Preparing your experience...
Preparing your experience...
Capability Domain
Quantify the balance between a drug's benefits and risks using the QBR framework with therapeutic window computation. Move from subjective judgment to reproducible, auditable numbers.
Benefit-risk analysis answers: for this patient population, do the benefits outweigh the harms? The QBR (Quantitative Benefit-Risk) framework treats benefits and risks symmetrically — both as contingency tables — then integrates them into a single dimensionless index.
QBR uses the same 2×2 contingency table structure as signal detection, applied separately to benefit outcomes and risk outcomes. Each cell captures observed vs. expected counts in treated vs. control populations.
Benefit Contingency Table
Drug responders vs. non-responders. Captures NNT — how many patients need treatment for one to benefit.
Risk Contingency Table
Adverse event rates in treated vs. control. Captures NNH — how many patients need treatment for one to be harmed.
QBRI (Quantitative Benefit-Risk Index) integrates the benefit and risk tables into a single value. A QBRI above 1.0 indicates benefit exceeds risk; below 1.0 indicates the opposite. The index is weighted by outcome severity using ICH E2A seriousness criteria.
QBRI Interpretation
The therapeutic window is the dose range where the benefit curve (Hill equation for efficacy) is above the risk curve (Hill equation for toxicity). NexVigilant computes both curves from clinical trial data and identifies the window boundaries numerically.
E_max
Maximum achievable efficacy from the benefit Hill curve
EC50 / TC50
Dose for 50% efficacy vs. dose for 50% toxicity
Therapeutic Index
TC50 / EC50 — the wider this ratio, the safer the drug
Number Needed to Treat (NNT) and Number Needed to Harm (NNH) are derived directly from the benefit and risk contingency tables. They translate statistical results into clinically interpretable quantities: for every NNT patients treated, one benefits; for every NNH patients treated, one is harmed. A high NNH/NNT ratio indicates a favorable benefit-risk profile.
5 Computation Methods
QBR computation, QBRI integration, therapeutic window, seriousness classification, NNH/NNT from contingency tables
Clinical Trial Data
SAE rates, efficacy endpoints, and study design from ClinicalTrials.gov — the primary source for benefit quantification
3 Reference Sources
FAERS outcomes for real-world risk, drug labeling for labeled safety events, and ICH E2A for seriousness weighting
Data Sources
Connect your AI agent to mcp.nexvigilant.com and run QBR analysis with therapeutic window computation.