|
Abstract
|
OBJECTIVES. This study evaluated the effectiveness of a community coalition to prevent severe injuries to children in Central Harlem, New York, NY. It was hypothesized that injury incidence rates would decline during the intervention (1989 through 1991) relative to preintervention years (1983 through 1988); that the decline would be greatest for the targeted age group (5 through 16 years) and targeted injury causes (traffic accidents, assaults, firearms, outdoor falls); and that the decline would occur in the intervention community rather than a control community. METHODS. Surveillance of injuries that result in hospitalization and/or death among children in the two communities has been under way since 1983. Data from this surveillance were used to test whether the incidence of severe injury declined during the intervention; other temporal variations were controlled by Poisson regression. RESULTS. The incidence of injury among school-aged children in central Harlem declined during the intervention. The decline was specific to the targeted age group and targeted causes. A nonspecific decline also occurred in the control community. CONCLUSIONS. The declining incidence rate in Central Harlem is consistent with a favorable program effect, but additional investigation of possible secular trend or spillover effects is needed.
VioLit summary:
OBJECTIVE:
The goal of this study by Davidson et al. was to evaluate the effects of a community-based program, the Safe Kids/Healthy Neighborhoods Coalition, aimed at decreasing the incidence of severe injury among children in Central Harlem, New York City. A public health perspective was utilized.
METHODOLOGY:
The authors conducted a primary analysis of longitudinal data, collected over a nine year period from two primary hospitals serving the Central Harlem and Washington Heights districts in New York City. These data were used to evaluate the Safe Kids/Healthy Neighborhoods Coalition injury prevention program, developed in 1988. The program was aimed at reducing outdoor injuries and assaults to children in Harlem, NY. Reasons for the development of the Coalition were based on findings obtained from the Northern Manhattan Injury Surveillance System (NMISS). The authors reported that the NMISS was developed in 1983 to measure severe pediatric injury incidence. The authors stated that NMISS pre-intervention surveillance data for 1983-1987 revealed three patterns of injury incidence: 1) there was double the incidence of severe injuries for Central Harlem compared with Washington Heights, 2) the incidence of severe injury was decreasing among children aged 0-4 years, but increasing among children aged 5-16 years, and 3) falls and motor vehicle accidents were found to be the leading causes of injury. 9% of injury hospitalizations and 36% of injury deaths were caused by assaults. These findings led to the development of the Safe Kids/Healthy Neighborhoods Coalition program. This program involved 1) the renovation of Central Harlem playgrounds, 2) the supervision of children and adolescents in safe activities aimed at skill development (e.g., dance, sports, art, carpentry, horticulture), 3) the provision of injury and violence prevention education, and 4) the provision of reasonably priced safety equipment. 26 community organizations participated in the program and included large numbers of community children. Severe injury was defined by the authors as "an injury resulting in hospitalization and/or death." It included "all trauma, poisonings and burns that resulted from intentional, unintentional, and undetermined external causes." This definition was drawn from the International Classification of Diseases (ninth revision). Children, younger than 17, from the district of Northern Manhattan, who suffered severe injuries between the years 1983-1991, were included in the study.
Two physicians and one graduate student extracted injury data (e.g., International Classification of Diseases [ninth revision] E-codes) from the medical records of the two major hospitals serving the community. Fatal injuries data were obtained from the New York death certificate files. Children living in the study vicinity who had died from injury, but were not admitted to the hospital, were added to the study.
The authors reported that the NMISS data was compared with the New York State Uniform Hospital Discharge Data (UHDD) to ensure that incidence rate data were as close as possible to being correct. It was found that the 76% of the children from Northern Manhattan district were hospitalized in one of the two chosen hospitals for the study. The UHDD revealed that the other resident children, not admitted to one of these two hospitals but treated in others, suffered the same type of external causes of injury as those included in the study hospitals.
The authors reported that the study design involved injury incidence rate comparisons between a pre-intervention period (1983-1988) and an intervention period (1989-1991). Comparison designs included: 1) 5-16 year old children vs. 0-4 year old children in Central Harlem, 2) 5-16 year old children in Central Harlem compared with 5-16 year old children in Washington Heights, and 3) targeted injuries (e.g., those related to vehicles, outdoor falls, assaults and guns) compared with non-targeted injuries (e.g., poisonings, burns, etc) among 5-16 year old children in both Central Harlem and Washington Heights.
The authors calculated incidence rates for all injuries and targeted and non-targeted injuries independently over the nine year period at three month intervals. Dates of hospital admission, or death dates, were used to indicate the time of injury occurrence.
The authors conducted a Poisson regression analysis to analyze the data. The quarterly injury rate was used as the dependent variable. Independent variables included the year (coded as a continuous variable), the season (coded as a 4-way categorical variable), and intervention period (coded as a categorical variable which accounted for the pre-intervention period and the intervention period following the program's development).
FINDINGS/DISCUSSION:
The authors found that between the period 1983-1991, the number of non-fatal injuries for Central Harlem children aged 0-4 years was 680, and for children aged 5-16 years was 1217. The number of fatal injuries for Central Harlem children aged 0-4 years was 29, and for children aged 5-16 years was 37. When compared with Washington Heights, non-fatal injuries for children aged 0-4 years were 975 and 1516 for children aged 5-16 years. Fatal injuries for children 0-4 years were 34 and for children 5-16 years, the total was 14.
The authors calculated a mean annual adjusted incidence rate for Central Harlem. These were: 1) all severe injuries (1983-1991) for children 0-4 years = 1251, and for children 5-16 years = 968, 2) injury mortality rates for children 0-4 years = 37.9 and for children 5-16 years = 21.2, 3) all severe injuries during the pre-intervention period (1983-1988) for children 0-4 years = 1382 and for children 5-16 years = 1035, and 4) all severe injuries during the intervention period (1989-1991) for children 0-4 years = 989 and for children 5-16 years = 833.
These rates were compared with Washington Heights. All severe injuries (1983-1991) for children 0-4 years were 611 and for children 5-16 years, 468. Injury mortality rate (1983-1991) for children 0-4 years was 17.6 and for children 5-16 years, 11.4. All severe injuries during the pre-intervention period (1983-1988) for children 0-4 years totalled 655, and for children 5-16 years, 509. All severe injuries during the intervention period (1989-1991) for children 0-4 years grossed 523 and for children 5-16 years, 387.
The results of the three major incidence rate study comparisons follow. 1) Central Harlem: comparison of incidence rates for the targeted and non-targeted age groups. The results revealed that during the intervention period there was a significant decrease in the overall injury incidence rate in the targeted age group of 26% (relative risk [RR]=0.74, 95% confidence interval [CI]= 0.62,0.89), compared with the 6 years before the onset of the program. The authors did not find a significant reduction in injury incidence in the younger (0-4 yrs), non-targeted group. 2) School-aged children: comparison of incidence rates between the targeted and non-targeted districts; the results revealed that for the age group 5-16 years, there was a significant reduction in the injury incidence rate among children in Washington Heights of 30% (RR=0.70, 95% CI=0.59, 0.83). 3) Comparison of incidence trends for targeted and non-targeted injuries for both Central Harlem and Washington Heights; revealed that in Central Harlem there was a significant decrease during the intervention period for targeted injuries (motor vehicles and assault injuries) among children 5-16 years of 44% (RR=0.56, 95% CI=0.45, 0.71). Yet there was not a significant reduction for non-targeted injuries (RR=1.03, 95% CI=0.78, 1.35). The authors found no significant difference between targeted and non-targeted injuries for Washington Heights. They reported a significant decline among children aged 5-16 years, in the incidence rate of targeted injuries for motor vehicles only (RR=0.47, CI 95%=0.34, 0.66).
The authors stated that overall, consistent with their hypotheses, there was a 44% reduction in the risk of injury for targeted injuries among Central Harlem's school-aged children, during the program's intervention period. Also these were mostly specific to the types of injury particular to the program. The authors also reported that, not consistent with their hypotheses, there was also a significant reduction in injury incidence in the control district, Washington Heights, of approximately 30% for all severe injuries to children aged 5-16 years.
The authors concluded that although mixed, their findings revealed that since there was a significant reduction in the incidence rate of targeted injuries during only the intervention period, there was reason to believe that the intervention program had been effective. However, the significant decrease in motor vehicle injuries in both the intervention and control districts was reported to have weakened the effectiveness of the program in preventing this type of injury. The authors attributed this decline to 1) an independent, general decline in motor vehicle injuries, supported by a national decline in motor vehicle death rates after 1989, or 2) an effect of the program that "spilled over" into the control district. This was supported by an increase of participants residing in Washington Heights.
AUTHORS' RECOMMENDATIONS:
The authors recommended that additional yearly follow-ups be required to assess the persistence of the decline in injury incidence. The authors also suggested yearly follow up program evaluations in order to provide insight into the effect of more recent interventions. Individual level analyses were also suggested.
(CSPV Abstract - Copyright © 1992-2007 by the Center for the Study and Prevention of Violence, Institute of Behavioral Science, Regents of the University of Colorado)
1990s
New York
Community Based
Public Health Approach
Child Safety
Juvenile Safety
Youth Safety
Child Injury
Juvenile Injury
Injury Prevention
Middle Childhood
Late Childhood
Late Adolescence
Early Adolescence
Community Violence
Environmental Factors
Prosocial Skills
Social Skills Development
Youth Development
Child Development
Child Victim
Juvenile Victim
Juvenile Development
Firearms Violence
Violence Prevention
Prevention Program
Program Evaluation
Program Effectiveness
Urban Youth
Urban Violence |