Welcome to the Disproportionality Analysis Tool

About This Tool

This interactive assistant will guide you through performing a disproportionality analysis for pharmacovigilance signal detection.

Disproportionality analysis identifies potential safety signals by checking if a drug and adverse event are reported together more often than expected by chance.

Workflow Overview
  1. Study Setup: Define your drug, adverse event, and analysis parameters
  2. Data Input: Enter contingency table data or individual reports
  3. Analysis: Calculate disproportionality measures (ROR, PRR, IC, etc.)
  4. Signal Assessment: Evaluate signal strength and significance
  5. Interpretation: Understand the clinical relevance of findings
  6. Export: Generate a report of your analysis
Key Concepts

Reporting Odds Ratio (ROR): Measures how much more likely an adverse event is reported with a specific drug compared to all other drugs.

Proportional Reporting Ratio (PRR): Compares the proportion of a specific adverse event for one drug versus all other drugs.

Information Component (IC): A Bayesian measure that quantifies the strength of association between a drug and adverse event.

Contingency Table: A 2x2 table that organizes data for disproportionality analysis.

Understanding the Information Component (IC)

The Information Component (IC) is a Bayesian measure used in pharmacovigilance to detect drug-adverse event associations. It's part of the Bayesian Confidence Propagation Neural Network (BCPNN) method used by the WHO Uppsala Monitoring Centre.

IC Formula: IC = log₂( (a / (a+b)) / ((a+c) / N) )
Where: a = drug-event reports, b = drug reports without event, c = event reports without drug, N = total reports

How to Interpret IC Values

The IC value represents the strength of association between a drug and adverse event:

IC ≤ 0
No Association

The event is not reported more frequently than expected

0 < IC < 1
Weak Signal

Minor association that may warrant monitoring

1 ≤ IC < 2
Moderate Signal

Noticeable association that requires investigation

IC ≥ 2
Strong Signal

Strong association that needs immediate attention

Advantages of IC

  • Bayesian approach: Incorporates prior knowledge and handles uncertainty better
  • Stable with sparse data: Performs better than frequentist methods with small numbers
  • Provides probability estimates: Gives the probability of a true association
  • Widely used: Employed by WHO's global pharmacovigilance database
Information Component Examples
Example 1: Strong Signal (IC = 2.8)

Scenario: A new antidepressant and reports of serotonin syndrome

  • Drug-event reports (a): 25
  • Drug reports without event (b): 2,500
  • Event reports without drug (c): 100
  • Total reports in database (N): 1,000,000
IC = log₂( (25/2525) / (125/1000000) ) = log₂(0.0099 / 0.000125) = log₂(79.2) = 2.8

Interpretation: Strong signal (IC > 2). Serotonin syndrome is reported much more frequently with this antidepressant than expected.

Example 2: Moderate Signal (IC = 1.4)

Scenario: A diabetes medication and reports of pancreatitis

  • Drug-event reports (a): 45
  • Drug reports without event (b): 14,955
  • Event reports without drug (c): 370
  • Total reports in database (N): 200,000
IC = log₂( (45/15000) / (415/200000) ) = log₂(0.003 / 0.002075) = log₂(1.45) = 1.4

Interpretation: Moderate signal (1 ≤ IC < 2). Pancreatitis is reported more frequently with this diabetes medication, warranting further investigation.

Study Setup

Define Your Study Parameters
Start Date
End Date

Data Input

Contingency Table

Enter the data for your disproportionality analysis. The tool will automatically calculate all measures.

Adverse Event Present Adverse Event Absent Total
Drug of Interest
(a)
(b)
15,000
(a+b)
All Other Drugs
(c)
(d)
185,000
(c+d)
Total 415
(a+c)
199,585
(b+d)
200,000
(N)
Quick Data Entry

Alternatively, you can enter total values and let the tool calculate the contingency table:

Analysis Results

Reporting Odds Ratio (ROR)
-
95% CI: -
Proportional Reporting Ratio (PRR)
-
95% CI: -
Information Component (IC)
-
-
Chi-Square
-
p = -
Detailed Calculations
ROR Formula: ROR = (a/b) / (c/d) = (a×d) / (b×c)
PRR Formula: PRR = (a/(a+b)) / (c/(c+d))
IC Formula: IC = log₂( (a/(a+b)) / ((a+c)/N) )
Measure Value Interpretation
ROR - -
PRR - -
IC - -
Chi-Square - -

Signal Assessment

Signal Strength
Weak Moderate Strong
ROR Signal
-
PRR Signal
-
IC Signal
-
Overall Assessment
-
Signal Criteria
Measure Threshold Current Value Met?
ROR > 2.0 - -
PRR > 2.0 - -
IC > 1.0 - -
Case Count > 3 - -

Interpretation

Signal Interpretation

Complete the analysis to see signal interpretation.

Clinical Context

Complete the analysis to see clinical context.

Recommendations

Complete the analysis to see recommendations.

Export Report

Generate Analysis Report