Challenges in Treatment Node Classification for Network Meta‐Analysis of Non‐Pharmacological Interventions: Insights From a Researcher Survey

Information from the abstract

OBJECTIVE: Non-pharmacological interventions (NPIs) are diverse and complex, making treatment node classification in network meta-analyses (NMAs) particularly challenging. Accurate node classification is essential, as it directly affects the validity and applicability of NMA findings. The aim of this study was to explore current practices in treatment node classification in NMAs of NPIs and identify key challenges. METHODS: An online survey was conducted among the first and corresponding authors of NMAs of NPIs published in first-quartile journals over the past five years. Eligible authors were identified through a systematic search of PubMed, Embase, and CINAHL on February 4, 2025. A self-administered questionnaire assessed experiences, perceptions, and recommendations regarding node classification. Quantitative data were summarized descriptively, and qualitative responses were analyzed thematically. RESULTS: Of 471 eligible authors contacted, 32 responded (response rate: 6.8%). Most (43.8%) had 5-10 years of experience in evidence synthesis, and 81.3% had published 1-5 NMAs of NPIs. Nearly all reported difficulties in node classification due to within-intervention and comparator variation, incorporating delivery features, balancing lumping versus splitting, and meeting the transitivity assumption. Respondents classified nodes by following prespecified protocols, relying on the characteristics of included studies, or using mixed approaches. While views on standard guidance were mixed, several respondents supported the development of simplified guidance for major research areas to support consistent node classification. CONCLUSIONS: Researchers encounter considerable difficulties in classifying treatment nodes in NMAs of NPIs. Findings highlight the need for prespecified and structured frameworks for certain types of NPIs to improve reproducibility, and reliability in future NMAs.

Why this record is monitored

This record has an Impact Signal of 75/100 based on recency, source, collaboration, and bibliographic signals. It prioritizes monitoring and is not a judgment of research quality.

Related topics: Mental Health Research Topics · Advanced Causal Inference Techniques · Meta-analysis and systematic reviews