*CONTEXT* Standard network meta‐analysis & indirect comparisons combine aggregate data from multiple studies, assuming that any effect modifiers are balanced across pops. & often incurring in aggregation bias 2/n
*AIM* Propose a new method extending the standard network meta‐analysis framework
*METHOD* Use of general numerical approach using quasi‐Monte‐Carlo integration 3/n
*FINDINGS* Uncertainty is substantially reduced by explaining within‐ & between‐study variation, and estimates are more interpretable 4/n
• • •
Missing some Tweet in this thread? You can try to
force a refresh