Abstract
Background
Falls are a public health concern among older adults, particularly in nursing home (NH) residents, but in Italy data on fall incidence and risk factors remain limited.
Aims
To estimate the incidence of falls, identify associated risk factors, and evaluate the predictive value of the Tinetti Performance-Oriented Mobility Assessment (T-POMA) in a large cohort of NH residents.
Methods
We conducted a retrospective cohort study using electronic health records from 32 NHs managed by Kos Care company across Italy. Residents aged ≥ 65 years with at least one functional assessment were included. Fall incidence was calculated per 100 person-years. Cox proportional hazards regression models were used to identify predictors of time to first fall.
Results
Overall, 754 residents (19.2%) experienced a fall. A T-POMA score ≤ 18 was a strong predictor of falls (HR 2.13; 95% CI: 1.61–2.81), along with age ≥ 85 years (HR 1.39; 1.02–1.90), male sex (HR 1.47; 1.27–1.71), cognitive impairment (HR 1.30; 1.10–1.53), hearing impairment (HR 1.26; 1.01–1.58), mood and behavioural disorders (HR 1.29; 1.09–1.54), and therapies known to increase fall risk (HR 1.29; 1.09–1.52). Poor general health (HR 0.63; 0.52–0.77) and frequent physical restraint (HR 0.70; 0.58–0.86) use were associated with lower fall risk. Severe fall consequences occurred in 15.9% of fallers and were significantly associated with low T-POMA scores and restraint use.
Discussion and conclusion
Falls are highly prevalent among Italian NH residents and influenced by multiple clinical and functional factors. These findings support the implementation of multifactorial fall prevention strategies and real-time risk monitoring in NH settings.
Supplementary Information
The online version contains supplementary material available at 10.1007/s40520-025-03111-7.
Keywords: Falls, Nursing homes, Older adults, Tinetti score, Functional assessment, Fall prevention
Introduction
As the population ages, falls among older adults are becoming an increasingly common occurrence and a significant public health challenge, with around one in four adults aged 65 years and older reporting a fall each year—although variability exists across different settings [1]. Falls are also a major cause of injury, with up to 30% of older adults who experience a fall sustaining moderate to severe consequences—such as fractures and head trauma—which can lead to hospitalization, disability, and, in the most severe cases, death [2]. Beyond the physical consequences, falls also have a significant psychological impact. They can lead to a loss of confidence, post-fall anxiety, and fear of falling, which may exacerbate activity limitations and functional decline, increase the risk of future falls, and further impair the quality of life in older adults [3].
The NICE quality standard (UK National Institute for Health and Care Excellence) highlights that there are over 400 risk factors associated with falls, with the likelihood of falling increasing as the number of these risk factors rises [4]. Among older adults residing in care facilities, such as nursing homes (NHs), the fall rate exceeds 2.5 falls per person per year, which is approximately three times higher than that observed among community-dwelling older adults [5– 7]. According to the US Centers for Disease Control and Prevention, up to one in five NH falls result in serious injuries, and approximately 1,800 NH residents die each year due to fall-related consequences [7].
The increased fall risk in NHs can be attributed to multiple factors, including residents’ advanced age, underlying health conditions and multimorbidity (e.g., moderate-to-high disability, neurological diseases, cognitive decline), polypharmacy, limited self-care abilities, use of walking aids, and weakened social support [6, 8]. Furthermore, falling is an independent predictor of future falls [9].
International guidelines emphasize the importance of using standardized tools to quantify fall risk through gait and balance assessments [4, 10], and the Tinetti Performance-Oriented Mobility Assessment (T-POMA) is widely recognized for evaluating balance and gait abnormalities, as well as for predicting fall risk [11, 12–15]. To consolidate the evidence on the predictive validity of the T-POMA in high-risk populations, particularly NH residents, previous studies have highlighted the need for large prospective investigations [13, 16]. Furthermore, comprehensive analyses of NH residents with a history of falls and associated risk factors are essential—particularly in countries such as Italy, where data on falls in long-term care settings remain limited and there is a pressing need to advance the current epidemiological understanding.
Building on this premise, we aimed to estimate the incidence and risk factors of falls in a large population of Italian NH residents, and to evaluate the added value of the T-POMA score in predicting fall risk in this population.
Methods
Study design
We conducted a retrospective cohort study using the databases of Kos Care company, an Italian healthcare group that provides socio-health services, rehabilitation, psychiatry, diagnostics, and acute care through a network of NHs across eight regions in Italy. These facilities are characterized by harmonized clinical-care processes, supported by standardized quality and safety protocols, and a unified electronic health record (EHR) system. The NHs deliver a range of services to older adults with varying levels of functional autonomy and cognitive capacity, offering both short-term and long-term (permanent) care. Ethical approval for the study was obtained from the Ethics Committee of the University of Milan–Bicocca (protocol no. 928/2025), and all data were analysed anonymously.
Study population
At the time of the study, the Kos Care databases included data from 4,745 residents across 32 NHs in Italy. In these facilities, residents underwent functional assessments (FAs), which evaluate an individual’s level of functioning and ability to perform daily tasks. The assessment includes mobility, basic activities of daily living (ADLs)—defined as essential tasks for survival, hygiene, and self-care—and participation in recreational activities. We selected a dynamic cohort of NH residents according to the following criteria: (1) aged 65 years or older; and (2) had at least one comprehensive clinical evaluation recorded at the time of admission or during residency, including physical examination, medical history, and FA. Residents whose FA was recorded after the occurrence of falls and those with a recorded T-POMA score of 0 were excluded from the analysis. For all participants, the index date for the analyses was set as January 1, 2022, and follow-up time was censored at the date of fall occurrence, exit from the cohort, or at the end of the follow-up period on December 31, 2022—whichever occurred first.
Study endpoints
According to the literature, a fall is defined as “an unexpected event in which an individual comes to rest on the ground, floor, or a lower level” [17]. Fall-related data were collected from daily logs or incident reports maintained by staff, as well as from EHR reviews. Documentation included details on the circumstances surrounding the fall, including location, time, contributing factors (e.g., clinical conditions, medication use), as well as the resident’s post-fall condition, such as the level of consciousness and use of mobility aids or restraints. The consequences of falls were also recorded in the Kos database and categorized as either severe (e.g., head trauma, hospitalization, fractures, surgery, or referral to the emergency department) or mild (e.g., bruises, abrasions, lacerations).
Study variables
Multiple variables were collected and categorized into key groups to assess fall risk and those measured in FAs were dichotomized from their original four-level categorical distribution for analysis. Demographic characteristics included sex and age, categorized into three groups (65–74, 75–84, and 85 + years). Balance and gait assessment was conducted using the T-POMA, which evaluates both balance (maximum score: 16) and gait (maximum score: 12), with a total possible score of 28. The balance component included tasks such as sitting balance, rising from a chair, and turning, while the gait component assessed step length, height, and symmetry. Scores were interpreted as follows: ≤18 indicated a high fall risk, 19–24 a moderate risk, and ≥ 25 a low fall risk. To assess the effect of long-term residency on fall risk and related outcomes, length of stay was categorized as ≤ 2 years or > 2 years. Physical restraint use was recorded based on frequency: less than once per week or one or more times per week/daily. The database includes a global evaluation of general health conditions, performed by NH physicians, which results in a classification of ‘good’, ‘discrete’, or ‘poor’ based on clinical judgment. The presence of moderate or severe visual and hearing impairments, was assessed using standardized clinical evaluations [18]. Cognitive function was assessed using the Mini-Mental State Examination (MMSE), with a cut-off score of 18 to distinguish no or mild cognitive impairment from moderate or severe cognitive impairment [19]. The functional status was evaluated using the Barthel Index [20], dichotomized as ≥ 75 (autonomous / mild dependence) and < 75 (moderate / severe dependence). Mood and behavioural disorders were evaluated by NH physicians, and categories were constructed as follows: the none/mild group included individuals with no symptoms or only occasional, minor issues (e.g., mild anxiety, irritability); the moderate/severe group included individuals with more severe symptoms, such as major mood (e.g., depression, mania) or marked behavioural disturbances (e.g., aggression, impulsivity, self-harm), which interfered with daily functioning and often required professional intervention. Comorbidities and therapies influencing fall risk were also considered (listed in Table S1). Comorbidities were classified as: i) absent or not relevant, i.e., no active medical conditions were present, or existing conditions did not contribute to fall risk; ii) relevant, i.e., at least one active medical condition potentially influencing fall risk was present (e.g., a condition causing postural instability, gait disturbances, or one that had previously led to falls). For medications, the risk categories were defined as follows: i) absent, i.e.) fewer than three medications, none associated with increased fall risk; ii) mild, i.e., fewer than five medications, none associated with increased fall risk; iii) moderate, i.e., fewer than three medications, none associated with increased fall risk, or more than five medications not associated with fall risk; iv) severe, i.e., more than three medications known to increase fall risk. Additionally, fall history was documented, specifically recording whether participants had experienced one or more falls in the previous 180 days.
Statistical analysis
Baseline characteristics were summarized as frequencies and percentages, and median and interquartile range (IQR). To evaluate fall incidence, we estimated the crude yearly incidence rate of first falls, expressed as the number of new fall events per person-year. To account for differences in fall risk stratification, we estimated the incidence of falls across three T-POMA categories: high fall risk (≤ 18), moderate fall risk (19–24), and low fall risk (≥ 25). Subsequently, we employed Cox proportional hazards regression models, both in univariable and multivariable analyses, to investigate the association between demographic and clinical factors and the time to first fall. Hazard ratios (HRs) with 95% confidence intervals (95% CIs) were reported to quantify the risk of experiencing a fall. The proportional hazards assumption was evaluated by means of martingale residuals and hypothesis testing [21]. Finally, we built a multivariable Poisson regression model to compute incidence rate ratios (IRRs) and 95% CIs of potential risk factors that predicted with the occurrence of severe fall consequences. All statistical analyses were performed using R 4.4.1 (R Foundation for Statistical Computing, Vienna, Austria) and SAS 9.4 (SAS Institute, Cary, North Carolina, USA). Significance level was set at p-value ≤ 0.05.
Results
A total of 3,921 individuals were included in the analysis according to the inclusion and exclusion criteria (Fig. 1), of whom 69.5% were female and 66.3% were aged 85 years or older. Of the total cohort, 754 residents experienced falls (19.2%). Table 1 presents the demographic and clinical characteristics of study participants, stratified by fallers and non-fallers. No differences in age distribution or median length of stay were observed between fallers and non-fallers. Fallers had a higher proportion of individuals with a T-POMA score ≤ 18 (30.5% vs. 17.5%) and a lower proportion of residents having poor general health conditions (21.8% vs. 36.7%). Frequent or daily use of physical restraints was less common among fallers compared to non-fallers (37.7% vs. 55.5%).
Fig. 1.
Flowchart of the cohort assembly. Abbreviations: T-POMA, Tinetti Performance-Oriented Mobility Assessment
Table 1.
Demographic and clinical characteristics of study participants, overall and stratified by fall status
| Residents’ characteristics | Total (N = 3921) | Fallers (N = 754) | Non-fallers (N = 3167) |
|---|---|---|---|
| Age class | |||
| 65–74 | 246 (6.3%) | 45 (6%) | 201 (6.3%) |
| 75–84 | 1074 (27.4%) | 204 (27.1%) | 870 (27.5%) |
| 85+ | 2601 (66.3%) | 505 (67%) | 2096 (66.2%) |
| Sex | |||
| Females | 2726 (69.5%) | 470 (62.3%) | 2256 (71.2%) |
| Time to first fall / censoring | |||
| Median (IQR) days | 365.0 (365.0–365.0) | 209.0 (132.0–277.8) | 365.0 (365.0–365.0) |
| Length of stay | |||
| Median (IQR) | 0.6 (0.4–1.1) | 0.7 (0.4–1.3) | 0.6 (0.3–1.1) |
| 0–2 years | 3332 (85%) | 645 (85.5%) | 2687 (84.8%) |
| > 2 years | 589 (15%) | 109 (14.5%) | 480 (15.2%) |
| T-POMA score | |||
| ≤18 | 784 (20%) | 230 (30.5%) | 554 (17.5%) |
| 19–24 | 787 (20.1%) | 190 (25.2%) | 597 (18.9%) |
| 25+ | 2350 (59.9%) | 334 (44.3%) | 2016 (63.7%) |
| Physical restraint | |||
| No / Sporadic | 1879 (47.9%) | 470 (62.3%) | 1409 (44.5%) |
| Frequent / Daily | 2042 (52.1%) | 284 (37.7%) | 1758 (55.5%) |
| General health conditions | |||
| Good / Discrete | 2596 (66.2%) | 590 (78.2%) | 2006 (63.3%) |
| Poor | 1325 (33.8%) | 164 (21.8%) | 1161 (36.7%) |
| Visual impairment | |||
| No / Mild | 3538 (90.2%) | 697 (92.4%) | 2841 (89.7%) |
| Moderate / Severe | 383 (9.8%) | 57 (7.6%) | 326 (10.3%) |
| Hearing impairment | |||
| No / Mild | 3465 (88.4%) | 660 (87.5%) | 2805 (88.6%) |
| Moderate / Severe | 456 (11.6%) | 94 (12.5%) | 362 (11.4%) |
| Cognitive impairment (MMSE) | |||
| 18+ (No / Mild) | 1772 (45.2%) | 353 (46.8%) | 1419 (44.8%) |
| < 18 (Moderate / Severe) | 2149 (54.8%) | 401 (53.2%) | 1748 (55.2%) |
| Functional assessment (Barthel Index) | |||
| 75+ (Autonomous / Mild Dependence) | 402 (10.3%) | 118 (15.6%) | 284 (9%) |
| < 75 (Moderate / Severe Dependence) | 3519 (89.7%) | 636 (84.4%) | 2883 (91%) |
| Mood and behaviour disorders | |||
| No / Mild | 2842 (72.5%) | 520 (69%) | 2322 (73.3%) |
| Moderate / Severe | 1079 (27.5%) | 234 (31%) | 845 (26.7%) |
| Comorbidities influencing the risk of falls | |||
| No / Not Relevant | 964 (24.6%) | 206 (27.3%) | 758 (23.9%) |
| Relevant | 2957 (75.4%) | 548 (72.7%) | 2409 (76.1%) |
| Therapies influencing the risk of falls | |||
| No / Mild Risk | 1570 (40%) | 286 (37.9%) | 1284 (40.5%) |
| Moderate / Severe Risk | 2351 (60%) | 468 (62.1%) | 1883 (59.5%) |
| Previous falls within the last 180 days | |||
| No | 3365 (85.8%) | 643 (85.3%) | 2722 (85.9%) |
| 1+ | 556 (14.2%) | 111 (14.7%) | 445 (14.1%) |
Abbreviations: IQR, interquartile range; T-POMA, Tinetti Performance-Oriented Mobility Assessment; MMSE, Mini-Mental State Examination
Males experienced falls at an incidence rate of 26.6 per 100 person-years (95% CI: 23.6–29.9), which was higher than that observed in females (18.6/100py; 95% CI: 17.0–20.4). Fall risk was also strongly associated with a T-POMA score ≤ 18, with an incidence rate of 33.6/100py (95% CI: 29.4–38.3), and this association was even more pronounced in the 85 + age group, where the incidence reached 38.8/100py (95% CI: 32.9–45.5). The highest incidence rates were observed among residents with moderate to severe mood and behavioural disorders (49.6/100py; 95% CI: 37.7–64.1) and those with hearing impairment (44.4/100py; 95% CI: 28.1–66.6), both within the T-POMA score range of 19–24 (Fig. 2 and Table S2, Supplementary material).
Fig. 2.
Crude incidence rates of first falls per 100 person-years by age classes, according to sex, physical restraint use, and T-POMA score. Abbreviations: T-POMA, Tinetti Performance-Oriented Mobility Assessment
Several factors were significantly associated with an increased risk of falls (Table 2). In the adjusted model, residents aged 85 years or older had a 39% higher risk of falling compared to younger (HR 1.39; 95% CI: 1.02–1.90), while male residents exhibited a 47% higher risk compared to females (HR 1.46; 95% CI: 1.27–1.71). A T-POMA score of ≤ 18 was strongly linked to an increased fall risk (HR 2.13; 95% CI: 1.61–2.81). An MMSE score indicative of moderate or severe cognitive impairment increased the risk of falling by 30% (HR 1.30; 95% CI: 1.10–1.53). An increased risk of over 25% was also associated with moderate or severe hearing impairment (HR 1.26; 95% CI: 1.01–1.58), the presence of moderate/severe mood and behavioural disorders (HR 1.29; 95% CI: 1.09–1.54), and the use of therapies influencing the risk of falls (HR 1.29; 95% CI: 1.09–1.52). Conversely, poor general health conditions and physical restrain showed HRs of 0.63 (95% CI: 0.52–0.77) and 0.70 (95% CI: 0.58–0.86), respectively.
Table 2.
Unadjusted and adjusted hazard ratios (HR) with 95% confidence intervals (CI) for first falls
| Residents’ characteristics | Unadjusted HR [95% CI] | Adjusted HR [95% CI] |
|---|---|---|
| Age class | ||
| 65–74 | Reference | Reference |
| 75–84 | 1.06 [0.77–1.46] | 1.20 [0.87–1.66] |
| 85+ | 1.09 [0.80–1.48] | 1.39 [1.02–1.90] |
| Sex | ||
| Females | Reference | Reference |
| Males | 1.44 [1.24–1.66] | 1.47 [1.27–1.71] |
| Length of stay | ||
| 0–2 years | Reference | Reference |
| > 2 years | 0.96 [0.78–1.17] | 1.09 [0.88–1.34] |
| T-POMA score * | ||
| 25+ | Reference | Reference |
| 19–24 | 1.78 [1.49–2.13] | 1.63 [1.30–2.05] |
| ≤18 | 2.24 [1.90–2.66] | 2.13 [1.61–2.81] |
| Physical restraint | ||
| No / Sporadic | Reference | Reference |
| Frequent / Daily | 0.52 [0.45–0.61] | 0.70 [0.58–0.86] |
| General health conditions | ||
| Good / Discrete | Reference | Reference |
| Poor | 0.51 [0.43–0.61] | 0.63 [0.52–0.77] |
| Visual impairment | ||
| No / Mild | Reference | Reference |
| Moderate / Severe | 0.74 [0.56–0.96] | 0.81 [0.61–1.07] |
| Hearing impairment | ||
| No / Mild | Reference | Reference |
| Moderate / Severe | 1.09 [0.88–1.35] | 1.26 [1.01–1.58] |
| Cognitive impairment (MMSE) | ||
| 18+ (No / Mild) | Reference | Reference |
| < 18 (Moderate / Severe) | 0.93 [0.81–1.07] | 1.30 [1.10–1.53] |
| Functional assessment (Barthel index) | ||
| 75+ (Autonomous / Mild Dependence) | Reference | Reference |
| < 75 (Moderate / Severe Dependence) | 0.58 [0.47–0.70] | 1.05 [0.82–1.34] |
| Mood and behaviour disorders | ||
| No / Mild | Reference | Reference |
| Moderate / Severe | 1.23 [1.05–1.43] | 1.29 [1.09–1.54] |
| Comorbidities influencing the risk of falls | ||
| No / Not Relevant | Reference | Reference |
| Relevant | 0.86 [0.73–1.01] | 1.16 [0.96–1.40] |
| Therapies influencing the risk of falls | ||
| No / Mild Risk | Reference | Reference |
| Moderate / Severe Risk | 1.11 [0.96–1.28] | 1.29 [1.09–1.52] |
| Previous falls within the last 180 days | ||
| No | Reference | Reference |
| 1+ | 1.07 [0.87–1.31] | 1.18 [0.96–1.44] |
Abbreviations: T-POMA, Tinetti Performance-Oriented Mobility Assessment; MMSE, Mini-Mental State Examination
The distribution of fall modalities and consequences is presented in Fig. 3. Severe fall consequences occurred in 120 out of 754 fallers, 15.1%. The multivariable analysis (Table S3, Supplementary Material) showed that lower T-POMA scores (≤ 18) and frequent or daily use of physical restraints were significant predictors of severe fall consequences, increasing the risk by 2.46 times (IRR, 95% CI: 1.18–5.12) and 2.00 times (IRR, 95% CI: 1.13–3.57), respectively.
Fig. 3.
Frequencies of fall modalities and consequences. More than one consequence of a single fall can be recorded per patient
Discussion
Fall incidence among long-term NH residents in Italy was high, with several key demographic and clinical characteristics influencing fall risk. The observed estimates align with previous studies showing that this population experiences falls at significantly higher rates than community-dwelling older adults [6, 21], likely due to a combination of advanced age, multimorbidity, cognitive decline, and mobility impairments, highlighting the need for comprehensive risk assessment and tailored intervention programs.
The stratification of fall incidence based on the T-POMA score supports its predictive validity, with residents in the high-risk category (≤ 18) exhibiting a significantly greater likelihood of falling compared to those in the moderate (19–24) and low-risk (≥ 25) groups. Poor balance and gait disorders were strongly associated with increased falls, a finding consistent with previous studies emphasizing balance disorder as a major risk factor in institutionalized older adults [9, 10]. Additionally, moderate-to-severe functional dependence was highly prevalent among fallers, indicating that mobility limitations substantially contribute to fall susceptibility. Comprehensive functional and mobility assessments in NHs are essential for identifying individuals at heightened risk and implementing early preventive measures. Notably, in our study, the level of independence in performing basic ADLs assessed by the Barthel Index was not associated with an increased fall risk. These findings emphasize the need for further research to refine evaluation methods and guide the development of tailored rehabilitation programs aimed at improving balance and mobility among NH residents [22].
In most older populations, falls are typically more common among older adult females, though males often experience more severe outcomes [23, 24]. Similarly to other studies conducted in NH settings, we found that male residents have a higher incidence of falls compared to their female counterparts. The increased risk of falls in male NH residents may firstly be attributed to a higher prevalence of health issues that predispose them to falls, including neurological disorders, balance impairments, or cardiovascular conditions [7]. Therefore, the higher fall incidence observed among males in this research may reflect these nuanced dynamics within the NH environment, highlighting that the relationship between sex and fall incidence can vary based on specific population characteristics and settings.
In addition to the well-established role of impaired mobility, in our multivariable analysis several other clinical factors emerged as a significant predictor of falls in NH residents. Moderate to severe hearing impairment emerged as a significant predictor, which reflects the impact of sensory deficits on balance, environmental awareness, and timely response to hazards, all of which are critical for fall prevention [25]. Similarly, cognitive impairment was associated with a 30% increased fall risk, aligning with evidence indicating that deficits in attention, executive function, and judgment can compromise safety and increase vulnerability to falls in institutionalized older adults, among whom the prevalence of conditions associated with cognitive decline is high [26, 27]. The presence of moderate to severe mood and behavioural disorders was also linked to higher fall risk. Symptoms such as agitation, impulsivity, or depressive withdrawal may influence fall either directly—by altering behaviour—or indirectly—through reduced engagement in preventive care or structured activities [28]. Finally, the use of therapies known to influence fall risk (Table S1) was associated with a 29% increase in fall risk. In long-term care settings, polypharmacy is common and often involves medications with sedative or hypotensive effects, underscoring the importance of carefully evaluating fall-related risks—an aspect that is often overlooked at the time of prescription due to pressing clinical needs [29, 30]. Together, these findings support a multifactorial approach to fall prevention that incorporates not only physical and functional assessments but also systematic evaluation of sensory, cognitive, psychological, and pharmacological risk factors. Tailoring interventions to address this complex interplay is critical to effectively reduce fall risk and improve resident safety.
Notably, residents with poor general health conditions showed a reduced fall risk. This counterintuitive finding underscores the complex relationship between frailty, mobility, and fall risk. It may be explained by the fact that residents with severe illnesses or extreme mobility limitations engage in fewer high-risk activities, spend more time in bed or seated, and receive closer supervision from caregivers. It also highlights that evaluating general health status alone is not sufficient for planning interventions, which should instead be tailored based on both functional and clinical profiles.
While results, in line with some reports, suggested that physical restraint may reduce immediate fall occurrence, the ethical and clinical implications of restraint use remain controversial [8, 31]. Prolonged restraint use has been linked to increased frailty, reduced muscle strength, and other adverse outcomes, necessitating a careful balance between fall prevention and residents’ functional autonomy [31–33]. Our findings suggest that the use of physical restraints may also be associated with more severe consequences when falls occur. Several possible explanations support this association, including reduced protective reflexes and mobility due to deconditioning, an increased risk of high-trauma falls—often abrupt or involving head-first impact—and heightened agitation, which may lead to more forceful or uncontrolled falls [32].
Several key clinical implications emerge from our findings. Routine mobility and functional assessments should be incorporated into NH care protocols to identify high-risk residents early, and personalized fall prevention strategies should consider sex-specific differences in fall patterns, activity levels, and clinical risk factors. Multifactorial interventions, including strength and balance training, medication reviews, environmental modifications, and tailored rehabilitation programs, should be prioritized over restrictive measures. In this sense, the integration of EHRs in NHs offers an opportunity to develop real-time fall risk prediction models based on longitudinal data. Also, technology-driven solutions, such as sensor-based monitoring and AI-assisted fall risk prediction models, could enhance early detection and intervention efforts [34].
Beyond their clinical implications, falls are recognized as a key quality indicator in NHs, reflecting the effectiveness of clinical risk management and safety protocols [31, 35]. A high incidence of falls can signal gaps in care pathways, insufficient staffing, or inadequate environmental adaptations, all of which impact the overall quality of care in NHs. Moreover, falls–particularly in NHs–carry substantial economic consequences. The costs associated with fall-related hospitalizations, fracture treatment, rehabilitation, and long-term disability care place a significant burden on healthcare systems. Implementing effective fall prevention strategies in NHs could not only reduce morbidity and mortality but also lower healthcare expenditures, making fall prevention a critical public health priority [36, 37].
A major strength of this study is its large sample size and the use of comprehensive EHR data from multiple nursing homes across Italy, enhancing the generalizability of our findings on fall risk factors in institutionalized older adults to similar long-term care settings. It is worth noting that, according to the most recent data from the Italian National Institute of Statistics (ISTAT), there are 12,363 active long-term residential care facilities in the country (regardless of facility type or number of beds per facility), accommodating approximately 362,850 residents as of December 31, 2022. More than 75% of these residents are aged 65 or older, and this number is once again increasing in line with the trend observed in the years preceding the COVID-19 pandemic [38].
Additionally, the inclusion of objective assessments (T-POMA) and detailed clinical characteristics enhances the robustness of our study. However, some limitations should be acknowledged. First, a standardized evaluation of functional status was not conducted uniformly across all residents at the same time prior to the fall. Instead, the analysis relied on the most recently recorded T-POMA scores, which were collected at varying time intervals prior to the fall event for each resident. Second, unmeasured confounders—such as environmental factors (e.g., flooring conditions, lighting), staff–resident ratio, and residents’ engagement in fall prevention programs—may have influenced fall risk but were not available for inclusion in the analysis. Although our dataset includes residents from 32 facilities, the absence of these covariates prevented us from conducting a facility-level adjusted analysis. Future studies should incorporate such structural characteristics to enable multilevel modelling and more accurately assess the impact of facility-level factors on fall risk. Third, the classification of health status reported in the database was based on the subjective clinical judgment of the attending physician, which may limit the reliability and interpretability of this variable in the analysis. Fourth, the database did not specify whether the date of cohort exit was due to death or discharge. As a result, it was not possible to apply a survival analysis framework that accounts for competing risks, preventing us from estimating marginal probabilities of events in the presence of competing outcomes. Fifth, the database did not include information on the fall-related outcome experienced by each individual resident, limiting the assessment of the health consequences of falls or their impact on healthcare resource utilization. An additional consideration is that the study period overlapped with a time when healthcare facilities were still coping with the challenges imposed by the COVID-19 pandemic [39]. As such, the results may have been influenced by pandemic-related disruptions in staffing, care routines, and resident activity levels—even though our findings are broadly consistent with those reported in non-Italian settings during other time periods.
In conclusion, this study provides valuable epidemiological data on fall incidence and risk factors among NH residents in Italy, highlighting the significant role of mobility impairments, age, sex, and clinical factors in determining fall risk. Our findings underscore the importance of routine functional assessments, structured fall prevention programs, and cautious evaluation of physical restraint use in NH settings. Future research should explore personalized, technology-assisted interventions to enhance fall risk detection and prevention in long-term care facilities. By implementing evidence-based fall prevention strategies, NHs can improve residents’ safety, mobility, and overall quality of life, ultimately reducing the healthcare burden associated with falls in older adults.
Electronic Supplementary Material
Below is the link to the electronic supplementary material.
Author contributions
P.F., C.C.M., D.R. and G.M. designed the study and drafted the manuscript. D.R. planned and performed the statistical analysis. C.F. participated in the data analysis and interpretation. All authors provided input on study design, participated in interpreting the results, contributed to the final version of the manuscript, and had final responsibility for the decision to submit the manuscript for publication.
Funding
This research received no external funding.
Data availability
No datasets were generated or analysed during the current study.
Declarations
Ethical approval
This study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of the University of Milan-Bicocca, under protocol no. 928/2025. All data were analysed anonymously.
Consent to participate
Not applicable.
Consent to publish
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Salari N, Darvishi N, Ahmadipanah M, Shohaimi S, Mohammadi M (2022) Global prevalence of falls in the older adults: a comprehensive systematic review and meta-analysis. J Orthop Surg 17(1):334. 10.1186/s13018-022-03222-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Hirase T, Okubo Y, Delbaere K, Menant JC, Lord SR, Sturnieks DL (2023) Risk factors for falls and Fall-Related fractures in Community-Living older people with pain: A prospective cohort study. Int J Environ Res Public Health 20(11):6040. 10.3390/ijerph20116040 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Schoene D, Heller C, Aung YN, Sieber CC, Kemmler W, Freiberger E (2019) A systematic review on the influence of fear of falling on quality of life in older people: is there a role for falls? Clin Interv Aging 14:701–719. 10.2147/CIA.S197857 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.UK National Institute for Health and Care Excellence. Falls in older people. NICE quality standards. Published online January 31, 2017. Accessed January 7 (2025) https://www.nice.org.uk/guidance/qs86
- 5.Cameron ID, Dyer SM, Panagoda CE et al (2018) Interventions for preventing falls in older people in care facilities and hospitals. Cochrane bone, joint and muscle trauma group. Ed Cochrane Database Syst Rev 2020(1). 10.1002/14651858.CD005465.pub4 [DOI] [PMC free article] [PubMed]
- 6.Huang L, Lai FTT, Yan VKC et al (2022) Comparing hybrid and regular COVID-19 vaccine-induced immunity against the Omicron epidemic. Npj Vaccines 7(1):162. 10.1038/s41541-022-00594-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Centers for Disease Control and Prevention. Falls in Nursing Homes. Accessed January 7 (2025) https://www.in.gov/health/files/CDC_Falls_in_Nursing_Homes.pdf [DOI] [PubMed] [Google Scholar]
- 8.Shao L, Shi Y, Xie XY, Wang Z, Wang ZA, Zhang JE (2023) Incidence and risk factors of falls among older people in nursing homes: systematic review and Meta-Analysis. J Am Med Dir Assoc 24(11):1708–1717. 10.1016/j.jamda.2023.06.002 [DOI] [PubMed]
- 9.Messinger-Rapport B, Dumas LG (2009) Falls in the nursing home: A collaborative approach. Nurs Clin North Am 44(2):187–195. 10.1016/j.cnur.2009.03.001 [DOI] [PubMed] [Google Scholar]
- 10.Montero-Odasso M, Van Der Velde N, Martin FC et al (2022) World guidelines for falls prevention and management for older adults: a global initiative. Age Ageing 51(9):afac205. 10.1093/ageing/afac205 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Omaña H, Bezaire K, Brady K et al (2021) Functional reach test, Single-Leg stance test, and Tinetti Performance-Oriented mobility assessment for the prediction of falls in older adults: A systematic review. Phys Ther 101(10):pzab173. 10.1093/ptj/pzab173 [DOI] [PubMed] [Google Scholar]
- 12.Köpke S, Meyer G (2006) The Tinetti test: Babylon in geriatric assessment. Z Für Gerontol Geriatr 39(4):288–291. 10.1007/s00391-006-0398-y [DOI] [PubMed] [Google Scholar]
- 13.Scott V, Votova K, Scanlan A, Close J (2007) Multifactorial and functional mobility assessment tools for fall risk among older adults in community, home-support, long-term and acute care settings. Age Ageing 36(2):130–139. 10.1093/ageing/afl165 [DOI] [PubMed] [Google Scholar]
- 14.Beck Jepsen D, Robinson K, Ogliari G et al (2022) Predicting falls in older adults: an umbrella review of instruments assessing gait, balance, and functional mobility. BMC Geriatr 22(1):615. 10.1186/s12877-022-03271-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Park J, Koh SB, Kim HJ et al (2018) Validity and reliability study of the Korean Tinetti mobility test for parkinson’s disease. J Mov Disord 11(1):24–29. 10.14802/jmd.17058 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Sterke CS, Huisman SL, Van Beeck EF, Looman CWN, Van Der Cammen TJM (2010) Is the Tinetti performance oriented mobility assessment (POMA) a feasible and valid predictor of short-term fall risk in nursing home residents with dementia? Int Psychogeriatr 22(2):254–263. 10.1017/S1041610209991347 [DOI] [PubMed] [Google Scholar]
- 17.Lamb SE, Jørstad-Stein EC, Hauer K, Becker C, on behalf of the Prevention of Falls Network Europe and Outcomes Consensus Group (2005) The Prevention of Falls Network Europe Consensus. J Am Geriatr Soc 53(9):1618–1622. 10.1111/j.1532-5415.2005.53455.x. Development of a Common Outcome Data Set for Fall Injury Prevention Trials: [DOI] [PubMed]
- 18.Kiani R, Miller H (2010) Sensory impairment and intellectual disability. Adv Psychiatr Treat 16(3):228–235. 10.1192/apt.bp.108.005736 [Google Scholar]
- 19.Folstein MF, Folstein SE, McHugh PR (1975) Mini-mental state. J Psychiatr Res 12(3):189–198. 10.1016/0022-3956(75)90026-6 [DOI] [PubMed]
- 20.GG (2015) The Barthel index: Italian translation, adaptation and validation. Int J Neurol Neurother. (2015)2(1). 10.23937/2378-3001/2/1/1028
- 21.Grambsch PM, Therneau TM (1994) Proportional hazards tests and diagnostics based on weighted residuals. Biometrika 81(3):515–526. 10.1093/biomet/81.3.515
- 22.Pham SN (2025) Wellness In Motion: A Quality Improvement Initiative to Support Resident Mobility and Reduce Falls. (Master’s Theses and Capstones). Published online 2024. Accessed January 7. https://scholars.unh.edu/thesis/1813/
- 23.Appeadu MK, Bordoni B (2025) Falls and Fall Prevention in Older Adults. In: StatPearls. StatPearls Publishing; Accessed January 7, 2025. http://www.ncbi.nlm.nih.gov/books/NBK560761/ [PubMed]
- 24.Gale CR, Cooper C, Aihie Sayer A (2016) Prevalence and risk factors for falls in older men and women: the english longitudinal study of ageing. Age Ageing 45(6):789–794. 10.1093/ageing/afw129 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Lach HW, Lozano AJ, Hanlon AL, Cacchione PZ (2020) Fear of falling in sensory impaired nursing home residents. Aging Ment Health 24(3):474–480. 10.1080/13607863.2018.1537359 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Kim SJ, Kim HD (2024) Relationship between falls, cognitive decline, and dementia in older adults: insights from the Korean longitudinal study of aging, 2006–2020. Exp Gerontol 194:112481. 10.1016/j.exger.2024.112481 [DOI] [PubMed] [Google Scholar]
- 27.Leroy V, Martinet V, Nunkessore O et al (2023) The nebulous association between cognitive impairment and falls in older adults: A systematic review of the literature. Int J Environ Res Public Health 20(3):2628. 10.3390/ijerph20032628 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Iaboni A, Flint AJ (2013) The complex interplay of depression and falls in older adults: A clinical review. Am J Geriatr Psychiatry 21(5):484–492. 10.1016/j.jagp.2013.01.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.De Vries M, Seppala LJ, Daams JG et al (2018) Fall-Risk-Increasing drugs: A systematic review and Meta-Analysis: I. Cardiovascular drugs. J Am Med Dir Assoc 19(4):371..e1-371.e9 [DOI] [PubMed] [Google Scholar]
- 30.Seppala LJ, Van De Glind EMM, Daams JG et al (2018) Fall-Risk-Increasing drugs: A systematic review and Meta-analysis: III. Others. J Am Med Dir Assoc 19(4):372e. 1-372.e8 [DOI] [PubMed] [Google Scholar]
- 31.Sze TW, Leng CY, Lin SKS (2012) The effectiveness of physical restraints in reducing falls among adults in acute care hospitals and nursing homes: a systematic review. JBI Libr Syst Rev 10(5):307–351. 10.11124/jbisrir-2012-4 [DOI] [PubMed] [Google Scholar]
- 32.Soeno S, Takada T, Takeshima T et al (2021) Association between the use of physical restraint and functional decline among older inpatients admitted with pneumonia in an acute care hospital: A retrospective cohort study. Arch Gerontol Geriatr 94:104330. 10.1016/j.archger.2020.104330 [DOI] [PubMed] [Google Scholar]
- 33.Chou MY, Hsu YH, Wang YC et al (2020) The adverse effects of physical restraint use among older adult patients admitted to the internal medicine wards: A Hospital-Based retrospective cohort study. J Nutr Health Aging 24(2):160–165. 10.1007/s12603-019-1306-7 [DOI] [PubMed] [Google Scholar]
- 34.González-Castro A, Leirós-Rodríguez R, Prada-García C, Benítez-Andrades JA (2024) The applications of artificial intelligence for assessing fall risk: systematic review. J Med Internet Res 26:e54934. 10.2196/54934 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Amato S, Aniuskevich A, Lombardi AM, Amato F (2024) Healthcare risk management and monitoring of patient falls in Italian National service. The experience of a local health authority. Ig E Sanita Pubblica 80(2):30–40 [PubMed] [Google Scholar]
- 36.LeLaurin JH, Shorr RI (2019) Preventing falls in hospitalized patients. Clin Geriatr Med 35(2):273–283. 10.1016/j.cger.2019.01.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Gannon B, O’Shea E, Hudson E (2008) Economic consequences of falls and fractures among older people. Ir Med J 101(6):170–173 [PubMed] [Google Scholar]
- 38.Istituto Nazionale di Statistica (ISTAT) (2025) Le strutture residenziali socio-assistenziali e socio-sanitarie al 1° gennaio 2023. ISTAT: Roma, 6 febbraio 2025. Available at: https://www.istat.it/wp-content/uploads/2025/02/Report-Presidi-2022.pdf Last accessed on June 10
- 39.Ferrara P, Albano L (2020) COVID-19 and healthcare systems: what should we do next? Public Health 185:1–2. 10.1016/j.puhe.2020.05.014 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Citations
- Lamb SE, Jørstad-Stein EC, Hauer K, Becker C, on behalf of the Prevention of Falls Network Europe and Outcomes Consensus Group (2005) The Prevention of Falls Network Europe Consensus. J Am Geriatr Soc 53(9):1618–1622. 10.1111/j.1532-5415.2005.53455.x. Development of a Common Outcome Data Set for Fall Injury Prevention Trials: [DOI] [PubMed]
Supplementary Materials
Data Availability Statement
No datasets were generated or analysed during the current study.