BENEFIT-HARM ASSESSMENT USING DATA DERIVED FROM NETWORK META-ANALYSIS- AN APPLICATION USING TREATMENTS FOR ACUTE VENOUS THROMBOEMBOLISM
Author(s)
Cameron C
Cornerstone Research Group Inc., Burlington, ON, Canada
Presentation Documents
OBJECTIVES: To present some simple graphical ways to illustrate benefits and harms of treatments derived from network meta-analysis, and provide an Excel-based tool so other can create similar graphical outputs. METHODS: We reanalyzed a published network meta-analysis comparing various treatment of acute venous thromboembolism using Bayesian approaches, and explored new forms of graphical presentations of results for benefits and harms. RESULTS: We present the findings for benefits and harms derived from network meta-analysis of venous thromboembolism using various approaches. Conventionally, research in this area has focused on presenting benefits and harms as relative effect measures or using probabilities and ranks. Findings have also traditionally not considered findings for benefits and harms together. We focus on methods for presenting findings on the absolute scale and integrating data related to a benefit and harm. We report findings using 7 approaches: scatterplot plots with error bars; confidence ellipses; bar charts with error bars, outcome acceptability curves, icon arrays, icon arrays depicting quality adjusted life days gained, and acceptability curves varying the decision makers’ ratio for preference of outcomes. CONCLUSIONS: We provide a multitude of ways to present findings on the benefits and harms of drugs derived from network meta-analysis on the absolute scale. We also provide a user friendly Excel-based tool to assist others in conducting benefit harms assessments using data derived from network meta-analyses.
Conference/Value in Health Info
2016-05, ISPOR 2016, Washington DC, USA
Value in Health, Vol. 19, No. 3 (May 2016)
Code
PRM18
Topic
Clinical Outcomes
Topic Subcategory
Clinical Outcomes Assessment
Disease
Cardiovascular Disorders, Systemic Disorders/Conditions