Methodologic issues in the conduct and interpretation of pediatric effectiveness research

Effectiveness research represents a number of methodologic challenges not shared with randomized, controlled clinical trials. This practice-based research attempts to translate clinical practices to a wide variety of different practice settings and situations and to diverse patient subgroups. However, because study designs most often used in the conduct of effectiveness research limit the ability to establish firm causal links between medical care and outcomes, it is important to address key methodologic features to generate sound, useable findings. Such features include selection of appropriate outcome measures (with a priori hypotheses linking care to the outcomes chosen), specification of appropriate primary sampling unit, specification of unit of analysis, establishment of appropriate comparison groups, and case-mix adjustment. Conduct of this type of research in pediatrics presents a number of unique methodologic concerns that either do not apply in adult medicine or are particularly acute in pediatrics. To alert policy makers and funders to the unique aspects of pediatric health services research and to provide guidance for the conduct and interpretation of pediatric effectiveness studies, we have organized and described the methodologic issues associated with the specific type of pediatric care under study (eg, specific disease-prevention, "bundled" care for chronic disease, care for problems with social etiologies, etc). We conclude with a summary of the methodologic steps that are critical to the conduct of sound effectiveness research in pediatrics.