 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
ARTICLE |
a Departments of Health Policy and Management
b Epidemiology
c Injury Free Coalition for Kids, Mailman School of Public Health
d Columbia University Center for the Health of Urban Minorities, Columbia University, New York, New York
e Department of Surgery, Columbia University at Harlem Hospital, New York, New York
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
METHODS. Injury analyses used national vital statistics data from January 1, 1981, to December 31, 2003, that were available from the Centers for Disease Control and Prevention. Rate calculations and 
2 test for trends (Mantel extension) used data that were collapsed into 3-year intervals to produce cell sizes with stable estimates. Percentage change for mortality rate ratios used the earliest (1981–1983) and the latest (2001–2003) study period for black, American Indian/Alaskan Native, and Asian/Pacific Islander children, with white children as the comparison group.
RESULTS. All-cause injury rates declined during the study period, but current mortality ratios for all-cause injury remained higher in black and American Indian/Alaskan Native children and lower in Asian/Pacific Islander children compared with white children. Trend analyses within racial groups demonstrate significant improvements in all groups for unintentional but not intentional injury. Black and American Indian/Alaskan Native children had higher injury risk as a result of residential fire, suffocation, poisoning, falls, motor vehicle traffic, and firearms. Disparities narrowed for residential fire, pedestrian, and poisoning and widened for motor vehicle occupant, unspecified motor vehicle, and suffocation for black and American Indian/Alaskan Native children.
CONCLUSIONS. These findings identify injury areas in which disparities narrowed, improvement occurred with maintenance or widening of disparities, and little or no progress was evident. This study further suggests specific mechanisms whereby new strategies and approaches to address areas that are recalcitrant to improvement in absolute rates and/or narrowing of disparities are needed and where increased dissemination of proven efficacious injury prevention efforts to high-risk populations are indicated.
Key Words: infants • children • health disparities • injury • injury prevention • race • trends
Abbreviations: ICD—International Classification of Diseases
Both intentional and unintentional injuries have a modifiable component when well-focused interventions are implemented.1–6 Injury prevention efforts of varying scope and intensity have targeted an array of potentially fatal injury mechanisms in the past 50 years, some directly aimed at reducing historically existing racial and ethnic disparities in residential fire, poisonings, pedestrian injury, motor vehicle traffic, suffocation, homicide, and others.7 When legislative or regulatory efforts have been combined with access to safety devices and parental education programs, notable successes have been reported in resource-limited neighborhoods.1–6,8–16 Although increasingly more effective methods have evolved to address excess mortality and serious morbidity that are associated with injury, application of this knowledge, access to safety devices, changes in behavior, and achieved rates of safety practices may be lower in resource-limited communities. To the extent that newly engineered products and modifications to behavior, the home, community, and environment are effective at preventing death from injury, lack of access or delayed implementation may be reflected in higher injury mortality and lowered life expectancy.17,18
Although examination of individual communities is important for assessment of how well interventions work, examination of national injury mortality trends is essential for assessment of the impact of how fully and effectively injury prevention efforts are penetrating at-risk populations. The value in conducting such analyses is that the findings may provide clues to population subgroups and/or injury mechanisms for which additional or redirected national efforts are indicated. This study examined >20 years of injury-related mortality for infants and young children by race for all-cause, unintentional, and intentional injury across specific injury mechanisms for which local and/or national interventions of varying intensities have been attempted.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Classification of Intent and Mechanism of Death
Classification of ICD-9 and ICD-10 coding for unintentional, intentional, undetermined intent, and specific mechanisms of death used standard definitions from the Centers for Disease Control and Prevention.19 Death as a result of undetermined intent accounted for a small number of deaths and is included in all-cause injury analyses only.
Unintentional motor vehicle traffic deaths were analyzed for the following categories: all motor vehicle traffic, occupant, pedestrian, and unspecified. Classifications of death that were attributable to other unintentional injury mechanisms in this study include drowning, residential fire/flame, suffocation, poisoning, and falls. Firearm injury includes all intents.
Transition From ICD-9 to ICD-10
Injuries were classified using ICD-9 for deaths that occurred between 1981 and 1998 and ICD-10 coding for 1999 to 2003.19,20 Homogeneous ICD classifications were maintained within groups by collapsing only years of the same ICD classification scheme.
Classification of Race and Ethnicity
Race classification was analyzed similarly throughout the study period for white, black, American Indian/Alaskan Native, and Asian/Pacific Islander children. Data on Hispanic origin were not available before 1990. Sensitivity analyses, performed after Hispanic origin became available, examined the robustness of rates that were calculated with the inclusion and exclusion of Hispanic ethnicity from each racial group and found small or negligible effects across most injury mechanisms. 2 test for trend is presented for Hispanic ethnicity with the data on which trend analyses are based beginning in 1990 rather than 1981. For consistency in graphic illustration of trends, Hispanic ethnicity is included as an addition to the graphs without adjustment of racial categories established in 1981.
Statistical Analyses
Mortality rates (per 100000 persons) used 3-year time intervals to create larger cell sizes and provide more stable numerical estimates. Linear trend in proportions used the 2 test for trends Mantel extension, available in Epi Info 3.2.2,23 to analyze trends in mortality rates within race and ethnicity with adjustment for the last uneven interval. Mortality rate ratios, used to examine the relative change in minority groups, compared the earliest (1981–1983) and the latest (2001–2003) study period for black, American Indian/Alaskan Native, and Asian/Pacific Islander children, with white children as the comparison group. All rates and rate ratios are for homogeneous ICD versions (ICD-9 or ICD-10, but not mixed within time periods) with 95% confidence intervals. The 2 test was used in univariate analyses of categorical variables. Statistical significance is defined as having a P 
.05.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
Unintentional and Intentional Injury Rates in 2003
Although the unintentional injury rate (per 100000) was 13.7 for the total population, rates varied widely by race, from a low of 5.9 in Asian/Pacific Islander children to a high of 28.0 in American Indian/Alaskan Native children and 20.6 in black children. Injury rates among Hispanic and white children were intermediate (11.1 and 13.0). Similarly, the intentional injury rate (per 100000) for the total population was 3.6 but was lowest in Asian/Pacific Islander children (1.9), highest in black (8.5) and American Indian/Alaskan Native children (8.1), and again intermediate in white (2.4) and Hispanic children (3.4).
Relative Rank of Injury Among All Causes of Death in 2003
The relative rank of injury among all causes of death differed significantly between those who were aged 0 to <1 year and 1 to 4 years. Injury was the leading cause of death in children who were aged 1 to 4 years across all race/ethnic groups, with all-cause injury (intentional and unintentional) accounting for >40% of all deaths in 2003, except among Asian/Pacific Islander children, for whom injury comprised 27.9% of all deaths. Among infants who were aged 0 to 12 months, unintentional injury ranked sixth among overall causes of death behind congenital anomalies, short gestation, sudden infant death syndrome, pregnancy complications, and placental conditions. Although there was considerable variation in the distribution of injury mechanisms across racial groups in infants who were aged 0 to 12 months, 
90% of all injury-related mortality was accounted for by 4 injury mechanisms: suffocation, motor vehicle traffic, drowning, and residential fire/burn.
Trends in Injury Mortality by Intent, Mechanism, and Race: 1981–2003
Trends in All-Cause Injury Mortality
All-cause injury rates declined significantly for all racial and ethnic groups examined (2 trend in white children: 1393.2 [P < .000001]; black children: 572.6 [P < 000001]; American Indian/Alaskan Native children: 66.4 [P < .000001]; Asian/Pacific Islander children: 50.0 [P < .000001]; and Hispanic children: 127.9 [P < .0000001]). Most of the observed improvement was attributed to declines in unintentional (Fig 1A) rather than intentional injury (Fig 1B). Despite declines in all-cause injury for all minority groups (data not shown), declines also occurred in the white comparison group, leaving mortality rate ratios for black and American Indian/Alaskan Native children relatively unchanged and nearly twice that of white children (Table 2).
|
|
80%. Asian/Pacific Islander children were notable for consistently having the lowest unintentional injury rates during the study time frame (Fig 1A).
Trends in Intentional Injury Mortality
Little progress was made in net improvement of intentional injury rates in 2003 compared with 1981. Small declines were observed for mortality rates among Asian/Pacific Islander and black children, with American Indian/Alaskan Native children showing wide fluctuations without improvement (Fig 1B). Black children showed steadily rising intentional mortality rates in the 1980s, with rates peaking in the early 1990s. Both Hispanic and black children showed improvement after 1990 (Fig 1B). Paradoxically, although mortality rate ratios declined slightly (Table 2), this was partially attributable to worsening rates in the white comparison group. When 20-year trends are examined, white children showed significant worsening, whereas black, American Indian/Alaskan Native, and Asian/Pacific Islander children showed insignificant 20-year trends in violence-related deaths (Fig 1B).
Trends in Motor Vehicle Traffic and Pedestrian Mortality
Motor vehicle traffic–related mortality rates were highest in black and American Indian/Alaskan Native children, lowest in Asian/Pacific Islander children, and intermediate for Hispanic and white children (Fig 2). Trend analyses demonstrated significantly declining 20-year injury rates for total motor vehicle traffic, occupant, and pedestrian injury deaths among all race and ethnic groups examined (Fig 2). Improvements in motor vehicle pedestrian traffic deaths continued for all race and ethnic groups during the latest study time frame (Fig 2C), whereas a leveling off/worsening occurred in the category of motor vehicle occupant mortality (Fig 2B) for all groups except Asian/Pacific Islander children. Additional examination of trends during the last 2 study periods for occupant deaths in infants (aged 0 to <12 months) reveal that non-Hispanic white, Hispanic, Asian/Pacific Islander, and Indian/Alaskan Native children showed a tendency toward continued improvement, whereas improvement in black children slowed. Occupant mortality among young children aged 1 to 4 years showed a tendency toward increased mortality in black, Hispanic, and American Indian/Alaskan Native children.
|
Drowning Mortality
Drowning rates improved significantly during the study period in all racial and ethnic groups examined, with the largest absolute improvement observed among American Indian/Alaskan Native children (Fig 3A). Although American Indian/Alaskan Native children continued to have the highest drowning rates among the racial groups examined, disparities in mortality rate ratios narrowed during the study time frame (Table 2). Despite improvements in white children outpacing those of black and Asian/Pacific Islander children, white children maintained their position as having the second highest mortality from drowning (Fig 3A).
|
Suffocation Mortality
Among children who were aged 0 to 4 in 2003, unintentional suffocation ranged between 13.9% of all injury-related mortality in Hispanic children to 26.0% in black children (Table 1). The relative racial disparities that were observed during the early study periods were maintained during the last study period, with American Indian/Alaskan Native and black children exhibiting higher rates than white children and Asian/Pacific Islander and Hispanic children having lower rates than white children (Fig 3C; Table 2). Mortality as a result of suffocation showed no significant trend toward improvement in any race or ethnic group and actually demonstrated a trend toward worsening in later study years (Fig 3C).
The majority of unintentional suffocation deaths were in the group aged 0 to 12 months, in which unintentional suffocation accounted for nearly two thirds of all injury-related deaths. In this group, suffocation as a proportion of injury-related deaths was lowest in Asian/Pacific Islander children (33.3%); highest in black children (75.1%); and intermediate in non-Hispanic white children (63.4%), Hispanic children (45.5%), and American Indian/Alaskan Native children (50.0%).
Poisoning Mortality
Deaths as a result of unintentional poisoning accounted for a relatively small proportion of all injury-related mortality (Table 1) and showed improvements in all race and ethnic groups during the study (Fig 3D). Poisoning mortality decreased from the 1980s until the late 1990s, when improvements leveled off, and by 2003, small increases in poisoning death rates were observed among white, black, Hispanic, and Asian/Pacific Islander children (Fig 3D). Despite a slight narrowing of disparities between white and black children, black children maintained a 2.7-fold higher rate of mortality as a result of poisoning at the last study period (Table 2).
Firearm-Related Injury Mortality, All Intent
Although firearm injury deaths declined during the study period, accounting for only 1.6% of all injury-related deaths in children who were aged 0 to 4 in 2003, 83.4% of these occurred in minority children, with black children accounting for more than half of all firearm injury deaths (Table 1). During the study period, 2167 deaths in children aged 0 to 4 years were attributable to firearms, with disparities during the study time frame widening in both black and American Indian/Alaskan Native children relative to white children. 2 for trend demonstrated significant lowering of firearm injuries during the study for white, black, and Hispanic children (Fig 3E). Despite this, black and American Indian/Alaskan Native children continued to have mortality rate ratios triple those of white children (Table 2).
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
A wide array of injury prevention efforts preceded or occurred concomitantly during this study.1,2,5–17,24–34 There are reports of declining injury rates with injury prevention efforts such as smoke detectors/smoke alarms, child restraint systems, speed limit and driver licensing laws, helmets, child safety caps, and window guards to protect against falls.1,2,6,8,9,25,32–34 Much of the historical decline in injury deaths is credited to a combination of passive and active injury prevention measures, including legislative and regulatory enforcement, educational efforts, and increased access to safety products1,2,6,9,16,24 as well as economic conditions and advances in trauma care.
Although there were significant trends toward improvement within racial groups, disparities in motor vehicle occupant mortality widened among black and American Indian/Alaskan Native children compared with white children. Mortality improvements were most notable for infants who were younger than 12 months, suggesting that additional work to increase use of booster seats in vulnerable populations is warranted. Pedestrian traffic–related mortality showed consistent downward trends in this study. Reports of efficacious prevention initiatives include "Safety Cities," rerouting traffic, improved traffic signs, speed controls, and building of playgrounds to provide alternative play spaces to the sidewalk and street.3,4,35,36 Despite improving rates in black children, significant disparities in pedestrian mortality remain.
Although the level of improvement of residential fire mortality in black children is impressive, a nearly threefold disparity in burn mortality continues for black compared with white children despite legislation/regulations, improved building codes, safety education, and promotion of smoke detectors/alarms. Some studies have questioned the effectiveness of smoke alarm giveaway programs,34 and updated building codes may have less benefit in urban black neighborhoods with infrequent new construction. Cigarette smoking, a reported contributor to house fires, is increasing in black individuals of childbearing age.37,38 Smoking seems to be a risk factor for residential fires through adults’ smoking in bed and increased accessibility of matches and lighters by children.12–14,39,40 Effective smoking cessation programs for adults may prove an efficacious injury prevention tool for residential fire prevention for very young children.
Mortality as a result of poisoning, once a leading cause of death in children aged 0 to 4 years, declined by >90% since the introduction of legislation that requires child safety caps on medicines, vitamins, and household products.9 Although racial/ethnic disparities remained, poisoning mortality declined during the study and was no longer a major source of mortality in 2003. However, we observed evidence of a troubling trend showing increases in recent poisoning deaths in Hispanic and black children, suggesting that injury surveillance systems should continue to monitor unintentional poisonings in this age group. In comparison, drowning accounts for more than twice as many fatalities in very young children but is reported to be an underrecognized source of injury, leading clinicians to omit this mechanism from their counseling sessions.41
Rates of unintentional suffocation mortality actually increased and disparities widened during the last 5 years of this study. Among black children, deaths from this cause began to increase in the late 1990s and rose to higher levels than that observed at the baseline, an increase that was accompanied by a significant widening of disparities between black and white children for this injury mechanism. The relative disparities that were observed during the early study periods were maintained during the last study period, with American Indian/Alaskan Native children also exhibiting higher mortality. Most of the increase in deaths occurred in infants who were aged <1 year, suggesting the need for additional study to assess whether this trend of increasing suffocation represents a data anomaly possibly associated with shifts in diagnostic classifications or a real increase in injury indicating a need for intensified culturally appropriate efforts directed toward known risk factors that are associated with an unsettling trend.42
All-intent firearm deaths declined in young children despite lack of significant improvement in intentional injury in most racial groups during the study period. Despite declines, disparities showed a trend toward widening in both black and American Indian/Alaskan Native children compared with white children. Previous studies noted disparities43 as well as effective preventions for lowering of firearm injury.43–46
This study has limitations. We did not have data that allowed investigation of contributing factors, such as socioeconomic status, social and environmental conditions, or ethnicity data necessary to produce 20-year trends for Hispanic children. Hispanic ethnicity was comprised of >90% white children, and, with few noted exceptions, Hispanic rates were similar to those of non-Hispanic white children. The composition of the study population changed during the study period to include increasing numbers of minority infants and children. This should have had minimal effect on our findings because we calculated race-specific mortality rates with denominators that had been established using standard means for intercensus estimates of population change calculated by the Census Bureau for use in national health statistics. Injury coding changed from ICD-9 to ICD-10 near the end of the study period. Because each period of analysis used the same ICD coding system for all race and ethnic groups, we believe that any effect of ICD coding on rates and thus rate ratios likely would have been similar across race.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
We are grateful to the Centers for Disease Control and Prevention for providing Web-based access to the data that made it possible to do the analyses necessary for this study.
|
FOOTNOTES |
|---|
Address correspondence to Joyce C. Pressley, PhD, MPH, Columbia University, Mailman School of Public Health, 722 W 168th St, Room 1712, New York, NY 10032. E-mail: JP376{at}columbia.edu
The authors have indicated they have no financial relationships relevant to this article to disclose.
An early version of this work was presented at the Seventh World Injury Conference; June 6–9, 2004; Vienna, Austria.
All interpretations are solely those of the authors and should not be attributed to the Centers for Disease Control and Prevention.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  Splinters & Fragments Inj. Prev., August 1, 2007; 13(4): 288 - 288. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2007 American Academy of Pediatrics. All rights reserved. |
||
|
||
