Poster Presentation
Wednesday, July 11, 2007
9:00 AM - 9:45 AM
Wednesday, July 11, 2007
2:45 PM - 3:30 PM
Translating extensive research as evidence for practice: Meta-analysis of interventions to improve metabolic control among adults with type 2 diabetes
Vicki Conn, PhD, RN, FAAN, School of Nursing, University of Missouri, Columbia, MO, USA
| Learning Objective #1: describe the importance of meta-analysis in summarizing research findings for implementation in evidence-based practice. |
| Learning Objective #2: describe the overall effect of interventions to improve diabetes self-management on metabolic outcomes, and identify significant moderators of effectiveness. |
Background: Evidence-based practice requires systematic evaluation of diverse and sometimes large bodies of existing research and program evaluations to determine which interventions should be incorporated in practice. Meta-analysis can quantitatively synthesize research to determine interventions that merit evidence-based practice implementation.
Aims: This meta-analysis integrates ambulatory care program evaluations and research testing the metabolic effects of interventions to improve diabetes self-management among adults with type 2 diabetes.
Methods: Comprehensive searching strategies located diverse international (Europe, Asia, North America) published and unpublished intervention reports that measured glycated hemoglobin (HbA1c) outcomes. Statistical results and study characteristics were coded. Fixed- and random-effects meta-analytic procedures were used to analyze weighted main effects and moderator influence.
Results: Data were synthesized from 103 research reports which included 10,455 subjects. Common intervention strategies included self-monitoring, goal setting, behavior modification, cognitive modification, social modeling, feedback, problem solving, social support, barriers management, contracting, stimulus control/cueing, and motivational interviewing. The overall mean weighted effect size (0.29 to 0.34) documents that intervention subjects experienced improvements in metabolic control. This effect size translates to a 7.38% HbA1c for treatment subjects versus a 7.83% HbA1c for control subjects. Interventions that focused only on exercise behaviors (0.45) were more effective than interventions which targeted multiple health behaviors (0.22). Studies with more male subjects reported higher effect sizes. Metabolic outcomes were unrelated to baseline HbA1c or body mass index. Too few studies documented cost factors related to interventions or outcomes to quantitatively synthesize the information.
Conclusions: These findings document significant metabolic control improvements from diabetes self-management interventions. Interventions that focus exclusively on exercise behavior may produce larger improvements. Interventions designed specifically for women need to be tested to ensure women attain improved metabolic outcomes. These findings document the importance of meta-analysis as a research methodology for translational research as a foundation for evidence-based practice.