Post-test probability
Enter valid values, then select Calculate.
Evidence tools
Practical tools for evaluating clinical evidence.
Convert a pre-test probability and likelihood ratio into a post-test probability, then examine the model, training data, human review, evidence retrieval and deployment controls behind a clinical AI product or API.
Educational use only. These tools are not medical advice or clinical decision support. Do not enter names, identifiers, health records or other patient information. Verify all inputs, outputs and assumptions independently.
Bayesian update
Post-test probability calculator.
Enter a pre-test probability as a percentage and a likelihood ratio for one finding or test result.
Calculation method
The calculator converts probability to odds, applies the likelihood ratio, then converts the result back to probability.
- pre-test odds = p ÷ (1 − p)
- post-test odds = pre-test odds × LR
- post-test p = post-test odds ÷ (1 + post-test odds)
Assumptions matter. A likelihood ratio is transferable only when the finding, threshold, test quality, setting and patient spectrum are sufficiently comparable to the source evidence. Sequential multiplication can be misleading when findings are not conditionally independent.
Clinical AI evaluation
Questions to ask before relying on a system.
Use the same scrutiny for a clinician product and a private API deployment. Mark only items verified from appropriate documentation or evidence; completion is not proof of quality, safety or regulatory status.
Diagnify applies the same questions to its own systems. Product copy, a certification logo or a single benchmark score cannot answer this checklist. Review the model and evaluation records, privacy and governance information, intended-use statement, regulatory position and local clinical-governance requirements.
Privacy by design
No clinical record belongs in these two browser tools.
The calculator needs only two numeric values, and the checklist needs only local checkbox state. They do not submit form entries or request a server response. This statement applies to these tools, not to data processing in the separate Diagnify product or a private API deployment.
No account
Neither tool requires sign-in, a name, an email address or an organisation.
No saved result
Selections and calculations remain in the current page and are cleared by reset or reload.
No patient input
Use abstract numeric values and system documentation only. Do not paste or infer patient information.
Two offerings, one evaluation standard