Journal Logo

Research: Research Article

The Accident Environment Resulting in Fragility Fractures: A 20-year National Epidemiologic Study

Makar, Gabriel S. MD; Davis, William MD; Suk, Michael MD, JD, MPH, MBA, FACS; Bowen, Thomas R. MD

Author Information
Journal of the American Academy of Orthopaedic Surgeons: July 1, 2022 - Volume 30 - Issue 13 - p e911-e918
doi: 10.5435/JAAOS-D-21-01169


Fragility fractures are an enduring source of morbidity in the elderly with unfortunate frequency and rising costs. Many health systems are expecting an increasing incidence of fragility fractures in future years. The United States is projected to have 72 million citizens older than 65 years by 2030 or approximately 20% of the population.1 Hip fractures alone cost Medicare $12 billion per year, and the overall fracture burden is projected to cost $25.3 billion per year by 2025.2-4 Fractures in the elderly yield decreased quality of life, place a strain on caregivers, and may be largely preventable. Furthermore, fractures in the elderly have been reported to generate psychological disturbance and an increased risk of suicide.5 Therefore, there is an ongoing need to optimize interventions aimed at preventing fractures in the aging.

Fragility fractures may be prevented by both the optimization of bone health and the prevention of trauma. Regarding trauma prevention, a better understanding of the location, circumstances, and mechanism of injury resulting in fragility fractures may direct where best to intervene. A 1985 study reported that 87% of fractures encountered in the emergency department resulted from falls.6 It is estimated that 22 to 32% of elderly patients have at least one fall per year, but less than half see a provider regarding their fall.7,8 One study reviewed trends in geriatric fractures from 2004 to 2014; however, the team did not study the product causes associated with the causes for fractures.9 Unfortunately, a detailed description of the physical location of the falls associated with the product type and the circumstances surrounding the trauma is otherwise lacking. As such, it is difficult for both the provider and the policymaker to implement interventions in a data-driven fashion to best minimize these injuries.

The purpose of this study was to identify the environmental location of accidents and commercial product types associated with the most morbid and common fragility fractures. The authors intend for this information to help direct specific, data-driven, practical patient interventions and perhaps even product innovations designed to decrease the incidence and severity of future fragility fractures.


We queried the National Electronic Injury Surveillance System (NEISS) database for all fractures found in patients older than 65 years presenting for care between January 1, 2000, and December 31, 2019. We recorded patient demographic data, such as age, sex, and race. We also recorded the environment in which injury occurred, such as home, public, unknown, street, and sport-associated. Low volume environments, such as schools, mobile home, farm, and industrial locations were grouped together and labeled “other” because individual numbers of injuries were too low, precluding additional analysis. The emergency department disposition was also extracted, including those who were “treated and released,” “admitted/hospitalized for treatment,” and “transferred to another facility for treatment.” Patients who “left without being seen,” have an “unknown” disposition, or were fatally injured were labeled “other” considering their disposition, again because of small individual sizes.

We analyzed the top 20 product codes for fractures. Patients who did not have a known sex, had an injury >25% of their body, or had an unknown location of injury were excluded from the analysis. Injuries >25% of the body and related only specifically to burns were excluded from the study.

Trends, proportions and distributions were analyzed using descriptive statistics. This study met the authors' institutional criteria for minimal risk to human subjects and was declared exempt from the Institutional Review Board review.

NEISS is an open, free database operated by the US Consumer Product Safety Commission (CPSC). The database collects data on consumer product-related injuries occurring in the United States. The CPSC data are intended to provide product-related statistics regarding emergency department visits. The data set and national estimates produced by the CPSC are derived from a nationally representative sample of 96 hospitals. Notably, each visit recorded by the database, or “case,” contains a brief narrative of the clinical scenario to assist in verification of the clinical data. NEISS coders use a body part diagram to assist in properly dictating the location of injury (Supplementary Figure 1, In addition, NEISS product codes are given for each case.


During the 20-year study period, there were 901,418 patient visits to the emergency department with acute injuries reported to the NEISS database. Of these, 216,657 patients (24%) were diagnosed with fractures. Both the number of patients presenting to the emergency department and the overall number of diagnosed fractures increased over time (Table 1 and Figure 1). Emergency department patient visits for all injuries increased from 31,135 in 2000 to 64,900 in 2019. Similarly, the total number of fractures rose from 8,087 to 14,823 by the end of the study period. The proportion of emergency department patients diagnosed with fractures remained largely unchanged, decreasing minimally from 26% in 2000 to 23% in 2019.

Table 1 - Top 20 Fractures, Total Fractures, and Total Emergency Department Visits Per Year
Year Age (Avg) Top 20 Fractures All Factures All Fracture Percentage Total Visits
2000 80.4 6,382 8,087 26 31,135
2001 80.3 6,421 8,290 25 32,915
2002 80.5 6,804 8,692 25 34,542
2003 80.1 7,244 8,824 26 34,132
2004 80.3 7,708 9,371 26 35,822
2005 80.3 7,697 9,402 25 37,256
2006 80.3 7,969 9,706 25 38,204
2007 80.4 8,127 9,964 25 39,942
2008 80.5 8,448 10,400 25 42,225
2009 80.2 8,709 10,757 24 45,483
2010 80.4 9,002 11,018 24 46,747
2011 80.2 8,490 10,608 23 45,434
2012 80.3 8,546 10,725 23 46,196
2013 80 8,897 11,243 24 47,629
2014 80 9,297 11,607 23 49,398
2015 80.1 9,839 12,259 24 51,688
2016 79.9 10,248 12,887 23 55,157
2017 79.8 10,855 13,882 23 60,814
2018 79.6 11,100 14,112 23 61,799
2019 79.6 11,774 14,823 23 64,900
Total 80.1 173,557 216,657 24 901,418

Figure 1:
Graph showing top 20 fractures causes, total fractures, and total emergency department visits per year.

The average age in our population was 80.1 years (SD 8.7). The average age of patients experiencing fractures trended minimally downward from 80.4 years in 2000 to 79.6 years in 2019. Women constituted most of the patients (127,753 [74%]), far outnumbering men (45,804 [26%]). Most of the patients were White (105,339 [61%]). Black/African American (7,371 [4%]), Hispanic (2,183 [1%]), other (1,940 [1%]), and Asian (1,157 [1%]) patients represented small fractions of the study population, although a large proportion of patients did not report their race (55,567 [32%]). Table 2 describes the patient characteristics of our cohort.

Table 2 - Patient Demographics
Category n (%)
Age (SD) 80.1 (8.7)
 Male 45,804 (26)
 Female 127,753 (74)
 White 105,339 (61)
 Black/African American 7,371 (4)
 Hispanic 2,183 (1)
 Others 1,940 (1)
 Asian 1,157 (1)
 Not specified 55,567 (32)

We analyzed the top 20 commercial product types associated with fractures in the study population which accounted for a total of 173,557 fractures, or 80% of all fractures in our cohort and 19% of all emergency department visits. The most common product types associated with fractures, floors/flooring material (58,347 [33.6%]), were more common than the second and third most common product types, combined (stairs/steps: 29,804 [17.2%] and bed/bed frames: 19,004 [10.9%]). Table 3 describes the top 20 most common product types associated with fractures in our population.

Table 3 - Top 20 Causes for Fractures in the Elderly
Category n (%)
Floors/flooring material 58,347 (33.6)
Stairs/steps 29,804 (17.2)
Bed/bed frames 19,004 (10.9)
Chair 11,055 (6.4)
Rugs/carpets 7,519 (4.3)
Bathtub/shower 6,163 (3.5)
Walkers 5,050 (2.9)
Toilets 4,662 (2.7)
Ladders 4,235 (2.4)
Footwear 3,783 (2.2)
Wheelchairs 3,299 (1.9)
Porches, balconies, open-side floors, or floor openings 3,195 (1.8)
Tables 3,162 (1.8)
Door associated 2,873 (1.7)
Bicycles 2,869 (1.7)
Wall 2,205 (1.3)
Sofas and couches 1,998 (1.2)
Day wear/clothing 1,519 (0.9)
Exercise 1,512 (0.9)
Pet supplies/pet associated 1,305 (0.8)
Total 173,557

We also analyzed the bone fractured, emergency department treatment disposition, and the environment of the injury (Table 4). Lower extremity fractures (97,195 [56%]) were more commonly reported than upper extremity fractures (63,899 [37%]). Of all fractures, the lower trunk (including the pelvis, femoral neck, and lower spine) was the most commonly fractured area reported (64,132 [37.0%]). Less common were upper trunk fractures (ribs, thoracic spine, and sternum) (20,245 [11.7%]), humerus fractures (12,479 [7.2%]), wrist fractures (10,996 [6.3%]), and nonfemoral neck femur fractures (9,993 [5.8%]).

Table 4 - Patient Injury and Hospital-specific Characteristics
Category n (%)
Fracture location
 Lower trunk (pelvis, femoral neck, and lower spine) 64,132 (37.0)
 Upper trunk (rib, thoracic spine, and sternum) 20,245 (11.7)
 Humerus 12,479 (7.2)
 Wrist 10,996 (6.3)
 Femur (nonfemoral neck) 9,993 (5.8)
 Shoulder (clavicle/scapula) 7,895 (4.5)
 Face (mandible) 7,600 (4.4)
 Ankle 7,574 (4.4)
 Lower leg (fibula/tibia) 5,948 (3.4)
 Lower arm (radius/ulna) 5,605 (3.2)
 Foot 3,918 (2.3)
 Neck (cervical spine) 3,900 (2.2)
 Knee (patella) 3,235 (1.9)
 Elbow 2,685 (1.5)
 Finger 2,423 (1.4)
 Toe 2,395 (1.4)
 Hand 1,571 (0.9)
 Head 821 (0.5)
 Pubic region 142 (0.1)
 Treated and admitted/hospitalized 87,610 (50.5)
 Treated and released 77,786 (44.8)
 Treated and transferred 5,835 (3.4)
 Held for observation 2,103 (1.2)
 Other 223 (0.1)
 Home 113,158 (65.2)
 Public 31,162 (18.0)
 Street 2,243 (1.3)
 Sports 1,110 (0.6)
 Other 175 (0.1)
 Unknown 25,709 (14.8)

Slightly more patients were treated and hospitalized (87,610 [50.5%]) than treated and released (77,786 [44.8%]). Few patients were transferred for additional management (5,835 [3.4%]) or held for observation (2,103 [1.2%]). Most fractures were sustained at home (113,158 [65.2%]), followed by a public setting (31,162 [18.0%]). A very few fractures occurred in the street (2,243 [1.3%]) or because of a sports injury (1,110 [0.6%]). Slightly more than one of every seven patients did not have the environment of their traumatic injury documented at the time of their emergency department visit (25,709 [14.8%]).


Fragility fractures negatively affect patients' quality of life, constitute a notable economic burden, and will only become more common as our population ages. Studies have reported falls as the most common cause of fractures in the elderly.8,10 The Centers for Disease Control reported that approximately 13.5% of Emergency Department visits were due to unintentional falls in patients older than 65 years, with that figure rising to 20% of visits in those older than 80 years.11 We sought to better understand fragility fractures by analyzing the associated product, resulting fractures, and environment of these injuries.

We found that approximately one-quarter of patients older than 65 years who presented to the emergency department after an incident were found to have a fracture, a much higher proportion than previously reported. One study investigating Emergency Department visits in the elderly found that hip fractures were among the least common cause of visits but were the most common cause of admission.12 This highlights the urgent nature and complexity of these injuries, which in turn generate greater cost and morbidity. The purpose of this study was to identify the environmental location of accidents and commercial product types associated with the most morbid and common fragility fractures with the goal of guiding-specific medical interventions to decrease the risk of fractures.

A similar study was done by Baidwan and Naranje reviewing NEISS data from 2004 to 2014.9 Their study notes a decrease in fractures occurring at home from 67% to 64%. On the contrary, our study found falls occurred more commonly at home during the last 5 years of our study compared with the first 5 years (65.7% and 63.7%, respectively). Although the differences in these rates are small and likely not clinically notable, it indicates that we have not decreased the rate of falls occurring at home, despite the home hosting most of these injuries. Their results support our findings that although lower trunk fractures are decreasing and upper trunk fractures are increasing, lower trunk injuries are still approximately 2.5 times more common. Our study and analysis largely focus on the products and causes for these fractures to provide more granular data on the underlying causes of these injuries.

Our study aims to direct specific interventions to decrease the fracture burden on the aging population. As mentioned above, our study found that nearly two-thirds of all fractures occurred at home. Perhaps unsurprisingly, “floor/flooring material” was far and away the most common cause of fractures in our study, representing over a third of all fractures in this population and more than the next two causes combined. Theoretically, these represent the most preventable type of fall because they indicate general instability due to various causes that may be correctable with tools such as assistance devices. Preventive medicine has become a large discussion in today's healthcare climate, but the discussion with patients to “accident-proof” their home is not an easy one. However, programs aimed at targeting patients who are at increased risk of falling at home could greatly alleviate the burden of fractures on the aging population. The annual Medicare wellness visit does include a fall screen and a Timed Up-and-Go test. However, it may prove cost-effective to include a more rigorous assessment, possibly including physical therapists, which evaluates the ability to turn and navigate stairs. Assistive devices, physical therapy, or other modalities may be offered earlier to patients who have difficulty walking and thus help prevent subsequent falls. In addition, our findings suggest that additional research into whether various flooring materials are higher risk or protective may be warranted; unfortunately, the NEISS database does not provide such granularity.

We found a declining rate of fractures compared with all-cause visits in the Emergency Department; the rate at which fractures increased per year did not match the overall rate of other visits to the Emergency Department. Approximately 26% of the Emergency Department visits were fractures in 2000, which fell to only 23% in 2019. Although there was a general decreased proportion of patients presenting to the Emergency Department with fractures compared with the overall number of visits, there was a positive trend in the number of fractures per year. The increased number of outpatient clinics, urgent care centers, and physician offices may have diverted some patients from the Emergency Department, which would not be captured by the NEISS database. However, one would expect nonfracture injuries to be diverted in greater proportions. As such, it is more likely that there is some other cause for this decline. Perhaps superior management of osteoporosis has helped prevent more fractures; additional research is needed to elucidate the cause of this change.

Interestingly, a study found that admissions for elderly patients decreased from 42% in 2006 to 33.6% in 2015.13 This is consistent with our finding of a falling proportion of Emergency Department visits discovering fractures. It is possible that the elderly population is generally healthier because of modern medical advances and skewing progressively younger as an ever-increasing number of Baby Boomers begin turning 65. It may be necessary to reconsider what we define as “elderly” because a greater proportion of these newly minted seniors are in relatively good health, whether through stratifying by age or by functional status or some other metric.


We found that elderly women were three times more likely to have an injury result in a fracture compared with men. This may be a result of factors such as women having a lower bone density and a higher life expectancy than men; unfortunately, the NEISS database does not include bone densitometry data to allow for us to adjust appropriately. It has been previously reported that women are more likely to seek medical care after falls and discuss fall prevention with their healthcare providers, which likely also contributes to this finding.8 Should women be more likely to seek care, this may provide an opportunity to examine the efficacy of fall-prevention education, as measured by the subsequent rate of fractures.


It is curious that few fractures were reported in the minority population, especially considering minorities are generally reported to have higher rates of various conditions that would predispose one to a fall. Perhaps this is a reflection of the decreased life expectancy of minorities, although this does not fully account for our observations; per 2019 government estimates, 76.1% of US seniors are White, compared with 9.5% Black and 7.8% Hispanic, with fewer minorities in each successive age group.14 It has been previously reported that Black and Hispanic patients are more likely to skip various questions on demographic questionnaires, although this is also not at a sufficient rate to make up the demographic gap that we have observed.15 Studies do show that Black patients have a higher bone mineral density compared with White patients, which may also provide a source for decreased fractures in Black patients.16

Furthermore, minority patients undergoing musculoskeletal care may have their pain dismissed by providers and lead to decreased radiographic evaluations for minority patients.17 In a recent study in the pediatric cohort, Marin et al18 studied diagnostic imaging in more than 3 million Emergency Department visits and found that 33.5% of non-Hispanic White patients underwent imaging studies compared with 24.1% of non-Hispanic Black patients. In the adult cohort, Schrager et al controlled for patient and facility-level factors and found that White patients underwent imaging in 51.3% of their encounters compared with 43.6% for Black patients. They additionally noted that Black patients had a lower odds of undergoing a CT scan (odds ratio [OR] = 0.86, 95% CI: 0.84 to 0.89), and Hispanic and Asian patients had increased odds of 1.36 and 1.25 of undergoing an ultrasonography during the encounters, respectively.19 Decreased radiographic imaging conducted on minority patients during their Emergency Department encounters may result in delayed/missed diagnoses of occult fractures that may later be identified during their outpatient follow-up visits.

Black and Hispanic patients were also previously reported as less likely to self-report falls to the Medicare Current Beneficiary Survey.20 These authors were somewhat skeptical of this finding, ascribing their results to self-reporting bias. However, here, we have somewhat objective evidence, albeit still tainted by a very high nonresponse rate to questions of ethnicity and race, that perhaps Black and Hispanic patients are in fact suffering fewer fractures than their White counterparts. Perhaps this is due to the relatively high rates of elderly minorities in nursing homes, which may represent decreased access to adequate home care options but also represents increased supervision and thus perhaps a decreased propensity for falls.21 Regardless, this is a curious finding and warrants additional exploration.


Our study is not without its limitations. One notable weakness in our study is in the inherit study design that being a database study. Database studies rely heavily on accurate documentation, coding, and stratification of important study data that are not available for our review. Without being able to thoroughly review the input data, we rely on each data center to thoroughly input correct and accurate data. The NEISS database only captures those presenting to the Emergency Department. Many of these injuries are somewhat indolent, and thus, the increasing number of outpatient centers and urgent cares may be shifting less severe fractures to the outpatient setting. Without data from ambulatory surgery centers, urgent cares, and outpatient offices, it is difficult to determine whether overall fractures are decreasing or simply moving away from the hospital setting. In addition, an increased emphasis on nonoperative and preventive management of fractures in the elderly has likely contributed to physicians feeling comfortable with entirely outpatient management. As with many retrospective studies, we cannot identify strict cause-and-effect relationships but rather highlight and describe the patterns of our findings. Similarly, we believe the discussion on race is critical, however, given the limited data set we cannot make large assumptions on race, and therefore, additional studies are necessary to extrapolate on racial disparities in care. In addition, a detailed description of the floor materials is not found in the NEISS database. Additional investigation is necessary to identify which particular floor material represents the greatest risk, which would then direct interventions.


We found most causes for fractures among mature adults are related to floors, stairs, or bedding. Most injuries resulted in pelvic, femoral neck, or lower spine fractures and occurred at home. This study offers a preliminary understanding for providers on the common products associated with fractures in mature adults and may be helpful when discussing with patients' preventive measures for lowering fracture risk. We propose that the most effective location of intervention to prevent major fractures in the aging should be directed toward improving the conditions of the floors, bedding, and stairs in the patient's home. However, we also see a clear need for additional investigation into these particular products.


1. Friedman SM, Mendelson DA: Epidemiology of fragility fractures. Clin Geriatr Med 2014;30:175-181.
2. Ferguson TA, Patel R, Bhandari M, Matta JM: Fractures of the acetabulum in patients aged 60 years and older: An epidemiological and radiological study. J Bone Joint Surg Br 2010;92:250-257.
3. Stevens JA, Corso PS, Finkelstein EA, Miller TR: The costs of fatal and non-fatal falls among older adults. Inj Prev 2006;12:290-295.
4. Burge R, Dawson‐Hughes B, Solomon DH, Wong JB, King A, Tosteson A: Incidence and economic burden of osteoporosis-related fractures in the United States, 2005-2025. J Bone Mineral Res 2007;22:465-475.
5. Jang SY, Yang DS, Cha YH, Yoo HJ, Kim KJ, Choy WS: Suicide in elderly patients with hip fracture: A South Korean nationwide cohort study. J Bone Joint Surg Am 2020;102:1059-1065.
6. Fife D, Barancik JI: Northeastern Ohio trauma study III: Incidence of fractures. Ann Emerg Med 1985;14:244-248.
7. Tinetti ME, Speechley M, Ginter SF: Risk factors for falls among elderly persons living in the community. N Engl J Med 1988;319:1701-1707.
8. Stevens JA, Ballesteros MF, Mack KA, Rudd RA, DeCaro E, Adler G: Gender differences in seeking care for falls in the aged Medicare population. Am J Prev Med 2012;43:59-62.
9. Baidwan NK, Naranje SM: Epidemiology and recent trends of geriatric fractures presenting to the emergency department for United States population from year 2004-2014. Public Health 2017;142:64-69.
10. Court-Brown CM, McQueen MM: Global forum: Fractures in the elderly. J Bone Joint Surg Am 2016;98:e36.
11. Albert M, McCaig LF, Ashman JJ: Emergency department visits by persons aged 65 and over: United States, 2009-2010. NCHS data brief, no 130. Hyattsville, MD, National Center for Health Statistics, 2013.
12. Ukkonen M, Jämsen E, Zeitlin R, Pauniaho S-L: Emergency department visits in older patients: a population-based survey. BMC Emerg Med 2019;19:20.
13. Sun R, Karaca Z, Wong HS: Trends in hospital emergency department visits by age and payer, 2006-2015, in Healthcare Cost and Utilization Project (HCUP) Statistical Briefs [Internet]. Rockville, MD, Agency for Healthcare Research and Quality, 2018. Available at:
14. National Population by Characteristics: 2010-2019. United States Census Bureau, 2020. Available at: Accessed June 17, 2020.
15. Dembosky JW, Haviland AM, Haas A, et al.: Indirect estimation of race/ethnicity for survey respondents who do not report race/ethnicity. Med Care 2019;57:e28-33.
16. Ettinger B, Sidney S, Cummings SR, et al.: Racial differences in bone density between young adult black and white subjects persist after adjustment for anthropometric, lifestyle, and biochemical differences. J Clin Endocrinol Metab 1997;82:429-434.
17. Ross AB, Kalia V, Chan BY, Li G: The influence of patient race on the use of diagnostic imaging in United States emergency departments: Data from the national hospital ambulatory medical care survey. BMC Health Serv Res 2020;20:840.
18. Marin JR, Rodean J, Hall M, et al.: Racial and ethnic differences in emergency department diagnostic imaging at US children's hospitals, 2016-2019. JAMA Netw Open 2021;4:e2033710.
19. Schrager JD, Patzer RE, Kim JJ, et al.: Racial and ethnic differences in diagnostic imaging utilization during adult emergency department visits in the United States, 2005 to 2014. J Am Coll Radiol 2019;16:1036-1045.
20. Sairafian K, Towe CW, Crandall M, Brown LR, Haut ER, Ho VP: Sociodemographic patterns of outpatient falls: Do minority patients fall less frequently? J Surg Res 2019;243:332-339.
21. Feng Z, Fennell ML, Tyler DA, Clark M, Mor V: The Care Span: Growth of racial and ethnic minorities in US nursing homes driven by demographics and possible disparities in options. Health Aff (Millwood) 2011;30:1358-1365.

Supplemental Digital Content

Copyright 2022 by the American Academy of Orthopaedic Surgeons.