ROC operating threshold audit
Inspect sensitivity, specificity and Youden index at each observed score threshold.
Your entries stay in this page’s memory and are not submitted to us. Downloads are generated locally. Reloading or leaving the page clears your entries. Do not enter patient information.
Larger scores indicate greater event risk. Outcomes must be zero or one with both classes present. Up to 100 rows; classification is positive when score ≥ threshold.
The starting values are hypothetical. Results update as you edit. Use aggregate or synthetic inputs only.
Calculation results
- Maximum exploratory Youden index
- 0.33333333
- Best observed thresholds
- 0.9
- Threshold table (threshold: sensitivity / specificity / J)
- Above maximum: 0 / 1 / 0 | 0.9: 0.3333 / 1.0000 / 0.3333 | 0.8: 0.3333 / 0.5000 / -0.1667 | 0.7: 0.6667 / 0.5000 / 0.1667 | 0.4: 0.6667 / 0.0000 / -0.3333 | 0.3: 1.0000 / 0.0000 / 0.0000
How to interpret the output
The table exposes the sensitivity–specificity trade-off. Maximum in-sample Youden index is an exploratory threshold, not a clinically validated decision rule; costs, prevalence and external validation matter. An additional above-maximum threshold predicts no positives. All tied best observed thresholds are reported.
Method referenceSource checked 2026-11-06. Read the reproducible calculator validation report.