Abstract
This study investigated the epidemiology of children treated in US emergency departments for radial head subluxation (RHS) associated with consumer products and recreational activities using data from the National Electronic Injury Surveillance System. An estimated 430 766 (95% confidence interval: 341 194-520 339) children ≤5 years of age were treated for RHS in US emergency departments from 1990 to 2011. The mean patient age was 2.1 years, and 56.5% of patients were girls. The most common mechanism of injury was a fall (43.2%), followed by arm pull (39.4%). The annual rate of RHS increased significantly by 190.1% during the 22-year study period. This is the largest study of RHS to date and the first to use a nationally representative sample to investigate secular trends and mechanisms of injury for RHS associated with consumer products and recreational activities. Events associated with consumer products and recreational activities are an important increasing source of RHS.
Introduction
Radial head subluxation (RHS), also known as annular ligament displacement, nursemaid’s elbow, or pulled elbow, is a common musculoskeletal disorder among young children. RHS was first described by Fournier in 1671; 1 however, references to it can be found in the records of Hippocrates. 1 Van Arsdale 2 conducted one of the first in-depth studies of this condition in 1889 based on 100 cases; he noted some of the defining characteristics of the injury, including children’s age and gender, and commented on possible mechanisms of injury. In 1971, Salter and Zaltz 3 performed classic anatomical studies on 12 child cadavers and demonstrated that traction on the pronated forearm caused RHS.
Anatomically, the annular ligament wraps around the radial head to support and keep it in place as the forearm is supinated or pronated. In the event of a subluxation, the radial head slips out from its normal position, allowing the ligament to slide into the gap between the radial head and the capitellum of the humerus. This prevents the radial head from returning to its normal position, causing immobility and pain. 3 Young children are predisposed to RHS because of annular ligament laxity and a weaker distal attachment of the ligament to the periosteum of the radial neck.3,4 Diagnosis of RHS is based on history and physical examination. Radiographs are used only to exclude more serious injuries, such as a fracture, when the diagnosis is unclear. 5 RHS is treated using either a supination or pronation maneuver to reposition the radial head and annular ligament. 6
Research on RHS has been devoted to evaluating the epidemiology, best treatment maneuvers, and waiting time in the emergency department (ED).5,7-9 Epidemiologic research has been based primarily on data from single EDs, although there has been 1 national study covering a 2-year period.7,9-11 The association of consumer products and recreational activities (and the related injury mechanisms) with RHS has not been previously investigated. Because children in the age group at risk for RHS spend most of their time in the home, where consumer product-related injuries commonly occur, this represents an important gap in our understanding of the epidemiology of RHS. To our knowledge, this is the largest study of RHS to date and the first to use a nationally representative sample to investigate secular trends and mechanisms of injury for RHS associated with consumer products and recreational activities.
Methods
Case Selection and Data Sources
Data were obtained from the National Electronic Injury Surveillance System (NEISS) for children 5 years and younger treated for RHS in US EDs from January 1, 1990, to December 31, 2011. Patients were included in this study if they sustained an injury to the elbow (NEISS body region code 32) and were diagnosed with a dislocation (NEISS diagnosis code 55) or if the NEISS narrative contained the terms “nursemaid’s elbow” or “radial head subluxation.” 12 NEISS coders are instructed by the US Consumer Product Safety Commission (CPSC) to code RHS as a “dislocation” of the elbow. 12 The NEISS narratives of all of these patients were then reviewed, and patients were excluded if they had an injury not consistent with a RHS, such as fractures with true dislocations. Thirty patients were excluded because they were admitted for inpatient care, which was judged to not be consistent with RHS; these patients also generally had high-energy mechanisms of injury. The final study dataset contained 15 141 actual cases, from which national estimates were calculated using sample weights provided by the CPSC.
The CPSC maintains the NEISS to monitor injuries associated with consumer products and sports and recreational activities treated in US EDs. The NEISS receives data from a network of approximately 100 hospitals, representing a stratified probability sample of the more than 5300 hospitals in the US and its territories with a 24-hour ED with at least 6 beds.13,14 Patients’ ED medical records are reviewed by professional NEISS coders, and data are entered and uploaded into the NEISS database.
The US Census Bureau’s July 1 intercensal and postcensal population estimates for children 5 years of age and younger from 1990 through 2011 were used to calculate RHS rates. 15
Study Variables
The NEISS database includes variables for patient age, gender, body region injured, injury diagnosis, location where injury occurred, and ED disposition, in addition to a short narrative describing the injury event. NEISS case narratives were used to categorize the mechanism of injury into the following: (1) fell down on outstretched arm/wrist; (2) fell on arm/wrist/shoulder; (3) fell, not otherwise specified; (4) grabbed by the arm in an attempt to prevent an injury; (5) lifted from a seated position; (6) arm pulled to move child; (7) arm pulled while putting on/taking off clothing; (8) object forcibly removed from hand; (9) swung by hands; (10) pull, not otherwise specified; (11) struck elbow on object; (12) arm wedged between 2 objects; (13) arm twisted or bent backward; (14) rolled over; and (15) other/unspecified. For analyses, the mechanism of injury was grouped into the following: (1) fall (mechanism codes 1-3), (2) pull (mechanism codes 4-10), (3) struck elbow (mechanism code 11), (4) twisted/entrapped (mechanism codes 12-13), (5) rolled over (mechanism code 14), and (6) other/unspecified (mechanism code 15). Disposition from the ED was categorized as (1) treated and released (patients who were treated and released or examined and released without treatment) and (2) left against medical advice. The location where the injury occurred was categorized as (1) home (including home, apartment/condo, and manufactured/mobile home) and (2) other (including school, farm, and other public property).
Statistical Analysis
Data were analyzed using SAS 9.3 (SAS Institute Inc, Cary, NC) statistical software. To account for the NEISS sampling design, complex survey procedures were used to calculate national estimates and the Taylor series linearization method was used to calculate the variance of the estimates. All estimates reported in this study are stable estimates unless stated otherwise. An estimate was judged as potentially unstable if the estimate was <1200 cases, the sample size was <20 cases, or the coefficient of variation was >30%. Trend analyses were performed using weighted linear regression with weights equal to the inverse of the variance of the estimated statistics. The estimated annual change in the number or rate from the regression analyses is denoted by m in this article. Other statistical analyses included the Rao-Scott chi-square test for association and calculation of relative risks (RRs) with 95% confidence intervals (CIs). Statistical significance was determined at the level α = .05. This study was approved by the Institutional Review Board of the Research Institute at Nationwide Children’s Hospital.
Results
Overall Trends and Demographics
An estimated 430 766 (95% CI = 341 194-520 339) patients ≤5 years were treated for RHS in US EDs from 1990 through 2011, which accounted for 1.1% of all nonfatal consumer product-related and sports and recreational activity-related ED visits by this age group during that period. On average, 19 580 (95% CI = 15 509-23 652) patients were treated annually for RHS during the 22-year study period, and the annual number of cases increased significantly (m = 783.4, P < .001) by 212.1%, from 9809 (95% CI = 5349-14 269) in 1990 to 30 616 (95% CI = 18 707-42 525) in 2011 (Figure 1). The annual rate of RHS per 10 000 children ≤5 years increased significantly (m = 0.31, P < .001) by 190.1% during the study from 4.35 (95% CI = 2.37-6.33) in 1990 to 12.62 (95% CI = 7.71-17.54) in 2011. Most of this increase was accounted for by rate increases among 1- and 2-year-olds, who represented 67.3% of RHS cases. RHS events occurred mostly at home (84.8%) (Table 1).
Annual number and rate of children 5 years of age and younger treated in US emergency departments for radial head subluxation by gender and year, National Electronic Injury Surveillance System 1990-2011.
Characteristics of Radial Head Subluxation Among Children 5 Years of Age and Younger Treated in US Emergency Departments, National Electronic Injury Surveillance System 1990-2011.
Percentages may not sum to 100.0% due to rounding error.
Potentially unstable estimate due to sample size <20 cases, estimate <1200 cases, or coefficient of variation >30%. No CI is given.
Mechanism of Injury
The most common mechanism associated with RHS was falls (43.2%), followed by pulls (39.4%), and struck elbow (5.2%) (Table 1). The annual rate of RHS per 10 000 children ≤5 years associated with pulls increased significantly by 245.1% (m = 0.16, P < .001) from 1.58 (95% CI = 0.67-2.49) in 1990 to 5.45 (95% CI = 3.14-7.75) in 2011 (Figure 2). Similarly, the annual rate of RHS due to falls increased significantly by 165.6% (m = 0.12, P < .001) from 1.99 (95% CI = 1.13-2.86) in 1990 to 5.29 (95% CI = 3.19-7.39) in 2011. Because of unstable annual rates, no trend subanalyses were performed for the other mechanism categories.
Annual rate of radial head subluxation among children 5 years of age and younger treated in US emergency departments by mechanism of injury and year, National Electronic Injury Surveillance System 1990-2011.
Age and Gender
The mean age of patients treated for RHS was 2.1 (standard deviation = 0.017) years with a median of 1.9 (interquartile range = 1.5-2.1) years (Figure 3). RHS was most common among 1- and 2-year-olds, accounting for 32.8% and 34.5% of all cases, respectively (Table 1). During the study period, the annual rate of RHS per 10 000 children for 1- and 2-year-olds increased significantly by 210.3% (m = 0.64, P < .001) and 183.1% (m = 0.69, P < .001), respectively (Figure 4). For 3-year-olds, the annual rate increased significantly by 105.3% (m = 0.24, P < .001) from 1991 through 2011 (1990 was excluded because the estimated rate was potentially unstable). No trend analyses were performed for other age groups because of potentially unstable rate estimates. Patients <1 year of age were more likely to experience RHS associated with rolling over (RR = 6.30; 95% CI = 4.02-9.87) or arms being twisted or entrapped between 2 objects (RR = 1.99; 95% CI = 1.56-2.54) than older patients.
Number and rate of children 5 years of age and younger treated in US emergency departments for radial head subluxation by age, National Electronic Injury Surveillance System 1990-2011.
Annual rate of radial head subluxation among children 5 years of age and younger treated in US emergency departments by age and year, National Electronic Injury Surveillance System 1990-2011.
Girls accounted for 56.5% of patients treated for RHS and had a higher rate of RHS than boys (9.6 [95% CI = 7.62-11.58] per 10 000 girls compared with 7.1 [95% CI = 5.58-8.55] per 10 000 boys). During the study period, the rate of RHS for boys and girls increased significantly by 169.1% (m = 0.25, P < .001) and 208.0% (m = 0.39, P < .001), respectively (Figure 1). Because of the greater rate increase among girls, there was a nonsignificant increase in the proportion of girls to boys from 1.17 in 1990 to 1.34 in 2011.
Associated Consumer Products and Activities
Among the consumer products and recreational activities associated with RHS, the top 17 categories accounted for 75.1% of the cases (Table 2). Beds or bed frames (16.0%), floors (13.4%), sofas or couches (10.7%), stairs (8.8%), chairs (4.9%), and slides (3.3%) were the top 6 products associated with RHS. Beds or bed frames was the category most frequently associated with RHS related to falls, and stairs was the category most frequently associated with the mechanism of pulling the arm.
Top 17 Consumer Products or Sport/Recreational Activities Associated With Radial Head Subluxation Among Children 5 Years of Age and Younger Treated in US Emergency Departments, National Electronic Injury Surveillance System 1990-2011.
Percentages may not sum to 100.0% due to rounding error.
Potentially unstable estimate due to sample size <20 cases, estimate <1,200 cases, or coefficient of variation >30%. No CI is given.
Includes consumer products and sport/recreational activities that each account for less than 1% of the total number of radial head subluxation cases.
Discussion
There were an estimated 430 766 children ≤5 years treated for RHS associated with consumer products and recreational activities in US EDs from 1990 through 2011, averaging 19 580 patients annually. This equals an average of 54 patients per day or 1 child every 27 minutes. By 2011, this had increased to 84 patients per day or 1 child every 17 minutes. Therefore, events associated with consumer products and recreational activities clearly are an important source of RHS. Patients one and two years of age accounted for the majority of the cases with a mean age of 2.1 years, which is consistent with previous studies.6,8,9,11,16 The female predominance (56.4%) seen in this study has also been described previously.6-9,11,16 There also appears to be a trend during the past 2 decades that reveals an increase in the proportion of girls to boys. The reason that RHS is more commonly seen among girls is unclear, but may include child gender-related differences in care-seeking behavior by parents, or that young girls more frequently have their hands held or are more frequently picked up when they fall due to societal traditions.
Contrary to previous studies, the most prevalent mechanism associated with RHS in this study was falls (43%), followed by pulls (39%). There are several possible explanations for the higher proportion of falls observed in this study, including that only events associated with consumer products and recreational activities were included in this study. In addition, because the mechanism of injury is based on parental report, it is possible that parents misattributed the injury to the fall. For example, if a child fell, started crying, was then picked up by the hands by the parent, and subsequently would not use an arm, the parent may assume that the condition was associated with the fall, when it was the action of picking up the child that caused a RHS and not the fall. As Van Arsdale 2 speculated, “In about one third of all the cases a fall was said to be the cause of the injury. But on closer inquiry, [Van Arsdale] generally found that the mothers retracted this statement.” Another possibility is that clinicians in previous studies dismissed reported falls as a cause of RHS because they were taught that pulling on the arm was the accepted causal mechanism. Errors in documentation and coding in the ED medical record or NEISS database are a third possible explanation for our findings.
In their classic study, Salter and Zaltz 3 were able to produce RHS by applying traction to the pronated forearms of 9 child cadavers younger than 5 years, but not in 3 child cadavers 5 to 9 years of age. Based on their findings and a review of the literature, they concluded that this was the mechanism of injury. Several subsequent studies of RHS further support their conclusion. In these studies, a pull on the arm was identified as the leading mechanism with percentages ranging from 50.6% to 93%.6-8,10,16 However, Salter and Zaltz 3 did not report testing of other possible mechanisms besides direct traction of the forearm in pronation. Therefore, their findings do not exclude the possibility of other untested mechanisms, such as a fall. RHS associated with a fall has been reported in previous studies with the percentage of cases ranging from 7% to 21.8%.6,8,10,16
Based on this large national study, we conclude that our findings support the existence of other mechanisms that cause RHS in addition to traction on a pronated forearm. In particular, falls appear to be an important causal mechanism among cases of RHS associated with consumer products and recreational activities. When young children fall forward, they often extend their arms out in front of them with the palms of the hands downward in an attempt to break their fall. This orientation places the forearms in pronation. We hypothesize that an impact in this orientation may be adequate to result in RHS. This proposed injury mechanism could be tested with child postmortem human subjects as an extension of the work of Salter and Zaltz. 3
This study is the first to describe an increase in the number and rate of RHS. This increase has been substantial, equaling approximately 190% during the 22-year study period. The factors contributing to this increase are not known. However, one study observed a relationship between RHS and children being overweight. 9 This observation suggests that the rise in the prevalence of overweight and obesity among young US children during recent decades may have contributed to the increased incidence of RHS observed in this study. According to the National Health and Nutrition Examination Survey, the prevalence of obesity among children 2 to 5 years of age increased from 7.2% during the 1988-1994 survey period to 12.1% for the 2009-2010 survey period, before decreasing to 8.4% during the 2011-2012 survey period. 17 This study identified falls and arm pulls as the mechanisms accounting for the increase in RHS during the past 2 decades. This is consistent with the association of obesity and RHS because increased body mass would result in increased traction force as a child is lifted by the hand and could contribute to an increase in impact force in a fall. Other factors also may have contributed to the increase in RHS observed in this study, including the increased use of EDs by parents for medical care for this condition. It is unlikely that the recognition and diagnosis by clinicians and the documentation in the ED medical record and the NEISS has changed substantially during the past two decades. Overall, it is similarly unlikely that changes in parental care-seeking behavior, diagnosis, and documentation could account for the almost three-fold increase in the incidence of RHS observed in this study. Further research to evaluate the secular trend in RHS incidence that includes all cases of RHS, and not just cases associated with consumer products and recreational activities, is warranted to corroborate these findings.
The American Academy of Pediatrics recommends that children be picked up by holding the body and lifting from under the arms, instead of lifting by the hands or wrists. It also recommends that parents abstain from swinging children by the hands. 12 Our study supports these recommendations, as well as exercising caution when removing clothing, such as a shirt or pullover. Preventing falls in this age group is difficult as they challenge their new-found mobility; however, many types of falls, for example from furniture and nursery products, can be reduced. To the extent that obesity predisposes to RHS, this offers another reason for maintenance of a healthy weight among young children.
Study Limitations
This study has several limitations. This study underestimates the true number of children with RHS because the NEISS only includes individuals treated in EDs and does not include children treated in urgent care facilities, clinics, or private practitioner offices, nor those who do not receive medical attention for their RHS. In addition, the NEISS only captures RHS associated with consumer products and sports/recreational activities, and therefore, the cases reported in this study are an underestimate and may not be representative of all RHS cases. Incomplete and variable amounts of detail in NEISS case narratives may have led to misclassification of the mechanism of injury in some cases. Mechanism of injury was based on parental report, and the retrospective study design precluded verification of documented details. Although our findings are ecologically consistent with the previously described association of obesity and RHS, this relationship could not be confirmed by our study. Despite these limitations, this is the largest study of RHS to date and the first to use a nationally representative sample to investigate secular trends and mechanisms of injury for RHS associated with consumer products and recreational activities.
Conclusions
Events associated with consumer products and recreational activities are an important increasing source of RHS. The rate of RHS increased significantly during the study period. This increase was attributable to increases in the number of cases caused by falls and arm pulls, which is consistent with the previously described association of child obesity and RHS. Furthermore, study findings support the importance of falls as a cause of RHS. This study also corroborated previous reports that RHS is more common in girls than boys, and revealed that there is an increasing trend in the proportion of girls to boys, although this did not achieve statistical significance. These study findings contribute importantly to our understanding of the epidemiology of RHS and point to future areas of needed research.
Author Contributions
RW conducted the data analysis, drafted and revised the manuscript, and approved the final manuscript. TC in data analysis, revised the manuscript, and approved the final manuscript. GAS conceptualized the study, assisted in data analysis, critically reviewed and revised the manuscript, and approved the final manuscript.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.