06	0.351	0.684	0.720	0.338
Total hip	−2.79±1.08	−2.85±1.21	−2.72±0.91	0.856	0.974	0.957	0.887
Lumbar	−2.29±1.29	−1.4±1.74	−1.55±1.72	0.020	0.064	0.139	0.932
BMD
Neck	0.49±0.11	0.47±0.10	0.50±0.10	0.371	0.713	0.754	0.338
Total hip	0.60±0.11	0.60±0.12	0.60±0.09	0.982	0.999	0.960	0.964
Lumbar	0.73±0.15	0.83±0.20	0.81±0.19	0.022	0.067	0.165	0.908
Past history of vertebral fracture (N)	4	6	6	0.405

The results of one-way analysis of variance (ANOVA) and multiple comparison tests are shown. The results of the χ² test are shown for patients with a history of vertebral fracture. P<0.05 is statically significant. Bold values indicate statistical significance.

FRI-21, Fall Risk Index-21; BMI, body mass index; TRACP-5b, tartrate-resistant acid phosphatase 5b; P1NP, procollagen type N-terminal propeptide of type I collagen; 25(OH)D, 25-hydroxy-vitamin D; DXA, dual energy X-ray absorptiometry; BMD, bone mineral density.

Fig. 2.

Comparison of scores on the three items that make up the Fall Risk Index-21 (FRI-21) between subgroups. The vertical axis in all the tables shows the scores for each item that makes up the FRI-21. ^a)P<0.05, ^b)P<0.01.

Table 3 and Figure 2 show the results of ANOVA and multiple comparison tests between groups calculated via cluster analysis. No differences in HSA and TBS parameters were observed among the three groups, and the results are shown in Supplementary Table 2. Compared with Groups B and C, Group A had significantly lower scores for the physical function, disease or geriatric syndrome aspects of the FRI-21. Group B had significantly higher physical function scores than Group A did, and the disease or geriatric syndrome scores were higher than those of Groups A and C. Group B also had significantly higher environmental factor scores than Group C did. On the other hand, Group C had higher physical function scores than Group A did, but the scores for disease or geriatric syndrome and environmental factors were significantly lower than those of Group B were.

4. Distribution of scores for each FRI-21 item among subgroups identified through cluster analysis

The distribution of scores for each item of the FRI-21 across the subgroups is shown in Table 4. The χ² test results revealed significant differences between subgroups in all the questions except for some questions about the environment (Q18, Q20, and Q21).

Table 4.

Distribution of scores for each Fall Risk Index-21 item among subgroups identified through cluster analysis

	FRI-21
	Physical function										Disease/Geriatric syndrome						Environmental
	Q1	Q2	Q3	Q4	Q5	Q6	Q7	Q8	Q9	Q10	Q11	Q12	Q13	Q14	Q15	Q16	Q17	Q18	Q19	Q20	Q21
Group A
Yes (N)	12	12	18	1	2	9	3	0	6	10	11	17	9	18	13	9	2	9	26	21	10
No (N)	26	26	20	37	36	29	35	38	32	28	27	21	29	20	25	29	36	29	12	17	28
Group B
Yes (N)	20	26	28	11	22	22	17	11	20	26	18	21	19	24	26	22	6	10	25	10	8
No (N)	8	2	0	17	6	6	11	17	8	2	10	7	9	4	2	6	22	18	3	18	20
Group C
Yes (N)	21	26	26	12	15	20	18	10	5	17	7	14	9	16	12	11	1	6	16	9	8
No (N)	8	3	3	17	14	9	11	19	24	12	22	15	20	13	17	18	28	23	13	20	21
P-value	<0.001	<0.001	<0.001	<0.001	<0.001	<0.001	<0.001	<0.001	<0.001	<0.001	0.003	0.036	<0.001	0.005	<0.001	<0.001	0.035	0.390	0.017	0.099	0.980

P<0.05 is statically significant. Bold values indicate statistical significance.

FRI-21, Fall Risk Index-21.

DISCUSSION

We investigated the distribution and characteristics of FRI-21 scores assessed based on pre-fracture lifestyle in patients with fall-related unilateral hip fracture. Our study revealed two novel findings. The first was the distribution of FRI-21 scores among hip fracture patients who sustained injuries due to falls. All participants had experienced a fall-related fracture, but 47 (49.5%) patients had FRI-21 scores below the cutoff (10 points). Although the FRI-21 is a useful tool for assessing fall risk, caution is warranted when relying solely on the total FRI-21 score in this population, as it may not adequately reflect the actual risk of falling. The second finding was the identification of distinct subgroups on the basis of the characteristics of the FRI-21 scores. The results of our cluster analysis suggest that tailored multidisciplinary interventions may be necessary and effective depending on the underlying causes of fall risk.

This study is the first to report the characteristics of the FRI-21 score based on preoperative living conditions in hip fracture patients. Falls are common among elderly individuals, with one in three adults aged 65 years or older experiencing a fall each year.[10] Risk factors that lead to falls in elderly people have also been identified and can be broadly categorized as either intrinsic (e.g., older age, history of falls, and muscle weakness) or extrinsic (e.g., lack of handrails, lack of lighting) factors.[10,11] The FRI-21 consists of 21 questions assessing three components, namely, physical function, disease or geriatric syndrome, and environmental factors, allowing the investigation of multiple factors related to falls. Furthermore, because the FRI-21 is a questionnaire, a major advantage is that the results are not affected by postoperative pain.

The Japanese FLS clinical standards for the prevention of secondary fractures state that a fall and fracture prevention program should be considered if the FRI-21 score is 10 or higher. In addition to the FRI-21 used in the present study, various methods, such as the Berg balance scale (BBS) and the Downton FRI, have been used for fall risk assessment,[ 12] and some studies have shown that fall risk assessment can be used to predict future falls and hip fractures in older adults.[13,14] As those fall risk assessments are screening tools for the general population, they may contribute to the prevention of primary fracture by identifying people at high risk of falling from the general population. Thus, considering the results of our study and the fact that the positive predictive value of the FRI-21 was approximately 44%,[7] the total FRI-21 score is not appropriate for use with populations who have fallen and experienced fractures, and caution is warranted when the total FRI-21 score is used as a criterion for intervention in patients with fall-related fractures with the aim of preventing secondary fractures.

All the subjects in this study had suffered hip fractures as a result of falls. Primary hip fracture is a significant predictor of contralateral secondary hip fracture, and fall prevention interventions for patients after hip fracture surgery are crucial. A recent meta-analysis revealed that the main risk factors for future falls were a history of falls and the presence or absence of comorbid conditions.[15] Additionally, a study of elderly people in Japan reported that the most influential predictor of future falls was a history of previous falls.[16] Therefore, based on the results of this study, even if the FRI-21 score based on preoperative living conditions is low (less than 10 points), it is important to take into account the fact that the patient has already experienced a fall and to provide fall prevention intervention and guidance.

Additionally, we performed cluster analysis using the scores of the three components of the FRI-21 (physical function, disease or geriatric syndrome, and environmental factors) as variables to classify the study participants into three subgroups (Table 3, Fig. 2). In general, physical therapists often play an important role in preventing falls. However, our cluster analysis and ANOVA results suggest that post-operative fall prevention requires comprehensive support from multiple professions, not just physical therapists. For example, for relatively young people with a low risk of falling, such as Group A, a realistic response might be to deepen their understanding of falls and provide them with appropriate exercise instructions at home. For groups such as Group C, which had high scores in the physical function categories, interventions by physical therapists aimed at improving physical function appeared to be particularly important in preventing falls. Furthermore, our χ² test revealed that more than 70% of individuals in Group C were marked ‘Yes’ on two FRI-21 items: Q1 (history of tripping) and Q3 (perceived decline in walking speed) (Table 4). Given the nature of these questions, our findings suggest a clearer need for fall prevention interventions targeting the lower limbs and trunk. On the other hand, Group B had high values for all the items related to physical function, disease or geriatric syndrome, and environmental factors. Fall prevention interventions for Group B suggest the need for support not only from physical therapists but also from medical staff, such as doctors and nurses, in collaboration. In other words, when considering the prevention of refractures, it is important to consider specific fall prevention intervention methods and the medical staff that should be involved, depending on the characteristics and distribution of the subjects’ FRI-21 scores.

This study has several important limitations. The most significant are recall bias, as the FRI-21 score was assessed after surgery based on the patients’ preoperative living situation, and the relatively small sample size (N=95), which may have reduced statistical power and limited the ability to detect significant differences, with some analyses showing only trends. Additionally, although we used the FRI-21 to assess fall risk, physical function was not evaluated using objective measures such as the BBS or TUG, making it difficult to fully characterize the participants’ functional status. Furthermore, age differences between groups may have influenced both the results and the interpretation of FRI-21 scores. Finally, because this was a cross-sectional study, longitudinal research is required to further evaluate the clinical utility of the FRI-21 in preventing refractures among patients with hip fractures.

Our study revealed important insights into the distribution and characteristics of FRI-21 scores assessed based on pre-fracture living conditions in patients with fall-related unilateral hip fracture. Notably, although all participants had experienced a fall-related fracture, approximately half were classified as having a low FRI-21 score (low fall risk). Although the FRI-21 is a useful tool for assessing fall risk, our findings highlight the need for caution when interpreting the total FRI-21 score alone in this population, as it may not fully capture the actual risk of falling. Furthermore, cluster analysis identified distinct subgroups according to the characteristics of the FRI-21 scores, suggesting that tailored, multidisciplinary prevention strategies may be necessary and effective. Overall, our results emphasize the clinical importance of developing individualized approaches that involve multiple healthcare professionals to prevent refractures in hip fracture patients.