Meta-analysis services for systematic reviews

Pooling is only sound when the studies are similar enough to be estimating a comparable effect. We start by checking that your included studies share a clinical and methodological logic, then judge whether the heterogeneity is low enough for a pooled number to mean something. When it is not, we are direct about it and run a structured narrative synthesis instead, rather than forcing a misleading single estimate.

The choice of effect measure matters as much as the model. We pick the odds ratio, risk ratio, mean difference, or standardized mean difference that fits your outcomes and your readers, and we explain the trade-off so the analysis answers your review question rather than the other way round.

1. 1

   Effect-size extraction

   We pull the effect estimate and its variance from every included study, converting between scales where studies report outcomes differently so they enter one analysis.

2. 2

   Model selection and pooling

   We choose fixed-effect or random-effects pooling on a stated rationale, then compute the pooled estimate with a confidence interval and a clear interpretation.

3. 3

   Heterogeneity and exploration

   We quantify heterogeneity with Cochran's Q, I-squared, and tau-squared, then explore it with subgroup analysis and meta-regression where the data allow.

4. 4

   Bias, robustness, and certainty

   We test small-study effects with funnel plots and Egger's test, run sensitivity analyses, and rate the certainty of evidence with GRADE.

The analysis is only as defensible as the tooling it runs on. We work in scripted, version-controlled environments so every number traces back to a line of code, and we match the method to your data rather than forcing your data into one default routine.

You get the pooled results, publication-quality forest plots and funnel plots, the heterogeneity and bias diagnostics, and a methods-and-results write-up aligned with the PRISMA 2020 reporting guideline. Every figure ships with the analysis script, so the work is reproducible and your reviewers can see exactly how each number was produced. If you would rather try the pooling yourself first, our free meta-analysis calculator runs both models on your own numbers.

• Pooled effect estimate with confidence interval and a plain-language interpretation
• Forest plot, funnel plot, and any subgroup or cumulative plots your review needs
• Cochran's Q, I-squared, and tau-squared with a reading of what the heterogeneity means
• Egger's test, trim-and-fill, and a small-study-effects assessment
• Subgroup, meta-regression, and sensitivity analyses specified in your protocol
• A methods-and-results section drafted for your manuscript, plus the annotated R script

The brief differs by who is asking, and we shape the analysis and the write-up to match. Doctoral candidates need a defensible quantitative chapter and a viva they can answer for; clinicians and guideline teams need a pooled estimate that will survive editorial and peer scrutiny; and authors in revision need a specific reviewer demand answered quickly and cleanly.

We quote a fixed fee per project rather than charging by the hour, so the cost is known before any work begins. What sits behind that number is the amount of analytical work your evidence actually requires: the count of included studies and outcomes, whether effect sizes need converting between scales, whether you need a single pairwise pooling or a full network meta-analysis, and how much of the methods-and-results writing you want us to draft. Send your included studies and your outcomes and we will scope the analysis and return a fixed quote, with no obligation to proceed.