The Fragility Index of meta-analyses

The conclusion of a meta-analysis of a binary outcome could be changed if the outcomes of a few patients were modified within the randomized controlled trials included in it.

The Fragility Index for meta-analyses of randomized controlled trials is an intuitive measure for the confidence we have in the conclusions of a meta-analysis.

It is defined as the minimum number of patients from one or more trials included in the meta-analysis for which a modification on the event status (ie, changing events to non-events, or non-events to events) would change the statistical significance of the pooled treatment effect.

After specific event-status modifications, a statistically significant pooled treatment effect could be turned non-significant, and a statistically non-significant treatment effect could be turned significant.

Evaluate the Fragility Index for your own meta-analysis

Evaluate the fragility of the conclusions of your own meta-analysis!

How does it work

The method used to evaluate the Fragility Index of meta-analyses is based on an iterative re-evaluation of the statistical significance of the pooled treatment effect of modified meta-analyses, iteratively derived from the original meta-analysis by performing single event-status modifications in each arm of each trial in turn.

A schematic represention is available here . You can also check our code on Github.

Contact: ignacioatal--at--gmail--dot--com

Example of a statistically significant meta-analysis with Fragility Index = 2

Evaluate the Fragility Index for your own meta-analysis

1- Upload meta-analysis data as CSV File

(See data requirements for help)

2- Select the parameters of the meta-analysis

3- Evaluate the Fragility Index



Data requirements

The data you upload should be a CSV file looking like this:

Where the following headers need to be present:

  • STUDY_ID: an ID for each trial in the meta-analysis
  • EVENTS_1: the number of events in Arm 1
  • TOTAL_1: the total number of patients included in Arm 1
  • EVENTS_2: the number of events in Arm 2
  • TOTAL_2: the total number of patients included in Arm 2

Fields in the CSV should be separated with a comma (,), and text should be delimited with quotation marks (")

Evaluate the Fragility Index of your own meta-analysis!

Schemamic representation of the iterative method used to evaluate the Fragility Index of a meta-analysis