JBJS | Fractures of the Femoral Shaft in Children. Incidence, Mechanisms, and Sociodemographic Risk Factors*

The Journal of Bone & Joint Surgery, Volume 81, Issue 4

Articles | April 01, 1999

Fractures of the Femoral Shaft in Children. Incidence, Mechanisms, and Sociodemographic Risk Factors*

RICHARD Y. HINTON, M.D., M.P.H., M.ED., P.T.†; ANDREW LINCOLN, M.S.‡; MICHELE M. CROCKETT, B.S.§; PAUL SPONSELLER, M.D.§; GORDON SMITH, M.B., CH.B., M.P.H.‡, BALTIMORE, MARYLAND

View Disclosures and Other Information

Investigation performed at The Johns Hopkins School of Hygiene and Public Health, The Union Memorial Hospital, and The Johns Hopkins Medical Institutions, Baltimore

The Journal of Bone & Joint Surgery. 1999; 81:500-7

5 Recommendations (Recommend) | 3 Comments | Saved by 3 Users Save Case

text A A A

Abstract

Background: Fractures of the femoral shaft in children are caused by major musculoskeletal trauma and result in high direct and indirect medical costs. To date, the American literature has focused on treatment options and outcomes, but the epidemiology of these injuries has been generalized from Scandinavian studies reported in the 1970s and early 1980s. The goals of the current study were (1) to determine the age, gender, and race-specific rates and mechanisms of fractures of the femoral shaft in children in a large United-States-based population and (2) to identify associations between the rates of these fractures and multiple sociodemographic indicators. Such information is vital for preventive efforts.Methods: The Hospital Discharge Database of the Maryland Health Services Cost Review Commission for the years 1990 through 1996 was used to obtain demographic data on 1485 cases of acute fracture of the femoral shaft in patients who were less than eighteen years old, and data from the United States Bureau of the Census for the state of Maryland for the year 1990 were used to obtain denominator data. Reliable external-cause data were available from the 1995 and 1996 databases for 472 patients. Small-area analysis was performed at the zip-code level to determine associations between numerous sociodemographic indicators and the rate of femoral shaft fracture.Results: The annual rate of femoral shaft fracture in children was 19.15 per 100,000. With regard to age, there was a bimodal distribution, with peaks at two and seventeen years. Boys had higher rates of fracture than did girls at all ages, and blacks had higher rates than did whites. The primary mechanisms of fracture were age-dependent and included falls, for children less than six years old; motor vehicle-pedestrian accidents, for those six to nine years old; and motor-vehicle accidents, for teenagers. Firearm-related injuries accounted for 15 percent of the fractures among black adolescents. Adverse socioeconomic conditions were significantly associated with higher rates of fracture.Conclusions: The rates and mechanisms of femoral shaft fractures in children depend on age, gender, and race. For children living in the United States today, the epidemiology of these fractures is different than that described in earlier, Scandinavian reports.

Figures in this Article

Introduction

Introduction | Materials and Methods | Results | Discussion | References

Trauma is the leading cause of death of children in the United States, and major sociodemographic variations in the rates of fatal injury in this population have been reported^{2,11,25,31,32,39}. However, there have been fewer studies of more common, nonfatal injuries in children. The incidence and mechanisms of such injuries cannot be reliably extrapolated from the existing mortality data^8,25; thus, a greater understanding of their epidemiology is required to guide preventive efforts.

Fractures of the femur in children are caused by major musculoskeletal trauma and result in high direct and indirect medical costs^7,18,26. Throughout childhood, changes in physical and social maturation place children at varying risk for femoral shaft fractures due to different mechanisms. Although investigators from a few institutions have reported on the frequency and mechanisms of these fractures^1,14, the United States-based literature has focused largely on treatment options and complications⁴. To date, the epidemiology of femoral shaft fractures in children in the United States has been generalized from Scandinavian studies reported in the 1970s and early 1980s^15,22,40.

The goals of the current study were to determine the age, gender, and race-specific rates and mechanisms of femoral shaft fractures in children in a large United States-based population and to identify, with use of small-area analysis, associations between the rates of fracture and multiple sociodemographic indicators.

*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.

†Department of Orthopaedic Surgery, The Union Memorial Hospital, Johnston Professional Building, 3333 North Calvert Street, Suite 400, Baltimore, Maryland 21218. Please address requests for reprints to Dr. Hinton c/o Elaine P. Bulson, Editor.

‡Center for Injury Research and Policy, The Johns Hopkins School of Hygiene and Public Health, 624 North Broadway, Baltimore, Maryland 21205.

§Department of Orthopaedic Surgery, The Johns Hopkins Medical Institutions, 601 North Caroline Street, Baltimore, Maryland 21287.

*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.

‡Center for Injury Research and Policy, The Johns Hopkins School of Hygiene and Public Health, 624 North Broadway, Baltimore, Maryland 21205.

§Department of Orthopaedic Surgery, The Johns Hopkins Medical Institutions, 601 North Caroline Street, Baltimore, Maryland 21287.

View Larger

TABLE I RATES OF HOSPITALIZATION FROM 1990 THROUGH 1996 FOR CHILDREN IN MARYLAND WHO HAD A FEMORAL SHAFT FRACTURE

*According to United States Bureau of the Census data for 1990³⁷.†The 95 percent confidence interval is given in parentheses.

Category	No. of Fractures	Maryland Population*	Annual Rate of Fracture per 100,000†	Incidence-Rate Ratio†
Gender
Female	412	540,802	10.88 (9.83—11.93)	1.00
Male	1073	566,865	27.04 (25.43—28.65)	2.48 (2.22—2.79)
Race
White	941	766,698	17.53 (16.41—18.65)	1.00
Black	544	340,969	22.79 (20.88—24.70)	1.30 (1.17—1.45)
Age-group
<2 yrs.	238	133,522	25.46 (22.23—28.69)	2.17 (1.79—2.65)
2—5 yrs.	424	274,252	22.09 (19.99—24.19)	1.89 (1.58—2.25)
6—9 yrs.	302	251,372	17.16 (15.22—19.10)	1.47 (1.22—1.77)
10—13 yrs.	188	229,274	11.71 (10.04—13.38)	1.00
14—17 yrs.	333	219,247	21.70 (19.37—24.03)	1.85 (1.54—2.23)
Total	1485	1,107,667	19.15 (18.18—20.12)	—

Add to My JBJS

TABLE I RATES OF HOSPITALIZATION FROM 1990 THROUGH 1996 FOR CHILDREN IN MARYLAND WHO HAD A FEMORAL SHAFT FRACTURE

*According to United States Bureau of the Census data for 1990³⁷.†The 95 percent confidence interval is given in parentheses.

Category	No. of Fractures	Maryland Population*	Annual Rate of Fracture per 100,000†	Incidence-Rate Ratio†
Gender
Female	412	540,802	10.88 (9.83—11.93)	1.00
Male	1073	566,865	27.04 (25.43—28.65)	2.48 (2.22—2.79)
Race
White	941	766,698	17.53 (16.41—18.65)	1.00
Black	544	340,969	22.79 (20.88—24.70)	1.30 (1.17—1.45)
Age-group
<2 yrs.	238	133,522	25.46 (22.23—28.69)	2.17 (1.79—2.65)
2—5 yrs.	424	274,252	22.09 (19.99—24.19)	1.89 (1.58—2.25)
6—9 yrs.	302	251,372	17.16 (15.22—19.10)	1.47 (1.22—1.77)
10—13 yrs.	188	229,274	11.71 (10.04—13.38)	1.00
14—17 yrs.	333	219,247	21.70 (19.37—24.03)	1.85 (1.54—2.23)
Total	1485	1,107,667	19.15 (18.18—20.12)	—

View Larger

TABLE II BIVARIATE ANALYSIS OF ASSOCIATIONS BETWEEN THE RATES OF HOSPITALIZATION, FROM 1990 THROUGH 1996 FOR CHILDREN WHO HAD A FEMORAL SHAFT FRACTURE AND THE SOCIODEMOGRAPHIC CHARACTERISTICS OF THE ZIP-CODE AREA OF THE BALTIMORE, MARYLAND, METROPOLITAN AREA*

*r = Pearson's correlation coefficients.

Indicator	<2-Yr. Age-Group		2-5-Yr. Age-Group		6-9-Yr. Age-Group		10-13-Yr. Age-Group		14-17-Yr. Age-Group		Total
r	p	r	p	r	p	r	p	r	p	r	p
Economic
Median family income	-0.336	0.520	-0.575	0.000	-0.577	0.000	-0.476	0.004	-0.449	0.008	-0.724	0.000
Percentage of children living below poverty level	0.299	0.086	0.576	0.000	0.534	0.001	0.377	0.028	0.516	0.002	0.685	0.000
Social
Percentage of blacks in population	0.218	0.216	0.303	0.082	0.324	0.061	0.344	0.046	0.513	0.002	0.507	0.002
Percentage of house- holds headed by single mother	0.320	0.065	0.558	0.001	0.539	0.001	0.430	0.011	0.551	0.001	0.713	0.000
Educational
Percentage of population 16—19 yrs. old that did not attend or did not complete high school	0.234	0.174	0.479	0.004	0.507	0.002	0.329	0.057	0.314	0.071	0.564	0.001
Percentage of population >25 yrs. old that attended some college	-0.112	0.529	-0.410	0.016	-0.397	0.020	-0.282	0.107	-0.242	0.168	-0.436	0.010
Geographic
Average no. of persons per room	-0.016	0.930	0.306	0.078	0.336	0.052	0.071	0.691	0.590	0.000	0.394	0.021
Percentage of housing structures that were not single-unit	0.414	0.015	0.438	0.010	0.564	0.001	0.430	0.011	0.282	0.106	0.658	0.000
Percentage of housing structures with >20 units	0.195	0.268	0.173	0.329	0.133	0.454	0.145	0.413	0.176	0.319	0.230	0.190

Add to My JBJS

*r = Pearson's correlation coefficients.

Indicator	<2-Yr. Age-Group		2-5-Yr. Age-Group		6-9-Yr. Age-Group		10-13-Yr. Age-Group		14-17-Yr. Age-Group		Total
r	p	r	p	r	p	r	p	r	p	r	p
Economic
Median family income	-0.336	0.520	-0.575	0.000	-0.577	0.000	-0.476	0.004	-0.449	0.008	-0.724	0.000
Percentage of children living below poverty level	0.299	0.086	0.576	0.000	0.534	0.001	0.377	0.028	0.516	0.002	0.685	0.000
Social
Percentage of blacks in population	0.218	0.216	0.303	0.082	0.324	0.061	0.344	0.046	0.513	0.002	0.507	0.002
Percentage of house- holds headed by single mother	0.320	0.065	0.558	0.001	0.539	0.001	0.430	0.011	0.551	0.001	0.713	0.000
Educational
Percentage of population 16—19 yrs. old that did not attend or did not complete high school	0.234	0.174	0.479	0.004	0.507	0.002	0.329	0.057	0.314	0.071	0.564	0.001
Percentage of population >25 yrs. old that attended some college	-0.112	0.529	-0.410	0.016	-0.397	0.020	-0.282	0.107	-0.242	0.168	-0.436	0.010
Geographic
Average no. of persons per room	-0.016	0.930	0.306	0.078	0.336	0.052	0.071	0.691	0.590	0.000	0.394	0.021
Percentage of housing structures that were not single-unit	0.414	0.015	0.438	0.010	0.564	0.001	0.430	0.011	0.282	0.106	0.658	0.000
Percentage of housing structures with >20 units	0.195	0.268	0.173	0.329	0.133	0.454	0.145	0.413	0.176	0.319	0.230	0.190

View Larger

TABLE III MULTIVARIATE ANALYSIS OF ASSOCIATIONS BETWEEN RATES OF HOSPITALIZATION, FROM 1990 THROUGH 1996, FOR CHILDREN WHO HAD A FEMORAL SHAFT FRACTURE AND THE SOCIODEMOGRAPHIC CHARACTERISTICS OF THE ZIP-CODE AREA OF THE BALTIMORE, MARYLAND, METROPOLITAN AREA

*S.E.= standard error.

Indicator	<2-Yr. Age-Group (Adjusted r² = 0.15)			2-5-Yr. Age-Group (Adjusted r² = 0.31)			6-9-Yr. Age-Group (Adjusted r² = 0.33)			10-13-Yr. Age-Group (Adjusted r² = 0.20)			14-17-Yr. Age-Group (Adjusted r² = 0.37)			Total (Adjusted r² = 0.52)
ß	S.E. (ß)	p	ß	S.E. (ß)	p	ß	S.E. (ß)	p	ß	S.E. (ß)	p	b	S.E. (ß)	p	b	S.E. (ß)	p
Economic
Median family income							-0.56	0.36	0.12	-0.51	0.17	0.00				-0.55	0.19	0.01
Percentage of individ- uals <18 yrs. old living below poverty level				0.40	0.10	0.00
Social
Percentage of blacks in population													0.14	0.08	0.09
Geographic
Average no. of persons per room													11.68	4.24	0.01
Percentage of housing structures that were not single-unit	0.47	0.18	0.02				0.22	0.17	0.19							0.12	0.09	0.19
Constant	-5.78	14.13	0.69	9.95	2.87	0.00	21.3	21.8	0.34	24.99	5.21	0.00	-38.00	17.4	0.04	25.9	11.78	0.04

Add to My JBJS

*S.E.= standard error.

Indicator	<2-Yr. Age-Group (Adjusted r² = 0.15)			2-5-Yr. Age-Group (Adjusted r² = 0.31)			6-9-Yr. Age-Group (Adjusted r² = 0.33)			10-13-Yr. Age-Group (Adjusted r² = 0.20)			14-17-Yr. Age-Group (Adjusted r² = 0.37)			Total (Adjusted r² = 0.52)
ß	S.E. (ß)	p	ß	S.E. (ß)	p	ß	S.E. (ß)	p	ß	S.E. (ß)	p	b	S.E. (ß)	p	b	S.E. (ß)	p
Economic
Median family income							-0.56	0.36	0.12	-0.51	0.17	0.00				-0.55	0.19	0.01
Percentage of individ- uals <18 yrs. old living below poverty level				0.40	0.10	0.00
Social
Percentage of blacks in population													0.14	0.08	0.09
Geographic
Average no. of persons per room													11.68	4.24	0.01
Percentage of housing structures that were not single-unit	0.47	0.18	0.02				0.22	0.17	0.19							0.12	0.09	0.19
Constant	-5.78	14.13	0.69	9.95	2.87	0.00	21.3	21.8	0.34	24.99	5.21	0.00	-38.00	17.4	0.04	25.9	11.78	0.04

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 1 Graph showing age and gender-specific rates of hospitalization from 1990 through 1996 for children living in Maryland who had a femoral shaft fracture (total number of cases, 1485).

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 2 Graph showing the age-group, gender, and race-specific rates of hospitalization from 1990 through 1996 for children living in Maryland who had a femoral shaft fracture (total number of cases, 1485).

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 3 Bar graph showing external causes of femoral shaft fracture, according to age-group, from 1995 through 1996 for children living in Maryland (total number of cases, 472). MVA = motor-vehicle accident.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 4 Bar graph showing external causes of femoral shaft fracture, according to age-group, gender, and race, from 1995 through 1996 for children living in Maryland (total number of cases, 472). MVA = motor-vehicle accident.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 5 Scatter plot showing correlations between the rate of femoral shaft fracture and the median annual income in the zip-code areas of the Baltimore metropolitan area from 1990 through 1996.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 6 Scatter plot showing correlations between the rate of femoral shaft fracture and the percentage of households headed by a single mother in the Baltimore metropolitan area from 1990 through 1996.

Materials and Methods

Introduction | Materials and Methods | Results | Discussion | References

The Hospital Discharge Database of the Maryland Health Services Cost Review Commission²³ for the years 1990 through 1996 was used to obtain diagnostic and demographic data on patients with a principal or secondary diagnosis of fracture of the femoral shaft who had been less than eighteen years old when they had been discharged from a nonfederal hospital in Maryland. Unlike most childhood fractures, almost all femoral shaft fractures necessitate hospitalization of some duration, for medical and often social reasons. A few patients in the youngest age-groups may have been managed on a strictly outpatient basis and would not have been included in the discharge database; however, we believe that this small percentage of unaccounted-for outpatients would affect the rates of femoral shaft fracture only slightly.

Definition of the Injury

The International Classification of Diseases, Ninth Revision, codes 821.00, 821.01, 821.10, and 821.11 were used to define fracture of the femoral shaft²⁴. This yielded narrowly defined diagnoses because there are separate codes for fracture of the femur involving the hip, the supracondylar region, the distal part of the femur, and the epiphyseal plates. The inclusion of nonacute fractures and readmissions was minimized with use of discharge database codes to exclude patients who had been seen initially at one hospital and had subsequently been admitted to a specialty center or another hospital. Diagnoses that had been made with use of codes 733.81 (malunion), 733.82 (nonunion), 958 (early complications), and 905—909 (late complications)²⁴ also were not included.

Mechanism of Injury

External-cause (E-code) data were obtained from the Hospital Discharge Database file for each case of femoral shaft fracture for the two years (1995 and 1996) for which reliable data were available. A total of 472 acute fractures were identified. On the basis of the recommendations of the Centers for Disease Control and Prevention⁵, individual mechanism codes were grouped into nine functionally descriptive categories (Fig. 3).

Demographic Data

Information on age, race, gender, and zip code of residence was obtained from the discharge database file. The thirty-four zip-code areas that make up the metropolitan Baltimore area³⁸ (those that are totally or partially within Baltimore city limits) were included in the analysis²⁸. The 1990 United States Census Data^36,37 (Database STF3B) for Maryland were used for population estimates and for sociodemographic profiling of the zip-code areas. Information that was gathered for each zip-code area included the median household income; the percentages of children living below the federal poverty level, of blacks and whites in the population (as defined in the 1990 Census CP-1-22 General Population Characteristics³⁶ [because of small numbers, other racial groups were not included]), of households headed by a single mother, of the population between the ages of sixteen and nineteen years that was not enrolled in or had not graduated from high school, of the population that was more than twenty-five years old and had attended college for some period of time, of housing units that were not single and detached, and of housing structures with more than twenty units; and the number of persons per room.

On the basis of developmental levels and possibly unique mechanisms of femoral shaft fractures, five age-groups were delineated. Group I comprised children who were less than two years old; group II, those two to five years old; group III, those six to nine years old; group IV, those ten to thirteen years old; and group V, those fourteen to seventeen years old.

Data gathered from administrative databases have potential flaws regarding migration of patients within and out of the state, reliability of the diagnoses, patient transfers from other hospitals, and readmissions. These problems have been discussed fully in previous reports^16,17,33 and are unlikely to have had any major effect on the rates presented in this report.

Statistical Analysis

Gender, race, age, and age-group-specific rates of femoral shaft fracture were calculated with use of hospital-discharge data (as the numerator) and 1990 census data (as the denominator). The rates of fracture were calculated in terms of 100,000 person-years, with 95 percent confidence intervals determined with use of the Cornfield approximation³⁵ (Table I). The level of significance was set at p < 0.05. Stata data-analysis software³⁵ was used to determine incidence-rate ratios and corresponding 95 percent confidence intervals for each stratum. The incidence-rate ratio represents the risk of femoral shaft fracture relative to the stratum with the lowest incidence.

Bivariate analyses, performed with use of Pearson's correlation coefficients (SPSS for Windows³⁴), were carried out to determine the relationships between the rates of femoral shaft fracture and the sociodemographic indicators in the various zip-code areas (Table II). Correlations were calculated for the entire study population and for each of the five age-groups to examine the associations between different indicators at different stages of childhood development.

Multivariate analysis, performed with use of forward stepwise linear regressions, was carried out to derive multivariate models of sociodemographic indicators for the entire study population and for each age-group. The criterion for entry into the multivariate analysis was a value of p < 0.20 from the bivariate analyses. Sociodemographic variables were entered in an order based on the magnitude of their correlation coefficients from the bivariate analyses. Accordingly, median income was usually the first entry. Beta coefficients representing the independent effects on the rates of femoral shaft fracture in the zip-code areas due to an increase of one percentage point were calculated for the following sociodemographic indicators: the percentage of individuals less than eighteen years old who were living below the federal poverty level, the percentage of the population that was black, the percentage of housing structures that were not single-unit, a $1000 increase in the annual median income, and an increase of 0.10 in the average number of persons per room within the zip-code area (Table III). (Bivariate variables not included in Table III were not significant contributors in the multivariate analysis.)

Results

Introduction | Materials and Methods | Results | Discussion | References

Use of the Hospital Discharge Database of the Maryland Health Services Cost Review Commission for the years 1990 through 1996²³ led to the identification of 1647 individuals less than eighteen years of age who had a fracture of the femoral shaft. One hundred and sixty-two were excluded from the study because they had had a nonacute fracture or had been admitted to the hospital more than once; this left 1485 children.

The annual rate of femoral shaft fracture was 19.15 per 100,000 (95 percent confidence interval, 18.18 to 20.12). Boys had a higher risk of fracture than did girls (relative risk, 2.49; 95 percent confidence interval, 2.22 to 2.79), and black children had a higher risk than did white children (relative risk, 1.30; 95 percent confidence interval, 1.17 to 1.45) (Table I). Among girls, the rate of fracture was highest in infants less than two years of age, and it decreased between the ages of two and five years. Among boys, the highest rate of fracture was in the two to five-year age-group. Both genders had similar escalating rates from the ages of fourteen through seventeen years (Fig. 1). The higher rates of fracture among blacks were driven by the disproportionately higher relative risk for black male patients compared with their white counterparts. The single highest-risk group was black male patients between the ages of fourteen and seventeen years (relative risk compared with that for whites, 1.68; 95 percent confidence interval, 1.21 to 2.33). Although a significant difference was not detected with the numbers available, black female teenagers had a lower relative risk of fracture than did their white counterparts (relative risk, 0.87; 95 percent confidence interval, 0.44 to 1.32) (Fig. 2).

With regard to the relative distribution of the mechanisms of the fractures, documented abuse accounted for ten (14 percent) of the seventy-three fractures in infants less than two years old, falls were the most common mechanism for children less than six years old, and motor vehicle-pedestrian accidents were the primary external cause for children between the ages of six and nine years old (Fig. 3). Among preadolescents, the most common causes of fracture were falls, motor vehicle-pedestrian accidents, and motor-vehicle accidents. Among teenagers, there was a drastic increase in the rate of fractures that were due to motor-vehicle accidents.

Race and gender-specific external-cause data demonstrated, most noticeably, that black children in all age-groups had a higher rate of fractures caused by motor vehicle-pedestrian accidents and a lower rate of fractures secondary to falls (Fig. 4). The rate of fractures that were due to documented abuse was slightly higher among blacks, and firearm-related fractures were found in black teenagers only (15 percent).

The bivariate analyses showed significant positive associations between the rates of fracture and the variables of low median income, high percentage living below the poverty level, high percentage of households headed by a single mother, and greater crowding of housing structures (Table II). Smaller but still significant positive associations were seen with the variables of lower educational attainment, higher percentage of the population that was black, and greater crowding of people per room. In the age-groups of two to five and six to nine years, the rates of fracture were more sensitive to adverse socioeconomic factors than were those for older children and teenagers. A significant positive correlation between the percentage of the population that was black and the rate of fracture was noted in the ten to thirteen-year-old age-group (r = 0.344, p = 0.046) and in the fourteen to seventeen-year-old age-group (r = 0.513, p = 0.002). Scatter plots for the entire study population revealed strong positive correlations between a lower median income and a higher percentage of households headed by a single mother and increased rates of fracture (Figs. 5 and 6).

According to the multiple regression analysis, a lower socioeconomic level and greater crowding of housing structures were the strongest predictors of fracture. A combination of median family income and housing-structure density accounted for 52 percent of the variance in the rates of fracture among the metropolitan zip-code areas for the entire study population. With each $1000 decrease in median family income, the rate of fracture increased by 0.55 per 100,000. Given the average rate of fracture of 19.15 per 100,000, this represents a 3 percent increase in the rate of fracture for each $1000 decrease in median family income. The annual median income in the Baltimore metropolitan zip-code areas ranged from $11,000 to $49,000; thus, there was the potential for a 108 percent difference in the rates of fracture on the basis of variations in income alone. The rate of fracture for the lowest-income zip-code area was 32.4 per 100,000, and that for the highest-income zip-code area was 4.1 per 100,000. For the age-groups of two to five and six to nine years, the differences in the percentage of children who were living below the federal poverty level and in the median family income accounted for 31 and 33 percent, respectively, of the variation in the rates of fracture among the zip-code areas. For children who were two to five years of age, each single-point increase in the percentage who were living below the federal poverty level resulted in an increase of 0.4 per 100,000, or of 2 percent, in the rate of fracture. In the zip-code area that had the highest percentage of children who were living below the federal poverty level (71.7 percent), the rate of fracture for the two to five-year age-group was 49.6 per 100,000 compared with 8.1 per 100,000 in the zip-code area with the lowest percentage of children (2.2 percent) living below the poverty level.

Changes in the order of entry of the variables into the regression model did not alter the paramount position of median income as a predictor of the rates of fracture. Although race and socioeconomic level are highly associated in urban populations, multivariate analysis revealed that the percentage of blacks living within a zip-code area was an important, independent predictor of the rates of fracture among children fourteen to seventeen years of age. Crowding within housing structures, determined by the average number of people per room in a zip-code area, was also a positive predictor that was unique to this age-group.

Discussion

Introduction | Materials and Methods | Results | Discussion | References

Most descriptive epidemiological studies of injuries in children have dealt with fatal trauma^{2,3,12,32,39,41}. Nonfatal injuries have been loosely categorized as serious, as requiring hospitalization, or as requiring medical attention^8,21,29,30. Orthopaedic studies of fractures in children often have lacked the appropriate population denominator and census-based demographic data necessary to generate true rates rather than frequencies or proportions^1,6,14,43. In the current study, we focused on a single, well defined orthopaedic diagnosis. To our knowledge, this is the first United States-based population study of the incidence, mechanisms, and socioeconomic risk factors for femoral shaft fracture in children. Previously, such information was extrapolated from Scandinavian population studies reported in the late 1970s and early 1980s^22,26; however, children living in the United States today are confronted with different situations and socioeconomic risk factors that may predispose them to injury.

The patterns of occurrence of femoral shaft fractures that are reported in the current study closely resemble previously described patterns of fatal childhood injuries with regard to the bimodal age distribution, the significantly higher rates among the socioeconomically disadvantaged, and the shared primary mechanisms of injury in several age-groups^3,39,42. Motor vehicle-pedestrian accidents and accidents involving motor-vehicle occupants are the leading causes of death of children between the ages of six to nine and fourteen to seventeen years, respectively^3,41. These mechanisms were also the primary causes of femoral shaft fractures in these age-groups. Child abuse is a common cause of fatal injury in children less than two years of age⁴¹. Although, in our study, only 14 percent of the fractures in this age-group were secondary to documented abuse (E codes E9670, E9671, E9679, and E9684), many of the falls may have been due to undetected abuse. Previous work has suggested that abuse is a leading cause of fractures in children less than one year old¹⁴. The higher rates of fracture due to abuse among black infants may be related to socioeconomic factors or to bias in the reporting and documentation of cases.

The overall rate of childhood fracture peaks between the ages of ten and thirteen years^6,40, whereas in the current study the rate of femoral shaft fracture was the lowest during this period. The bimodal age distribution that we noted, with rates that peaked at the ages of two and seventeen years, may be due to a number of factors. Abuse was a mechanism of injury unique to the youngest age-groups. Femoral remodeling is greatest during the first two years of life²⁷. Commonly occurring low-energy mechanisms, such as falls, may not injure an older child but may produce a femoral fracture in an infant or a toddler²⁰. In toddlers, the protective reflexes are not fully developed and physical capabilities outpace judgment skills in assessing potentially dangerous situations and activities¹⁰. Differences between genders with regard to accepted play activities and reinforced behaviors make this relative mismatch more of a problem for boys than for girls⁴². This appears to be reflected by the widened gap in the rates of fracture between boys and girls in the two to five-year age-group.

The second peak in the rate of femoral shaft fracture is probably associated with increased rates of trauma due to motor-vehicle accidents. Although boys have higher rates, the increases in the rates for teenage boys and girls are similar. The finding that the boy:girl ratio for nonfatal motor-vehicle injuries is considerably lower than that reported for fatal accidents suggests that there are gender differences in exposure to injury during vehicular accidents^10,39. For example, girls are more often relatively protected backseat passengers, they use seatbelts more frequently, and they are involved in fewer alcohol-related accidents¹⁰. Although the overall rates of femoral shaft fracture in children in Scandinavian studies are slightly higher than those in our United States-based study (thirty-five per 100,000 compared with twenty-six per 100,000 for boys and fifteen per 100,000 compared with eleven per 100,000 for girls), Scandinavian investigators did not find a second peak among girls¹⁹, suggesting that there are cultural differences in adolescent behavior and exposure to injury.

Black children had a much higher rate of femoral shaft fracture caused by motor vehicle-pedestrian accidents in the present study. Similarly, black children have been reported to have a higher risk than white children for all types of motor vehicle-pedestrian injuries requiring hospitalization^21,29,30. King and Palmisano found an increased risk of such injury among black children even after the implementation of gross controls for socioeconomic status²¹. The differences may have been related to patterns of play, availability of transportation, and housing environments.

Firearm-related violence is another mechanism of injury that disproportionately affects black children, particularly boys. For the black male children between the ages of fourteen and seventeen years in the current study, the rate of fracture due to gun-related violence was nearly half that due to the more common mechanism of motor-vehicular trauma. Racial disparity was greatest in this age-group. This finding is in agreement with those of previous studies on adolescent mortality, which found that black male teenagers were exposed to disproportionately greater risks of injury^2,11,12. Although black female teenagers also had some gun-related fractures, their overall rate of femoral shaft fracture was actually slightly lower than that of their white counterparts (Fig. 2). This protective effect may be due to lower exposure to motor-vehicular trauma or high-energy sports activities. For example, several fractures in the white female teenagers were related to horseback riding, whereas this mechanism was not associated with the fractures in the black children. Baker et al. found that adolescents living in more densely populated, inner-city areas had a substantially lower risk of mortality due to motor-vehicle accidents than did those living in more rural areas³. Black girls also have slightly higher bone density and reach skeletal maturity sooner than do age-matched white girls^9,13. The role of gender-related differences in the rates of injury among black teenagers is an interesting area that deserves additional research in terms of potentially causal behavioral and environmental factors.

In summary, in the current study we detailed the epidemiology of femoral shaft fractures in children as well as the host, agent, and environment-related factors that contribute to increased risk of such fractures. Our findings should serve as the basis for preventive programs aimed at reducing the occurrence and consequences of these common childhood injuries²⁵.

NOTE: The authors thank Monique Sheppard, formerly of the Maryland Department of Health and Mental Hygiene; Gregg Burleyson of the Department of Case Mix Management at The Johns Hopkins Hospital for supplying the databases for this study; and Elaine Bulson for her assistance in project development.

References

Introduction | Materials and Methods | Results | Discussion | References

Anderson, W. A.: The significance of femoral fractures in children. Ann. Emerg. Med.,11: 174-177, 1982.11174 1982 [PubMed]

Baker, S. P.; Whitfield, R. A.; and O'Neill, B.: Geographic variations in mortality from motor vehicle crashes. New England J. Med.,316: 1384-1387, 1987.3161384 1987

Baker, S. P.; O'Neill, B.; Ginsberg, M. J.; and Li, G.: Appendix. In The Injury Fact Book. Ed. 2, pp. 297-320. New York, Oxford University Press, 1992.

Beaty, J. H.: Femoral-shaft fractures in children and adolescents. J. Am. Acad. Orthop. Surgeons,3: 207-217, 1995.3207 1995

Centers for Disease Control and Prevention: Recommended framework for presenting injury mortality data. Morbid. and Mortal. Weekly Rep., 46: 1-30, 1997.

Cheng, J. C., and Shen, W. Y.: Limb fracture pattern in different pediatric age groups: a study of 3,350 children. J. Orthop. Trauma,7: 15-22, 1993.715 1993 [PubMed]

Coyte, P. C.; Bronskill, S. E.; Hirji, Z. Z.; Daigle-Takacs, G.; Trerise, B. S.; and Wright, J. G.: Economic evaluation of 2 treatments for pediatric femoral shaft fractures. Clin. Orthop.,336: 205-215, 1997.336205 1997 [PubMed]

Durkin, M. S.; Davidson, L. L.; Kuhn, L.; O'Connor, P.; and Barlow, B.: Low-income neighborhoods and the risk of severe pediatric injury: a small-area analysis in northern Manhattan. Am. J. Pub. Health,84: 587-592, 1994.84587 1994

Ettinger, B.; Sidney, S.; Cummings, S. R.; Libanati, C.; Bikle, D. D.; Tekawa, I. S.; Tolan, K.; and Steiger, P.: Racial differences in bone density between young adult black and white subjects persist after adjustment for anthropometric, lifestyle, and biochemical differences. J. Clin. Endocrinol. and Metabol.,82: 429-434, 1997.82429 1997

Evans, L.: Driver behavior. In Traffic Safety and the Driver, pp. 133-161. New York, Van Nostrand Reinhold, 1991.

Fingerhut, L. A.; Ingram, D. D.; and Feldman, J. J.: Firearm and nonfirearm homicide among persons 15 through 19 years of age. Differences by level of urbanization, United States, 1979 through 1989. J. Am. Med. Assn.,267: 3048-3053, 1992.2673048 1992

Fingerhut, L. A.: Firearm mortality among children, youth, and young adults 1—34 years of age, trends and current status: United States, 1985—90. Advance Data,231: 1-17, 1993.2311 1993 [PubMed]

Gilsanz, V.; Roe, T. F.; Mora, S.; Costin, G.; and Goodman, W. G.: Changes in vertebral bone density in black girls and white girls during childhood and puberty. New England J. Med.,325: 1597-1600, 1991.3251597 1991

Gross, R. H., and Stranger, M.: Causative factors responsible for femoral fractures in infants and young children. J. Pediat. Orthop.,3: 341-343, 1983.3341 1983

Hedlund, R., and Lindgren, U.: The incidence of femoral shaft fractures in children and adolescents. J. Pediat. Orthop.,6: 47-50, 1986.647 1986

Hinton, R. Y., and Smith, G. S.: The association of age, race, and sex with the location of proximal femoral fractures in the elderly. J. Bone and Joint Surg.,75-A: 752-759, May 1993.75-A752 1993

Hinton, R. Y.; Lennox, D. W.; Ebert, F. R.; Jacobsen, S. J.; and Smith, G. S.: Relative rates of fracture of the hip in the United States. Geographic, sex, and age variations. J. Bone and Joint Surg.,77-A: 695-702, May 1995.77-A695 1995

Hughes, B. F.; Sponseller, P. D.; and Thompson, J. D.: Pediatric femur fractures: effects of spica cast treatment on family and community. J. Pediat. Orthop.,15: 457-460, 1995.15457 1995

Jacobs, J. T.: Achilles tenodesis for paralytic calcaneocavus foot. Clin. Orthop.,47: 143-149, 1966.47143 1966 [PubMed]

Kasser, J. R.: Femoral shaft fractures. In Fractures in Children, edited by C. A. Rockwood, Jr., K. E. Wilkins, and J. H. Beaty. Ed. 4, pp. 1195-1230. Philadelphia, Lippincott-Raven, 1996.

King, W. D., and Palmisano, P. A.: Racial differences in childhood hospitalized pedestrian injuries. Pediat. Emerg. Care,8: 221-224, 1992.8221 1992

Landin, L. A.: Fracture patterns in children. Analysis of 8,682 fractures with special reference to incidence, etiology and secular changes in a Swedish urban population 1950—1979. Acta Orthop. Scandinavica, Supplementum 202, 1983.

Maryland Health Services Cost Review Commission Hospital Discharge Data Base 1990—1996. Baltimore, Maryland Health Services Cost Review Commission, 1997.

Medicode: Fracture of lower limbs. In 1998 Physician ICD-9-CM. International Classification of Diseases, Ninth Revision, Clinical Modification. Ed. 5, pp. 201-203. Salt Lake City, Medicode Publications, 1998.

National Committee for Injury Prevention and Control: Injury prevention: meeting the challenge. Am. J. Prevent. Med.,5: 1-303, 1989.51 1989

Newton, P. O., and Mubarak, S. J.: Financial aspects of femoral shaft fracture treatment in children and adolescents. J. Pediat. Orthop.,14: 508-512, 1994.14508 1994

Ogden, J. A.; Ganey, T. M.; and Ogden, D. A.: The biological aspects of children's fractures. In Fractures in Children, edited by C. A. Rockwood, Jr., K. E. Wilkins, and J. H. Beaty. Ed. 4, pp. 19-52. Philadelphia, Lippincott-Raven, 1996.

Pittman, J., and Andrews, H.: The use of zip coded population data in social area studies of service utilization. Eval. Prog. Plan.,9: 309-317, 1986.9309 1986

Rivara, F. P., and Barber, M.: Demographic analysis of childhood pedestrian injuries. Pediatrics,76: 375-381, 1985.76375 1985 [PubMed]

Rivara, F. P.: Child pedestrian injuries in the United States. Current status of the problem, potential interventions, and future research needs. Am. J. Dis. Child.,144: 692-696, 1990.144692 1990 [PubMed]

Rodriguez, J. G.: Editorial. Childhood injuries in the United States. A priority issue. Am. J. Dis. Child.,144: 625-626, 1990.144625 1990 [PubMed]

Rutledge, R.; Smith, C. Y.; and Azizkhank, R. G.: A population-based multivariate analysis of the association of county demographic and medical system factors with per capita pediatric trauma death rates in North Carolina. Ann. Surg.,219: 205-210, 1994.219205 1994 [PubMed]

Smith, G. S.; Langlois, J. A.; and Buechner, J. S.: Methodological issues in using hospital discharge data to determine the incidence of hospitalized injuries. Am. J. Epidemiol.,134: 1146-1158, 1991.1341146 1991 [PubMed]

SPSS: SPSS Advanced Statistics 7.5. Chicago, SPSS, 1997.

Stata Press: STATA Reference Manual, Release 4, pp. m-z. College Station, Texas, Stata Press, 1997.

United States Bureau of the Census: 1990 Census of Population and Housing Guide. Washington, D.C., United States Bureau of the Census, 1991.

United States Bureau of the Census: 1990 United States Census Lookup: STF3B, 1999.

United States Postal Service: Postal Zip Code Map: Baltimore, Maryland Delivery Area & Near-by Post Offices, 1997.

Waller, A. E.; Baker, S. P.; and Szocka, A.: Childhood injury deaths: national analysis and geographic variations. Am. J. Pub. Health,79: 310-315, 1989.79310 1989

Wilkins, K. E.: The incidence of fractures in children. In Fractures in Children, edited by C. A. Rockwood, Jr., K. E. Wilkins, and J. H. Beaty. Ed. 4, pp. 3-17. Philadelphia, Lippincott-Raven, 1996.

Wilson, M. H.; Baker, S. P.; Teret, S. P.; Shock, S.; and Garbarino, J.: The injury problem. In Saving Children: A Guide to Injury Prevention, pp. 3-13. New York, Oxford University Press, 1991.

Wilson, M. H.; Baker, S. P.; Teret, S. P.; Shock, S.; and Garbarino, J.: Pedestrians. In Saving Children: A Guide to Injury Prevention, pp. 57-67. New York, Oxford University Press, 1991.

Worlock, P., and Stower, M.: Fracture patterns in Nottingham children. J. Pediat. Orthop.,6: 656-660, 1986.6656 1986

Topics

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 1 Graph showing age and gender-specific rates of hospitalization from 1990 through 1996 for children living in Maryland who had a femoral shaft fracture (total number of cases, 1485).

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 5 Scatter plot showing correlations between the rate of femoral shaft fracture and the median annual income in the zip-code areas of the Baltimore metropolitan area from 1990 through 1996.

View Large | Download Slide(.ppt)
Add to My JBJS

View Larger

TABLE I RATES OF HOSPITALIZATION FROM 1990 THROUGH 1996 FOR CHILDREN IN MARYLAND WHO HAD A FEMORAL SHAFT FRACTURE

*According to United States Bureau of the Census data for 1990³⁷.†The 95 percent confidence interval is given in parentheses.

Category	No. of Fractures	Maryland Population*	Annual Rate of Fracture per 100,000†	Incidence-Rate Ratio†
Gender
Female	412	540,802	10.88 (9.83—11.93)	1.00
Male	1073	566,865	27.04 (25.43—28.65)	2.48 (2.22—2.79)
Race
White	941	766,698	17.53 (16.41—18.65)	1.00
Black	544	340,969	22.79 (20.88—24.70)	1.30 (1.17—1.45)
Age-group
<2 yrs.	238	133,522	25.46 (22.23—28.69)	2.17 (1.79—2.65)
2—5 yrs.	424	274,252	22.09 (19.99—24.19)	1.89 (1.58—2.25)
6—9 yrs.	302	251,372	17.16 (15.22—19.10)	1.47 (1.22—1.77)
10—13 yrs.	188	229,274	11.71 (10.04—13.38)	1.00
14—17 yrs.	333	219,247	21.70 (19.37—24.03)	1.85 (1.54—2.23)
Total	1485	1,107,667	19.15 (18.18—20.12)	—

Add to My JBJS

TABLE I RATES OF HOSPITALIZATION FROM 1990 THROUGH 1996 FOR CHILDREN IN MARYLAND WHO HAD A FEMORAL SHAFT FRACTURE

*According to United States Bureau of the Census data for 1990³⁷.†The 95 percent confidence interval is given in parentheses.

Category	No. of Fractures	Maryland Population*	Annual Rate of Fracture per 100,000†	Incidence-Rate Ratio†
Gender
Female	412	540,802	10.88 (9.83—11.93)	1.00
Male	1073	566,865	27.04 (25.43—28.65)	2.48 (2.22—2.79)
Race
White	941	766,698	17.53 (16.41—18.65)	1.00
Black	544	340,969	22.79 (20.88—24.70)	1.30 (1.17—1.45)
Age-group
<2 yrs.	238	133,522	25.46 (22.23—28.69)	2.17 (1.79—2.65)
2—5 yrs.	424	274,252	22.09 (19.99—24.19)	1.89 (1.58—2.25)
6—9 yrs.	302	251,372	17.16 (15.22—19.10)	1.47 (1.22—1.77)
10—13 yrs.	188	229,274	11.71 (10.04—13.38)	1.00
14—17 yrs.	333	219,247	21.70 (19.37—24.03)	1.85 (1.54—2.23)
Total	1485	1,107,667	19.15 (18.18—20.12)	—

View Larger

*r = Pearson's correlation coefficients.

Indicator	<2-Yr. Age-Group		2-5-Yr. Age-Group		6-9-Yr. Age-Group		10-13-Yr. Age-Group		14-17-Yr. Age-Group		Total
r	p	r	p	r	p	r	p	r	p	r	p
Economic
Median family income	-0.336	0.520	-0.575	0.000	-0.577	0.000	-0.476	0.004	-0.449	0.008	-0.724	0.000
Percentage of children living below poverty level	0.299	0.086	0.576	0.000	0.534	0.001	0.377	0.028	0.516	0.002	0.685	0.000
Social
Percentage of blacks in population	0.218	0.216	0.303	0.082	0.324	0.061	0.344	0.046	0.513	0.002	0.507	0.002
Percentage of house- holds headed by single mother	0.320	0.065	0.558	0.001	0.539	0.001	0.430	0.011	0.551	0.001	0.713	0.000
Educational
Percentage of population 16—19 yrs. old that did not attend or did not complete high school	0.234	0.174	0.479	0.004	0.507	0.002	0.329	0.057	0.314	0.071	0.564	0.001
Percentage of population >25 yrs. old that attended some college	-0.112	0.529	-0.410	0.016	-0.397	0.020	-0.282	0.107	-0.242	0.168	-0.436	0.010
Geographic
Average no. of persons per room	-0.016	0.930	0.306	0.078	0.336	0.052	0.071	0.691	0.590	0.000	0.394	0.021
Percentage of housing structures that were not single-unit	0.414	0.015	0.438	0.010	0.564	0.001	0.430	0.011	0.282	0.106	0.658	0.000
Percentage of housing structures with >20 units	0.195	0.268	0.173	0.329	0.133	0.454	0.145	0.413	0.176	0.319	0.230	0.190

Add to My JBJS

*r = Pearson's correlation coefficients.

Indicator	<2-Yr. Age-Group		2-5-Yr. Age-Group		6-9-Yr. Age-Group		10-13-Yr. Age-Group		14-17-Yr. Age-Group		Total
r	p	r	p	r	p	r	p	r	p	r	p
Economic
Median family income	-0.336	0.520	-0.575	0.000	-0.577	0.000	-0.476	0.004	-0.449	0.008	-0.724	0.000
Percentage of children living below poverty level	0.299	0.086	0.576	0.000	0.534	0.001	0.377	0.028	0.516	0.002	0.685	0.000
Social
Percentage of blacks in population	0.218	0.216	0.303	0.082	0.324	0.061	0.344	0.046	0.513	0.002	0.507	0.002
Percentage of house- holds headed by single mother	0.320	0.065	0.558	0.001	0.539	0.001	0.430	0.011	0.551	0.001	0.713	0.000
Educational
Percentage of population 16—19 yrs. old that did not attend or did not complete high school	0.234	0.174	0.479	0.004	0.507	0.002	0.329	0.057	0.314	0.071	0.564	0.001
Percentage of population >25 yrs. old that attended some college	-0.112	0.529	-0.410	0.016	-0.397	0.020	-0.282	0.107	-0.242	0.168	-0.436	0.010
Geographic
Average no. of persons per room	-0.016	0.930	0.306	0.078	0.336	0.052	0.071	0.691	0.590	0.000	0.394	0.021
Percentage of housing structures that were not single-unit	0.414	0.015	0.438	0.010	0.564	0.001	0.430	0.011	0.282	0.106	0.658	0.000
Percentage of housing structures with >20 units	0.195	0.268	0.173	0.329	0.133	0.454	0.145	0.413	0.176	0.319	0.230	0.190

View Larger

*S.E.= standard error.

Indicator	<2-Yr. Age-Group (Adjusted r² = 0.15)			2-5-Yr. Age-Group (Adjusted r² = 0.31)			6-9-Yr. Age-Group (Adjusted r² = 0.33)			10-13-Yr. Age-Group (Adjusted r² = 0.20)			14-17-Yr. Age-Group (Adjusted r² = 0.37)			Total (Adjusted r² = 0.52)
ß	S.E. (ß)	p	ß	S.E. (ß)	p	ß	S.E. (ß)	p	ß	S.E. (ß)	p	b	S.E. (ß)	p	b	S.E. (ß)	p
Economic
Median family income							-0.56	0.36	0.12	-0.51	0.17	0.00				-0.55	0.19	0.01
Percentage of individ- uals <18 yrs. old living below poverty level				0.40	0.10	0.00
Social
Percentage of blacks in population													0.14	0.08	0.09
Geographic
Average no. of persons per room													11.68	4.24	0.01
Percentage of housing structures that were not single-unit	0.47	0.18	0.02				0.22	0.17	0.19							0.12	0.09	0.19
Constant	-5.78	14.13	0.69	9.95	2.87	0.00	21.3	21.8	0.34	24.99	5.21	0.00	-38.00	17.4	0.04	25.9	11.78	0.04

Add to My JBJS

*S.E.= standard error.

Indicator	<2-Yr. Age-Group (Adjusted r² = 0.15)			2-5-Yr. Age-Group (Adjusted r² = 0.31)			6-9-Yr. Age-Group (Adjusted r² = 0.33)			10-13-Yr. Age-Group (Adjusted r² = 0.20)			14-17-Yr. Age-Group (Adjusted r² = 0.37)			Total (Adjusted r² = 0.52)
ß	S.E. (ß)	p	ß	S.E. (ß)	p	ß	S.E. (ß)	p	ß	S.E. (ß)	p	b	S.E. (ß)	p	b	S.E. (ß)	p
Economic
Median family income							-0.56	0.36	0.12	-0.51	0.17	0.00				-0.55	0.19	0.01
Percentage of individ- uals <18 yrs. old living below poverty level				0.40	0.10	0.00
Social
Percentage of blacks in population													0.14	0.08	0.09
Geographic
Average no. of persons per room													11.68	4.24	0.01
Percentage of housing structures that were not single-unit	0.47	0.18	0.02				0.22	0.17	0.19							0.12	0.09	0.19
Constant	-5.78	14.13	0.69	9.95	2.87	0.00	21.3	21.8	0.34	24.99	5.21	0.00	-38.00	17.4	0.04	25.9	11.78	0.04