tracenma

Tracenma provides 217 datasets extracted from published systematic reviews with network meta-analysis for developing and appraising methodology to assess the transitivity assumption.

Key Features:

• Network Meta-Analysis Dataset Collection: Contains 217 eligible connected networks from 453 networks in the nmadb database, with each network comprising a dataset.
• Characteristic Extraction and Classification: Extracts at least four characteristics per network, classified into three types (clinical, demographic, and methodological) and ten subtypes (participants, treatments, outcomes, age, sex, ethnicity, study design, study setting, risk of bias, and withdrawals).
• Data Access Functions: The tracenma R package provides functions that facilitate access to the database.
• Characteristic Distribution Summaries: Summarises the distribution of characteristics and missing data across all datasets by type and subtype.

Scientific Applications:

• Transitivity Assessment Methodology Development: Initiates methodological advances in the evaluation of the transitivity assumption in network meta-analysis using empirical data.
• Empirical Evaluation of Novel Methods: Enables development and empirical evaluation of novel quantitative methods for assessing transitivity, addressing the current lack of such methods.
• Effect Modifier Distribution Analysis: Supports analysis of how treatment comparisons are similar regarding the distribution of important characteristics that act as effect modifiers.

Methodology:

The nmadb R package was used to build a database of connected networks. Relevant literature was consulted to determine necessary characteristics comprising potential effect modifiers. For each eligible network, the systematic review report and supplementary material were referenced to retrieve at least four extractable characteristics, comprising studies, corresponding treatment comparisons, and aggregated characteristics at the study level. Characteristics were classified into three types and ten subtypes, with some characteristics transformed when reported inconsistently across studies concerning measure scale and summary statistics.