Our Process

The Best Practices in Prevention-Oriented Child Death Review site is tailored for child death review teams. These local- or state-level multidisciplinary teams regularly review child deaths to better understand how and why children die, and use the findings to take action to prevent other deaths and improve the health and safety of children.

Because injury remains the most prevalent cause of child death after the first year of life, this Best Practices site has been developed specifically for CDR teams as a flexible, searchable database of interventions that aim to prevent child injury death. Our site takes a critical look at the evidence behind each intervention; rating the intervention and providing additional resources for those that are promising or recommended.

Stage One was conducted by the Harborview Injury Prevention and Research Center, supported in part by a HRSA EMS–C Targeted Issues Grant Program (1H34MC02543–01–033). The Harborview team first developed a logic model for five causes: Child Abuse, Drowning, Unintentional Firearms, Motor Vehicle, and Youth Suicide.

Through extensive literature searches and peer review by experts in each content area, they reviewed and rated the strength and quality of published evidence supporting the efficacy of these interventions. They established through consensus, a rating of the research supporting each intervention and identified evidence-based best practices in injury prevention for each studied mechanism.

Their first step in building the site was to develop mechanism-specific logic models for each mechanism. These logic models were based on a broad review of literature, along with input from experts in the field. Each logic model was tailored to reflect the types of intervention most useful and relevant to CDR teams, especially those operating in a public health system. Interventions that fall outside of the scope of a state or local multidisciplinary review team were usually not included.

Using the logic models as guides, the Harborview Team then completed a careful review of the literature, summarizing interventions based on the strength of evidence available in published studies. The Harborview Team preferentially weighted studies that were methodologically rigorous (e.g. randomized trials), carefully conducted, or used serious injury and mortality as outcome measures. Other studies of lesser methodological rigor or those that tracked intermediate outcomes (such as behavior change) were included in the reviews but — in most cases — were not as influential in determining the ratings. In all cases, the Harborview Team required included studies to measure effectiveness with, at least, a comparison group of some type. Uncontrolled case series were not considered.

The study group incorporated the results of this literature review and the input from content experts to rate each intervention in terms of the strength and quality of evidence supporting the injury prevention approach. The goal was to highlight as "recommended" those interventions that should be the first consideration for CDR teams wanting to take action to address a state or local child injury concern. In some cases, with newer approaches that have not been fully evaluated or did not demonstrate changes in deaths or injuries but behavior or other change, the Harborview Team use the rating of "promising."

Stage Two was conducted by the Michigan Public Health Institute’s National Center for Child Death Review in partnership with the Education Development Center’s Children’s Safety Network, supported in part by a grant from the HRSA Maternal and Child Health Bureau and the Department of Justice Coordinating Council on Juvenile Justice and Delinquency. The five additional causes of deaths were completed in Stage Two.

Stage Two differed from the Harborview approach, in that following extensive literature review, a team internal to the two participating agencies assigned the ratings, and then submitted their findings to a small group of peer reviewers (three for each cause) but were not seeking consensus from a larger body of experts.

Assigning Ratings:

Both groups recognize that this process of literature review and intervention rating is inherently subjective. Editorial decisions were made with respect to the breadth of study quality that the teams would include in their considerationsą as well as the depth to which their librarian would go to uncover descriptions or studies of interventions in lesser known, poorly indexed or less recent media˛. The decision-making was always driven by three considerations:

  1. Is the intervention or program in question likely to be relevant to the public-health or other prevention oriented focus of most CDR team deliberations?
  2. Is the intervention widely practiced or promoted, to the degree that a clear understanding of the relevant evidence base is needed to guide resource allocation?
  3. Would additional data from these sources substantively change the overall assessment of the evidence base supporting the intervention in question?

Click here for a table describing our criteria for assigning ratings:


References

  1. Ogilvie D, Egan M, Hamilton V, Petticrew M., Systematic reviews of health effects of social interventions: 2. Best available evidence: how low should you go? J Epidemiol Community Health. 2005 Oct;59(10):886-92.


  2. Ogilvie D, Hamilton V, Egan M, Petticrew M., Systematic reviews of health effects of social interventions: 1. Finding the evidence: how far should you go? J Epidemiol Community Health. 2005 Sep;59(9):804-8.