#NetworkMetaAnalysis

Sparse networks are where frequentist and Bayesian NMA part ways. With 1 or 2 trials per comparison, REML can put heterogeneity at zero and hand you tight intervals; a weakly informative prior on τ won't. Before trusting a ranking, run both and look at the gap.

CoreSR runs frequentist and Bayesian NMA in real R, in the same project as screening, extraction, RoB and GRADE: (free plan, or 7-day trial of paid) #NetworkMetaAnalysis #Bayesian #EvidenceSynthesis

Component NMA isn't just another league table. You estimate effects for treatment components and combinations — including ones no trial ran directly. Walkthrough in CoreSR: #ComponentNMA #NetworkMetaAnalysis #SystematicReview #EvidenceSynthesis #CoreSR

Owen & Abrams (BMC Med Res Methodol) let the log HR change linearly with log(time), then pool both terms. Gastric cancer PFS: constant HR 0.78, but 0.83 at 6 months and 0.99 at 3.5 years. Neat. But with 9 of 13 melanoma trials extrapolated, is a 5-year HR data or assumption?

Open access, code on GitHub: Pair it with their leave-one-out advice: in the gastric set, one trial's time trend ran the opposite way. #MetaAnalysis #NetworkMetaAnalysis #SurvivalAnalysis #HTA

Just published key insight into #GRADE approach to #NetworkMetaAnalysis. For assessing certainty of indirect #evidence GRADE relies on first order loop, ignores others. Maybe big problem, but review of 103 comparisons suggests almost no misleading conclusions. Hurrah for GRADE #NMA! @JClinEpi #EBM

Nice theory paper in Res Synth Methods: Rücker, Davies & Schwarzer show an NMA's covariance matrix is a geometric series of diffusion matrices, no inversion. Hat-matrix entries become random walks between treatments. Could slow convergence flag thinly connected networks?

Convergence was fastest with a central node like placebo as the absorbing node. The authors are candid that it changes no day-to-day NMA practice. Open access, R code on Zenodo: #NetworkMetaAnalysis #EvidenceSynthesis #RStats

When a frequentist and a Bayesian NMA disagree, that's information, not noise. In a sparse network it's usually the heterogeneity prior doing the work. In CoreSR you can fit both in real R on the same data and see where estimates move before you write up any ranking.

#NetworkMetaAnalysis #MetaAnalysis #SystematicReview CoreSR runs pairwise, frequentist and Bayesian NMA, component NMA and dose-response in real R, with GRADE in the same project. 7-day trial, from $39 CAD/mo:

Multicomponent interventions (exercise + education + CBT) almost never get tested in every combination. Component NMA asks which parts carry the effect, but it assumes they add up. Check that against the standard NMA fit before you trust it. CoreSR runs CNMA in real R.

Component NMA sits in the same project as screening, extraction, RoB and GRADE: (7-day trial, from $39 CAD/mo) #NetworkMetaAnalysis #SystematicReview #EvidenceSynthesis

This paper describes theoretical aspects of #Bayesian and #frequentist models for #NetworkMetaAnalysis (NMA), presents evidence from simulations and empirical studies comparing frequentist and Bayesian models and provides an overview of automated software packages to perform #NMA.

This #BMJ article, crucial for anyone conducting a #NetworkMetaAnalysis (#NMA) and using #GRADE, provides the key approach to concluding what treatments work best, what are intermediate, and what are among the worst.

Professional workshop: Meta-Analysis & Network Meta-Analysis with R 23–28 November 2026 | 7–10 PM WAT daily For clinicians, professionals & researchers. Certificate included. Register: WhatsApp: +234 805 338 2872 #MetaAnalysis #NetworkMetaAnalysis #RStats

Looking forward to starting this @NIHRresearch funded #networkmetaanalysis on #ADHD treatments in adults, with @CinziaGiovane, @srchamb, A. Philipsen, @UK_ADHD, @DrSusanYoung1, and @And_Cipriani Protocol out today in @BMJ_Open @SotonPsych, @SolentNHSTrust

In 2014, #GRADE introduced approach to rating #CertaintyofEvidence from #NetworkMetaAnalysis. Four years later an update that now represents current definitive GRADE guidance for rating certainty of evidence from an #NMA.

🚨 ICYMI - Diagnostic test accuracy #MetaAnalysis methods View the recording: #ResearchMethods #NetworkMetaAnalysis

Primary treatments for solitary hepatocellular carcinoma ≤ 3 cm: A systematic review and network meta-analysis 🌷 Ann Hepatobiliary Pancreat Surg 2024;28(4). #NetworkMetaAnalysis #HepatocellularCarcinoma #OverallSurvival #RecurrenceFreeSurvival

One network, very different certainty. A vs B can rest on high-certainty direct trials while A vs C is very low, inferred through a thin loop. NMA GRADE rates each comparison: direct, indirect, and incoherence between them. In CoreSR it sits beside the R output.

GRADE for NMA, frequentist and Bayesian NMA, CNMA and dose-response, all in one project. 7-day free trial: #NetworkMetaAnalysis #GRADE #SystematicReview #EvidenceSynthesis

Would you like to learn more about how we explore in our anaylses which #selfmanagement characteristics work (or do not work)? In our latest blog post @dimi_mavridis explaines the benefits and limitations of two different #networkmetaanalysis approaches:

📢RECENTLY UPDATED #COCHRANE REVIEW! 🗯️ What are the benefits and risks of different treatments for central serous chorioretinopathy (CSC)? 📊 67 studies, 4015 adults #NetworkMetaAnalysis 🔗Read the full review: @cochranecollab

Delayed gastric emptying can be troublesome after pancreaticoduodenectomy. What reconstructive approach best mitigates this? Answer from this #systematicreview and #networkmetaanalysis is in the picture. Or read via link: #openaccess #some4surgery

Day 3 of teaching with @AChaimani at @EvidSynIRL! Participants learn how to fit in R the penalized likelihood #NetworkMetaAnalysis model for rare events!

A recent #NetworkMetaAnalysis in #BloodAdvances compared the efficacy of #BTKinhibitors in patients with high-risk R/R CLL from 3 RCTs. Read the full publication here: or full visual summary here: #IndirectTreatmentComparison #CLL