📘 Meta-Analysis Web Application – Help Guide

📌 Overview

This web application performs quantitative meta-analysis for both continuous and dichotomous outcomes. It generates publication-ready statistical outputs, including forest plots, funnel plots, risk-of-bias visualizations (ROB 2), certainty tables, and journal-ready narrative results.

The tool is intended for academic and research use and follows established methodological standards, including guidance from the Cochrane Handbook, PRISMA, and ROB 2.

---

1️⃣ Enter Studies (Effect Size Data)

You may enter study-level data using any of the following methods:

• Manual Entry (Recommended)
  Click “+ Add study row” and enter values directly into the table. You may paste data directly from Excel or CSV into any cell. Rows auto-expand as needed.
• Paste CSV Text
  Paste rows copied from Excel or CSV. A header row is optional and will be auto-detected.
• Upload CSV File
  Upload a .csv file containing your study data.

📄 Required Columns

Continuous Outcomes (MD / SMD)

study, year, n_c, mean_c, sd_c, n_t, mean_t, sd_t

Dichotomous Outcomes (RR / OR)

study, year, event_c, total_c, event_t, total_t

⚠️ Sample sizes and event counts must be numeric. Standard deviations must be greater than zero.

---

2️⃣ Missing Standard Deviation (SD Helper)

If a study does not report standard deviation (SD), select the study row and use the built-in SD Helper.

The following validated transformations are supported:

• From Standard Error (SE): SD = SE × √n
• From 95% CI: SD = ((Upper − Lower) / (2 × 1.96)) × √n
• From IQR: SD ≈ IQR / 1.35
• From Range: SD ≈ Range / 4 (n < 70)

Computed SD values are automatically inserted into the table.

---

3️⃣ Run Meta-Analysis

Select the desired effect measure and statistical model:

• Mean Difference (MD)
• Standardized Mean Difference (SMD, Hedges g)
• Risk Ratio (RR)
• Odds Ratio (OR)

Choose either a Fixed-effect or Random-effects (DerSimonian–Laird) model, then click “Run Meta-Analysis”.

---

4️⃣ Outputs Generated Automatically

• 📊 Forest plot (600 dpi, publication-ready)
• 📉 Funnel plot with 95% pseudo-confidence limits
• 📄 Study-level effects table (CSV)
• 🚦 ROB 2 traffic-light plot
• 📋 ROB 2 certainty table
• 📝 Journal-ready narrative results

All outputs appear in the Downloads section after analysis.

---

5️⃣ Publication Bias Assessment

Funnel plots are generated for all analyses. Egger’s regression test is automatically performed when at least 10 studies are included.

Results are summarized in both the statistical output and narrative results.

---

6️⃣ Risk of Bias (ROB 2)

Risk of bias may be entered manually or pasted as CSV.

Required Columns:

study, randomization, allocation, blinding, incomplete, reporting, overall

Allowed values (case-insensitive):

• Low
• Some concerns
• High

The application generates a ROB 2 traffic-light plot and a publication-ready certainty table.

---

7️⃣ Narrative Results (Journal-Ready)

Automatically generated narrative text summarizes:

• Forest plot results
• Heterogeneity interpretation
• Funnel plot and publication bias
• Risk of bias and certainty of evidence

These narratives are suitable for direct inclusion in the Results section of a manuscript and remain fully editable after download.

---

📊 Interpretation Notes

• RR and OR are analyzed on the log scale and back-transformed
• I² interpretation:
  • 0–25%: low heterogeneity
  • 25–50%: moderate heterogeneity
  • >50%: substantial heterogeneity

---

📚 Recommended Citation

Rajput R. Meta-Analysis Web Tool for Effect Size Synthesis, ROB 2 Visualization, and Narrative Reporting. Version 1.0, 2026. Available at: [Application URL].

© 2026 Ravindrasingh Rajput. All rights reserved. Users are responsible for verification of analyses prior to publication.