The Growing Burden of Fall-Related Injuries among Older Adults: A Seven-Year Study from a Tertiary Medical Center in Taiwan
Article information
Abstract
Background
As Taiwan’s population ages, falls among older adults have become a critical public health concern. However, limited data exist regarding temporal trends and injury patterns in fall-related emergency department (ED) visits. This study aimed to examine trends in fall-related ED visits and hospitalizations among older adults in Taiwan and to explore injury distributions by age group.
Methods
We conducted a retrospective cohort study using data from the National Taiwan University Hospital between 2011 and 2017. Patients aged ≥65 years were compared with those aged 20–64 years. Fall-related visits were identified using chief complaints and the International Classification of Diseases 9th/10th revision (ICD-9/ICD-10) codes. Outcomes included hospitalization rates, length of stay, and 30-day mortality.
Results
A total of 22,471 fall-related ED visits were analyzed. While visits among younger adults declined (annual growth rate, -1.34%), visits among older adults increased (2.37% annually), with the steepest rise in those aged ≥85 years. Hospitalization occurred in 27.1% of older adults, nearly double that of younger adults (14.4%). Older adults also had longer hospital stays and higher 30-day mortality rates, findings consistent even when restricted to lower limb fractures.
Conclusion
Fall-related ED visits and hospitalizations are rising disproportionately among Taiwan’s older population. Targeted prevention strategies and transitional care interventions are urgently needed to address the growing clinical and economic burden of falls in aging societies.
INTRODUCTION
As Taiwan became an aged society in 2018 and is projected to become a super-aged society in 2025, falls among older adults have emerged as a critical public health concern due to their increasing prevalence and the severe health consequences they pose. Previous studies have shown a year-on-year increase in emergency department (ED) visits related to falls or fall-related injuries among older adults.1,2) Falls in older adults are closely associated with a range of adverse outcomes, including recurrent ED visits, hospitalizations, long-term disability, and mortality.3,4) Frailty further amplifies these risks, making older adults particularly vulnerable to complications following a fall.5) Beyond clinical consequences, the economic burden is substantial. According to the United States Centers for Disease Control and Prevention (CDC), the costs associated with hospitalizing older adults due to falls have risen by 7.8% over the past 8 years.6) These trends underscore the urgent need for effective fall prevention strategies that address both health outcomes and economic sustainability.
While Taiwan has implemented various fall prevention initiatives, the burden of fall-related ED visits appears to be increasing.7) Interestingly, previous studies from other countries have reported no significant upward trend in fall-related ED visits among older adults,8,9) suggesting that cross-national differences in healthcare systems, aging perceptions, and preventive policies may influence patterns of care-seeking. The contrast between Taiwan’s rising trend and the plateau observed in other countries highlights the importance of contextualized evidence for designing region-specific interventions.
Moreover, although fractures—particularly hip fractures—are recognized as the leading cause of fall-related hospitalizations in older adults,10) there remains limited understanding of the broader spectrum of injury types and their distribution among different age groups. A comprehensive analysis of injury patterns can help inform emergency care planning and resource allocation for aging populations.
This study aims to examine temporal trends in fall-related ED visits and hospitalizations among older adults in Taiwan and to explore the distribution of injury types based on age groups. Despite the growing urgency of the issue, few studies in Asia have quantified these trends using data from large tertiary medical centers. By addressing this gap, our findings may provide valuable insights for both national policy development and global discussions on fall prevention in rapidly aging societies.
MATERIALS AND METHODS
Study Design and Time Period
This observational cohort study retrospectively collected data from the integrated medical database of National Taiwan University Hospital (NTUH). The database serves as a centralized clinical data repository for electronic health records (EHR) across the healthcare system, covering the main hospital and six branch hospitals. It includes information from inpatient, outpatient, and ED visits, such as demographics, diagnoses, treatments, imaging, laboratory results, prescriptions, nursing notes, and administrative data. The database is maintained and updated by dedicated research staff and has been utilized in prior research studies.11,12)
For this study, we retrieved seven years of ED data from the main hospital, spanning January 1, 2011 to December 31, 2017. All data, including demographic information and the International Classification of Diseases (ICD) diagnostic codes, were registered as part of the EHR for the specific ED encounter; therefore, the ICD code registration date corresponds exactly to the ED visit date. The NTUH main hospital is a tertiary academic medical center with approximately 2,400 beds and handles around 100,000 ED visits annually. The case selection process, including all inclusion and exclusion criteria, followed the flow outlined in Fig. S1. This study was approved by the Institutional Review Board of National Taiwan University Hospital (Protocol Number: 201905088RIN).
Study Population
The study focused primarily on ED patients aged 65 and older, with those aged 20 to 64 included as a comparison group. As the study period coincided with the transition to the ICD-10 system at NTUH, the database contains diagnostic codes from both the ICD-9 and ICD-10 classification systems. Patients were included if they met any of the following criteria: (1) a mention of "fall" in the chief complaint or present illness, (2) ICD-10 codes W00-W19, (3) ICD-9 codes E880-E888, which correspond to external cause of injury codes specifically indicating falls. Cases with missing data were excluded.
As outlined in Fig. S1, the initial sample included 25,374 visits. We first excluded 156 visits for patients under 20 years of age and 3 visits unrelated to falls, resulting in 25,215 visits. We further excluded 205 visits in which the ED disposition was coded as hospitalization, but no corresponding inpatient record was identified. In addition, we excluded 2,539 visits with missing ICD codes. This selection process yielded a sample of 22,471 visits for analysis.
Outcome Measurements
In addition to basic demographic variables, outcome variables following the emergency visit included subsequent hospitalization rate, length of stay and 30-day mortality. The database is linked to Taiwan’s national cause of death registry, enabling the assessment of 30-day mortality.
Cases were classified based on ICD codes into the following injury categories:
• Minor injuries: this category was defined by diagnoses that did not involve fractures or internal organ injuries. It primarily included ICD-9 and ICD-10 codes representing abrasions, contusions, sprains/strains, and superficial open wounds.
• Skull fractures or intracranial hemorrhages: ICD-10 codes S02, S06 or ICD-9 codes 800, 801, 802, 803, 804, 850, 851, 852, 853, 854.
• Spinal fractures and neurological injuries: ICD-10 codes S12, S14, S22.0, S32.0, S32.1, S32.2, S32.9, S24, S34 or ICD-9 codes 805, 806, 905, 907, 952.
• Severe chest injuries: ICD-10 codes S22.2, S22.3, S22.4, S22.5, S27 or ICD-9 codes 807, 860, 861, 862.
• Severe abdominal and pelvic injuries: ICD-10 codes S36, S32.3, S32.4, S32.5, S32.6, S32.7, S32.8 or ICD-9 codes 863, 864, 865, 866, 867, 868, 808, 902, 926.
• Upper limb fractures and crush injuries: ICD-10 codes S42, S47, S52, S57, S62, S67 or ICD-9 codes 808, 810, 811, 812, 813, 814, 815, 816, 867, 868, 927.
• Lower limb fractures and crush injuries: ICD-10 codes S72, S75, S77, S82, S87, S92, S97 or ICD-9 codes 820, 821, 822, 823, 824, 825, 826, 896, 928.
• Burns: ICD-10 codes T20, T21, T22, T23, T24, T25, T26 or ICD-9 codes 940, 941, 942, 943, 944, 945, 947, 948, 949.
• Major trauma: defined by ICD-10 code T07 (unspecified multiple injuries). This category inherently overlaps with other specific injuries (e.g., fractures) but is analyzed distinctly. In Taiwan, T07 serves as a specific clinical and administrative marker for severe polytrauma (often Injury Severity Score ≥16).
Statistical Analysis
We estimated the annual growth rate (AGR) of ED visits and hospitalizations from 2011 to 2017 by fitting a log-linear regression model to the annual case counts. For this analysis, the natural logarithm of the annual case count was the dependent variable, and the calendar year was the independent variable, i.e., log(case count)=β0+β1 (year). The AGR was then calculated by exponentiating the slope coefficient (β1) and subtracting 1, using the formula:
This analysis was stratified by age groups. Trends were also visualized using linear regression plots. The distribution of injury types was compared between older patients (aged ≥65 years) and non-older patients (aged 20–64 years), with injury types classified into the categories. All analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA).
RESULTS
A total of 25,374 ED visits were initially identified. After excluding patients aged under 20, non-fall-related cases, cases with a hospitalization disposition but without inpatient record, and cases missing ICD codes, a final sample of 22,471 fall-related ED visits was included in the present analysis, accounting for 3.5% of the 639,482 total adult ED visits. Among these, 12,175 cases involved older adults (aged ≥65 years) and 10,296 involved younger adults (aged 20–64 years). A divergent trend in fall-related ED visits was observed across age groups from 2011 to 2017. Among adults aged 20–64 years, the number of fall-related visits slightly declined, with an average AGR of -1.34% (Fig. 1A). In contrast, older adults (aged ≥65 years) showed an increasing trend, with a 2.37% average AGR (Fig. 1B; 0.67% for aged 65–74 years in, Figs. S2 and 2.79% for aged 75–84 years in Fig. S3). Notably, the most rapid increase was seen in those aged ≥85 years, with an AGR of 3.36% (Fig. S4).
Annual trends in fall-related emergency department visits: (A) aged 20–64 years group, (B) older adults (aged ≥65 years) group. The solid line indicates the observed annual number of cases; the dashed line represents the linear trend. Annual growth rates are -1.34% (aged 20–64 years) and 2.37% (aged ≥65 years).
Approximately 80% of all fall-related ED visits involved minor injuries. However, among the older adults, the most common moderate to severe injuries were lower limb fractures, followed by upper limb fractures. Compared to adults aged 20–64 years, older adults were significantly vulnerable to skull fracture or intracranial hemorrhage, compression spinal fracture with neurological injuries, abdominal and pelvic fractures and lower limb fractures and crush injuries (all p<0.001) (Table 1). Furthermore, 27.1% of older adults required hospitalization due to falls, a significantly higher proportion compared to 14.4% in the younger adults (p<0.001) (Table 2). The AGR of fall-related hospitalizations among older adults was 3.51% (Fig. 2B), in sharp contrast to a decline of 2.20% observed among younger adults (Fig. 2A). These findings indicate that the burden of fall-related hospitalizations is increasing disproportionately in the older population.
Comparison of injury patterns in fall patients between older and younger adults
Distribution of outcomes for patients who visited the emergency department (ED) due to falls
Annual trends in fall-related hospitalization: (A) aged 20–64 years group, (B) older adults (aged ≥65 years) group. The solid line indicates the observed annual number of cases; the dashed line represents the linear trend. Annual growth rates are -2.20% (aged 20–64 years) and 3.51% (aged ≥65 years).
When focusing on patients who required hospitalization due to falls, older adults had significantly poorer outcomes than their younger counterparts. As detailed in Table 3, older adults had a significantly longer mean length of stay for all fall-related hospitalizations (12.07 vs. 9.04 days, p<0.001). Most notably, the 30-day mortality rate was nearly three times higher for older adults (15.32% vs. 5.81%, p<0.001). This pattern of longer hospital stay and higher mortality was also observed in the subgroup of patients with lower limb fractures, the most common injury in the older cohort (Table 3).
Comparison of outcomes between older and younger adults following fall-related hospitalization
DISCUSSION
This study underscores a concerning upward trend in fall-related ED visits among Taiwan’s older adult population. This trend likely reflects not only the demographic shift toward an aging society but also increases in frailty and changes in care-seeking behaviors. This is particularly evident in our hospital's catchment area (Taipei City), which aged more rapidly during this period (16.4% older adults by 2017) than the national average (approx. 13.9%), serving as a "harbinger" for future national trends. Our findings are consistent with previous research identifying falls as a leading cause of injury-related morbidity and mortality among older adults,1,13) highlighting the urgent need for age-specific fall prevention strategies.
Notably, the particularly steep rise in fall-related ED visits among individuals aged over 85 underscores their heightened vulnerability. This trend may be attributed to both the growing number of very old adults and the increasing burden of frailty within this subgroup, emphasizing the need for tailored preventive measures for the oldest old.
A clear difference in injury severity was also observed between older and younger adults. Among older adults, lower limb fractures were the most frequent moderate-to-severe injuries, whereas younger adults more often sustained minor injuries. Although previous studies have identified hip fractures as the most common injury following a fall in older adults,14) our broader classification under "lower limb fractures" remains consistent with these findings. This pattern may be due to older adults' greater susceptibility to falls caused by impaired balance15) and the well-established association between osteoporosis and fall-related fractures.16)
Hospitalization was required in 27.1% of fall-related ED visits among older adults, nearly twice the rate observed in younger patients. This is comparable to previous studies reporting a hospitalization rate of 19.44% among older fall patients.17) Furthermore, our findings demonstrate that the likelihood of hospitalization increases with advancing age, indicating that the oldest individuals are more vulnerable to sustaining severe injuries requiring inpatient care.18) These results underscore the substantial clinical and systemic burden posed by falls in older populations and highlight the need for targeted post-fall care and secondary prevention models.
Among hospitalized patients, older adults had significantly longer hospital stay and higher 30-day mortality rates compared to younger adults, consistent with prior research.19) These patterns persisted even when analyses were limited to patients with lower limb fractures. Such findings suggest that implementing targeted interventions to prevent recurrent falls among hospitalized older adults could substantially improve outcomes.
Although this study did not directly measure economic impact, the increasing hospitalization rates and initiation of long-term care20) suggest an increasing financial burden. Similar trends observed internationally21) suggest that without effective intervention, fall-related healthcare costs in Taiwan will continue to escalate. Evidence-based strategies—including exercise programs to enhance balance and strength, home environment modifications, and public health education—are critical to mitigating both the human and economic costs associated with falls.
Importantly, 72.5% of older adults in our study were discharged home after their ED visit, representing a large and often overlooked group for secondary prevention efforts. Prior studies have demonstrated that even a single fall substantially increases the risk of future falls,22,23) particularly among the very old,23) who also exhibited the most rapid growth in ED utilization in our data. Given that many older adults may not perceive themselves to be at risk after a fall, the ED provides a unique opportunity to identify high-risk individuals and intervene early before more serious injuries occur.
Thus, the ED can serve not only as a site for acute care but also as a critical gateway to community-based fall prevention services, including balance training, medication review, home safety assessments, and social support linkages. Integrating fall risk screening and referral pathways into routine ED workflows—especially for patients discharged home—could help prevent recurrent falls and reduce long-term disability. Future research should focus on evaluating the feasibility, effectiveness, and outcomes of such transitional care interventions initiated in the ED setting.
This study has several limitations. First, it was conducted using data from a single tertiary medical center, which may limit the generalizability of the findings to other healthcare settings in Taiwan. Although the NTUH database provides comprehensive clinical information, it may underrepresent individuals seeking care at regional hospitals or private clinics.
Second, our study period concluded at the end of 2017, as this was the extent of the data available to us from the database. This limitation prevents the analysis of more recent trends. These post-2017 trends are of significant interest, as they may have been influenced by factors such as the COVID-19 pandemic and other recent public health initiatives, warranting separate investigation in future studies.
Third, our study is subject to case-definition limitations inherent to retrospective database research. Due to the de-identified nature of the data, we could not perform manual chart reviews to improve specificity. This prevented us from definitively distinguishing between mechanical falls and falls secondary to other acute medical events (e.g., syncope, seizure, or motor vehicle accidents). Similarly, we could not reliably differentiate between fall types, such as ground-level falls versus falls from height. This is a direct consequence of our hospital's ED coding system, which, for clinical efficiency, defaults to the general 'W19 (Unspecified fall)' code. More specific codes (e.g., W09–W17 or Y92.*) are rarely used by clinicians. Therefore, our cohort represents the total burden of visits coded as falls in our system, rather than a perfectly defined clinical subgroup of mechanical falls.
Fourth, while the study identified trends in ED visits, hospitalizations, and injury patterns, it did not explore the underlying causes driving the observed increase in falls, nor did it evaluate the effectiveness of existing fall prevention interventions. This means our trend analysis could not adjust for unmeasured confounders (e.g., changes in population comorbidity or frailty over the study period), which is a potential source of bias. Future studies should leverage population-based datasets and incorporate intervention evaluations to inform the development of comprehensive, evidence-based fall prevention policies for Taiwan’s rapidly aging population.
In conclusion, this study demonstrates a growing burden of fall-related ED visits and hospitalizations among Taiwan’s older adult population, particularly among the oldest old. Older adults not only experienced a higher rate of hospitalization following falls, but also had longer hospital stays and significantly higher 30-day mortality rates, compared to younger adults. These findings highlight the urgent need for targeted fall prevention strategies tailored to the needs of older adults, both within the ED and through community-based interventions. Given the increasing clinical and economic burden associated with falls, integrated, system-level approaches are essential to improving outcomes for aging populations and mitigating the rising healthcare demands. Future research should focus on developing and evaluating transitional care models that link ED-based risk assessments to comprehensive fall prevention programs in the community.
Notes
We thank the staff of the Department of Medical Research, National Taiwan University Hospital for the Integrative Medical Database (NTUH-iMD).
CONFLICT OF INTEREST
The researchers claim no conflicts of interest.
FUNDING
None.
AUTHOR CONTRIBUTIONS
Conceptualization, YCT, FPL; Data curation, YCT, FPL; Methodology, YCT, SYC, FPL; Supervision, YMC, PCH; Formal analysis, YCT, FPL; Writing-original draft, YCT; Writing-review & editing, SYC, YMC, FPL.
SUPPLEMENTARY MATERIALS
Supplementary materials can be found via https://doi.org/10.4235/agmr.25.0128.
Flow diagram of case selection.
Annual change in number of falls (aged 65–74 years).
Annual change in number of falls (aged 75–84 years).
Annual change in number of falls (aged ≥85 years).