Verflux: A browser-based platform for end-to-end systematic reviews and meta-analysis

Abstract

A systematic review or meta-analysis is seldom carried out inside one program. In practice the work is divided among a literature database, a reference manager, a screening tool, a spreadsheet, one or more appraisal templates, and a statistics package such as R or Stata, and the final analytic step demands programming skill that many researchers lack and a licence that many cannot afford. Verflux closes that division. It is a browser-based platform that carries a review through its full course inside one project: protocol registration, searching across about nine databases and a trial registry, forward and backward citation chasing, screening by two independent reviewers, data extraction, risk-of-bias appraisal, GRADE certainty rating, meta-analysis, figure generation, and PRISMA 2020 reporting, with no installation and no code. The analytic engine pools effect sizes with inverse-variance fixed-effect and random-effects models (DerSimonian–Laird and Paule–Mandel), reports heterogeneity statistics and prediction intervals, screens for small-study effects, and extends to subgroup analysis, meta-regression, cumulative meta-analysis, and a frequentist network meta-analysis with treatment rankings; every estimator is tied to its original published source. Its outputs were validated both by hand and against established meta-analysis software, agreeing to within rounding. By folding the whole process into one tool and removing the coding barrier, Verflux makes rigorous evidence synthesis more accessible without relaxing the standards reviewers and journals expect.