To evaluate the occurrence of emergency department (ED) visits due to humerus fractures in the US.
To evaluate the occurrence of emergency department (ED) visits due to humerus fractures in the US.
We analyzed the 2008 Nationwide Emergency Department Sample, which contained approximately 28 million ED records. We identified the cases of interest using diagnostic codes for proximal, shaft, and distal humerus fractures.
In 2008, approximately 370,000 ED visits in the US resulted from humerus fractures. Proximal humerus fractures were the most common, accounting for 50% of humerus fractures. The incidence rate of proximal humerus fractures followed the shape of an exponential function in the age groups 40–84 years for women (R2 = 97.9%) and 60–89 years for men (R2 = 98.2%). After the exponential increase in these age intervals, the growth rate of proximal humerus fracture slowed and eventually decreased. The peak occurrence of distal humerus fractures was in children ages 5–9 years; however, elderly women had an increased risk. As the baby boomer generation ages, unless fracture prevention programs improve, more than 490,000 ED visits due to humerus fractures are expected in 2030 when the youngest of the baby boomers turn age 65 years.
Compared to epidemiologic studies in Japan and European countries, the incidence rates of humerus fractures are substantially higher in the US. The high incidence rate of humerus fractures in the expanding elderly population may contribute to the recent trend of rapid increase in shoulder arthroplasty in the US. Rigorous safety measures to reduce falls and improved preventive treatments of osteoporosis are needed.
Humerus fractures have a substantial impact on personal function and well-being and are one of the leading causes of excessive mortality among the elderly (1–3). A humerus fracture is often caused by a fall directly on the shoulder or arm (4), and the proximal part of the humerus fracture is commonly encountered in patients with osteoporosis. Published epidemiologic studies have reported widely diverse incidence rates of humerus fracture (5–7). Whereas one study has reported that humeral fractures occur predominantly in those ages <50 years (5), another study has shown that the highest incidence occurs in patients ages >50 years (6). One study reported a bimodal age distribution of humeral shaft fractures (6), while another found that the majority of humeral shaft fractures occurred among people ages <35 years (7). Conflicting results in these studies likely reflect wide variation among samples. To date, to our knowledge there has been no published study on humerus fractures based on a representative sample in the US; consequently, the annual cases (and incidence rates) of humerus fractures remain largely unknown. Without such data, it is difficult to plan and monitor the effectiveness of measures to prevent humerus fractures.
The objective of this study was to evaluate the annual incidence rate of emergency department (ED) visits for humerus fractures in the US. Furthermore, we wanted to identify the demographic characteristics of people at higher risk for these fractures as well as elucidate the mechanism of injury. In the absence of a registry system to track all fractures in the population, we believe that ED visits best reflect the incidence of humerus fractures because such injuries require immediate medical attention. We also provided here the future demands of ED care for humerus fractures in the US relevant for aging baby boomers. Compared with 2008, the sector of the population age ≥65 years will grow by 32.8 million in 2030 (from 38.7 million to 71.5 million) (8). Although it is well understood that the number of osteoporotic fractures will increase as the elderly population expands, the impact of this expansion on the number of humerus fractures is not well described. We projected the number of humerus fractures in 2030, when the youngest of the baby boomers will turn age 65 years. Patients and physicians may consider these data in making public health and clinical decisions on osteoporosis management.
This is the first representative study of emergency department visits with humerus fractures in the US.
The age- and sex-specific incidence rates for the proximal, shaft, and distal humerus fractures could be used to plan and monitor the effectiveness of measures to prevent humerus fractures.
We projected the number of humerus fractures in 2030, when the youngest of the baby boomers will turn age 65 years.
Patients and physicians may consider these data in making public health and clinical decisions on osteoporosis management.
We studied the annual incidence of humerus fractures in the US on the basis of the Nationwide Emergency Department Sample (NEDS). The institutional review board (IRB) of our institution determined this study as “Not human subject research. No IRB review is needed.”
NEDS, a part of the Healthcare Cost and Utilization Project, is a database of hospital-based ED visit reports conducted annually under the auspices of the Agency for Healthcare Research and Quality (AHRQ). The objective of NEDS is to produce national estimates of ED visits in the US. In 2008, probability samples of 28 million ED visits were selected from a total of 125 million ED visits. Nearly 1,000 stratified samples of EDs were selected from the 2008 American Hospital Association Annual Survey. All data obtained from uniform billing (UB-04 form) records from the selected EDs were included in the NEDS database. More details concerning sampling procedures have been published elsewhere (9).
We identified humerus fracture cases using the diagnosis coding from the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM): 812.0 and 812.1 for proximal humerus, 812.2 and 812.3 for humeral shaft, and 812.4 and 812.5 for distal humerus. The diagnostic code is based on the patient medical record as documented by clinicians. A hospital's medical records department assigns ICD-9-CM codes by reviewing patient medical records to identify all pertinent diagnoses and procedures performed. When an injury is the result of an external cause, an external cause code (E-code) is noted in addition to the injury code. On the basis of the E-code, we further described the most common mechanisms of injury for each site of humerus fracture.
Because the NEDS database utilizes a probabilistic sample survey scheme, the sampling weight and sampling design were considered in estimating the total number of cases and their SEs. Sampling weights, provided by the AHRQ, were used to account for the unequal sampling probabilities and to produce estimates for all EDs in the US. We estimated the SE using Taylor linearization. As suggested by the National Center for Health Statistics, a relative SE in excess of 30% is considered to be unreliable. Whenever appropriate, a 95% confidence interval was reported along with a point estimate.
The incidence rate, also referred to as the cumulative incidence rate, is a measure of the risk of developing a new condition during a specified time period (10). When the incidence (or occurrence) is divided by the number of people at risk, the incidence becomes the incidence rate (11), as follows:
In this study, the number of new events was the number of humerus fracture cases that occurred during 2008. The population at risk is frequently estimated using the midinterval population during the specified time period (11), and we used the population estimate as of July 1, 2008. The base rate we used was 105, which is a frequently used base rate in published studies. The annual incidence rate in this study indicates the risk of an individual sustaining a humerus fracture within a 1-year period. To determine the demographic-specific annual incidence rates of humerus fractures, we calculated the number of ED visits in 2008 for every 100,000 people in each respective demographic group on the basis of the US Census (12).
To reflect the nonlinear growth of humerus fracture risk with respect to increasing age, we fitted an exponential equation to the data from which the percentage increase of fracture risk was calculated. The exponential function fitted was y = α exp(βx), where y = the incidence rate of humerus fracture, x = the numerical age, and exp (approximately 2.718) = the base of natural logarithms. The constant “α” is the vertical stretch factor, whereas the constant “β” is the horizontal stretch factor. The fit of the model was evaluated by the coefficient of determination and the residual plot.
The purpose of providing a future projection in this study was to estimate the impact of an aging population; therefore, the projection was solely based on the demographic changes projected by the US Census. Because the incidence rate of humeral fractures among Medicare patients in the US has not changed over time (13), we applied the 2008 age- and sex-specific fracture incidence rates to the projected US population size in 2030 (8). The baby boomer generation will be nearly age 65–84 years in 2030. Previously, a study reported a 2–3-fold larger fracture risk for white Americans ages >65 years (14). Compared with 2008, the proportion of white Americans ages >65 years is expected to decrease; therefore, we also adjusted for race distribution on the basis of the US Census projection (8). We applied the age- and sex-specific fracture rates from 2008 to the projected population size of whites in the US in 2030. For the nonwhite elderly, we applied one-third of age- and sex-specific fracture rates.
In 2008, approximately 370,000 ED visits occurred in the US as a result of patients with humerus fractures (Table 1). The most common site of fracture was the proximal humerus, accounting for 50% of humerus fractures. The next most common site of fracture was the distal humerus. Less than 1% of patients experienced fractures at more than 1 anatomic part of the humerus. The female to male fracture ratio varied by fracture site: 2.3:1 for the proximal humerus, 1.3:1 for the humeral shaft, and 0.9:1 for the distal humerus.
|Anatomic site||%||Rate*||Frequency||95% confidence interval for frequency|
To estimate the burden of humerus fractures in the ED, we first estimated the occurrence of ED visits by age group (in 5-year intervals) without an adjustment to the population size. As shown in Figure 1, the peak incidences for the proximal, shaft, and distal humerus fractures varied by age. Compared to the overall trend, a larger number of proximal humerus fractures were observed in baby boomers in the 45–64 years age group. The number of ED visits with proximal humerus fractures increased continuously until age 84 years. A relatively small number of cases (approximately 11,300 cases) occurred in those ages ≥90 years (Figure 1). This was mainly due to the small number of people in this age group in the US (the population size of this age group was 2,150,000, or 0.7% of the US population).
The distal part of the humerus was the second most common anatomic site of fracture, with a peak incidence in children ages 5–9 years. Children ages <15 years had 84,000 ED visits with distal humerus fractures, accounting for 64% of all distal humerus fractures. The number of ED visits with distal humerus fractures decreased quickly after age 15 years (Figure 1). Unlike other parts of the humerus, the occurrence of shaft fractures was relatively infrequent and remained approximately the same for all ages.
As noted, the annual incidence rate of humerus fracture is the risk of an individual sustaining a humerus fracture within a 1-year period. Taking into account the population of the respective groups in the US (12), we calculated age- and sex-specific fracture rates to estimate the risk of fractures. Overall, nearly 122 visits per 100,000 people in the US were made for humerus fractures. The risk for all types of humerus fractures was low for both men and women in the 20–44 years age group (Figure 2).
For proximal sites, the rates of humerus fractures (per 100,000 people) increased continuously until age 89 years, and this pattern was particularly pronounced in women. The rate of proximal humerus fractures followed the shape of an exponential function in the age groups 40–84 years for women (R2 = 97.9%) and 60–89 years for men (R2 = 98.2%). The fitted lines are as follows: for women, proximal humerus fracture rate = 18.857 × exp(0.3792 × age group), and for men, proximal humerus fracture rate = 31.021 × exp(0.3596 × age group).
Among women, the humerus fracture rate (per 100,000) at age 40–44 years was 28 and the rate of fracture increased by 46% for every 5-year increase in age. That is, compared with women ages 40–44 years, the risk of sustaining a proximal humerus fracture was nearly 5 times greater for women ages 60–64 years and 21 times greater for women ages 80–84 years. At age 90 years, the growth rate of proximal humerus fracture slowed and eventually decreased. The numbers of proximal humerus fractures in women and men ages ≥70 years were 424 and 150 (per 100,000), respectively. Among men, the incidence rate of proximal humerus fracture at age 60 years was 44 and the rate of fracture increased by 43% for every 5-year increase in age.
For the distal humerus, the rate of fractures also showed the highest peak in children ages 5–9 years, with nearly 185 and 200 ED visits per 100,000 girls and boys, respectively (Figure 2). Although this injury is relatively uncommon among adults, there is an increasing trend of distal humerus fractures as people age as evidenced by approximately 21 and 47 ED visits per 100,000 men and women ages ≥65 years. Likewise, the fracture rate of shaft fractures increased as people age.
The combination of a larger elderly population and the increased risk of humerus fractures for this age group may result in a tremendous increase in the total number of humerus fractures (Figure 3). Assuming no change in age- and sex-specific incidence, more than 500,000 ED visits due to humerus fractures can be expected (projection A, Figure 3). Of these, approximately 275,000 will be due to proximal humerus fractures, which represent a 50% increase as compared with an incidence of 184,000 found in 2008. When we assume that the humerus fracture risk for nonwhites will remain at one-third that of whites (14), more than 490,000 ED visits due to humerus fractures can be expected in 2030 (projection B, Figure 3).
The overwhelming mechanism for all types of fractures of the humerus was a fall, which accounted for 88% of the injuries. The next most common mechanisms of injury were motor vehicle accidents (8%) and injury from striking or being struck by an object or person (5%).
To our knowledge, this is the first representative study of ED visits with humerus fractures in the US. A number of important findings emerged in this study. Although the incidence rate of humeral fractures among elderly patients in the US is not increasing (13), the rate of humerus fractures in the US is considerably higher than that of countries with a larger proportion of elderly persons in the population. Epidemiologic studies suggest that the annual incidence rates of humerus fractures in the US are higher than those in Japan and European countries (15–22) (Table 2). Finland is considered to be a country with high fragility fracture rates (23). However, among women ages ≥80 years, the proximal humerus fracture rate (per 100,000) was 298 in Finland (19) compared with 565 in the US. The US population is young by the standards of developed countries (24). Considering the US population is younger than that of Japan and the European countries listed in Table 2 (24), the US incidence rate will remain higher than that of these countries, even if differences in age distribution are accounted for.
|Author, year (ref.)||Origin of study||Data year||Age of patients, years||Sex||Annual rate||US rate*|
|Hagino et al, 1999 (15)||Japan||1995||≥35||Male||17||43|
|Oinuma et al, 2010 (16)||Japan||2006||≥65||All||38||259|
|Sakuma et al, 2008 (17)||Japan||2004||≥50||Male||9||73|
|Pentek et al, 2008 (18)||Hungary||1999–2003||≥50||Male||47||73|
|Kannus et al, 2009 (19)||Finland||2007||≥80||Female||298||565|
|Palvanen et al, 2010 (20)||Finland||2007||≥60||Female||25||42|
|Ekholm et al, 2006 (21)||Sweden||1998–1999||≥16||All||14.5||20|
|Van Staa et al, 2001 (22)||England and Wales||1988–1998||≥20||Male||53||74|
As baby boomers age, an expanding elderly population will inevitably result in an increase in the number of humerus fractures, the number of ED visits, and the demands for operation and inpatient care. A recent study reported a rapid increase in shoulder arthroplasty in the US (25). As baby boomers begin reaching age 65 years in 2011, the high proximal humerus fracture rate among the expanding elderly population is likely to continuously increase the demand for shoulder arthroplasty. With an expected nearly 275,000 ED visits for proximal humerus fractures in 2030, the increase in the demand for shoulder arthroplasties is inevitable. Rigorous safety measures to reduce falls and preventive treatments of osteoporosis are needed in this age group.
The most common mechanism of a fracture was falling. The increase in the rate of injurious falls is inevitably associated with a higher risk of proximal humerus fractures (26, 27). It is not clear whether Americans live in a more hazardous environment with respect to falling. Further research is needed to better explain the higher rate of humerus fractures in the US as compared to the rate in other developed countries. A number of strategies have been recommended for fall prevention in the elderly (28), including making the home safer by removing hazards, increasing physical activity, installing more light in living spaces, using assistive devices, educating about the side effects of medication, and wearing sensible shoes. A clinical trial showed that a multifactorial prevention program significantly reduced the odds of falls in the elderly (29).
Despite the incidence of falls in men being more frequent than in women among community-dwelling older adults (30), it was clear that women experienced more humerus fractures (per population size) and this finding is consistent with those of other studies (17, 18, 22, 31). As low bone mineral density is the strongest predictor of humerus fractures, we believe that this statistic is due to the high prevalence of osteoporosis among elderly women. Indeed, in a case–control study of proximal humerus fractures, Chu et al found that decreased bone mass and low dietary calcium were risk factors for proximal humeral fractures, whereas menopausal hormone replacement and calcium supplementation reduced the risk of fracture (4). There are likely still many unknown physical, dietary, and genetic factors to consider that may become more apparent with population studies. As we learn more about osteoporosis, we need to develop strategies to identify individuals who are at high risk for proximal humerus fragility fractures, quantify fracture risk, and develop prevention and treatment algorithms in order to better manage this exponentially increasing problem.
Nearly two-thirds of distal humerus fractures occurred among children ages <15 years in our study. In general, ED visits with elbow fractures reflect hazards in the environment of children; therefore, distal fractures could be preventable to some degree. The most common pediatric elbow fracture is the supracondylar fracture and it is well understood that the majority of these elbow injuries occur at playgrounds when children are playing on the monkey bars (32). In an effort to improve playground safety, various prevention strategies have been suggested (or required by law) over the last couple of decades, i.e., lowering the monkey bar height, installing an impact-absorbing surface underneath the bars, adult supervision, etc. The degree of implementation and success rates are unknown, yet nearly 39,000 children ages 5–9 years are still visiting the ED with distal humerus fractures every year.
The strength of this study lies in the large sample size of the NEDS database, which constitutes more than 28 million visits selected with known probability. Despite using a probabilistic sampling technique, the large sample size yields reliable statistics with small SEs. A study reported that E-codes reported in hospital discharge data are a reliable source of information on the mechanism of injury (33); however, the accuracy of ICD-9-CM codes submitted by each hospital is unknown. Although large databases are prone to some degree of error based on the accuracy of data entry and diagnostic coding, these databases are still valuable for insight into large-scale population effects. The NEDS database contains event-level records but no patient-level records. A patient-level analysis of humerus fracture injuries is not allowed in this data set. Although we believe that ED visits best reflect the incidence of fractures, we recognize that not all cases of shoulder and arm fractures were represented in the ED. Making inferences regarding the incidence of humerus fractures based on ED visits could be biased; therefore, our study should be understood as presenting the minimum estimation of humerus fractures in the population. Although there is no evidence that the incidence rate of humeral fractures in the US is changing (13), we want to note that our projection was based on data from a single year. In interpreting this projection, it is important to recognize the uncertainty inherent in the calculation process. To illustrate the impact of changing demographics, the projection of future cases of humerus fracture was based on the projected elderly population in 2030. Future improvement in preventing fractures is another important factor that may decrease the occurrence of humerus fractures but is not reflected in the projection.
In conclusion, this study documents for the first time the incidence of ED visits for humerus fractures in the US. Regardless of the anatomic site of humerus injury, the most common mechanism of a fracture was falling. The risk of proximal humerus fractures, frequently attributed to osteoporosis, is growing exponentially as people get older. Distal humerus fractures were most common among children and the risk decreased rapidly after age 15 years. Shaft humerus fractures occurred relatively infrequently, but the rate of occurrence increased with age. The rates of proximal, distal, and shaft humerus fractures were considerably higher compared with those of other countries. As the baby boomers age, an expanding elderly population will inevitably increase the number of humerus fractures and ED visits. Rigorous safety measures from injurious falls and preventive treatments of osteoporosis are needed.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Kim had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Kim, Szabo, Marder.
Acquisition of data. Kim.
Analysis and interpretation of data. Kim, Szabo, Marder.
The authors thank Dr. Betty Guo at the University of California, Davis, School of Medicine, Office of Research for commenting on an earlier draft of the manuscript.