Diagnostic Math

Between the deep learning work I’ve done in low vision and glaucoma and my time spent as a deeptech investor, I’ve spent a great deal of time looking at diagnostic technologies of all sorts and thinking about diagnostic tests and how to make them successful.

While each technology and entrepreneurial team is different, there are some commonalities that drive whether a diagnostic test is useful around concepts like sensitivity, specificity, positive predictive value, and AUROC.

I summarized these insights in the comprehensive essay “Why is it so Hard to Build a Diagnostic Business?”. Below are the embedded computational models and Google Sheets workbooks developed as part of that research.

1. Interactive Clinical Utility Model

Test sensitivity and specificity alone can be deceptive because disease prevalence in the target screening population heavily dictates whether a positive result is real. Use the interactive calculator below to adjust prevalence, sensitivity, and specificity:

2. Interactive Diagnostic Economics Model

Unlike therapeutics which command high prices for treating disease, diagnostics face intense payer constraints. Calculate the cost to screen one true positive across test price, confirmation procedure costs, and population prevalence:

3. Master Financial & Clinical Trial Workbook

Workbook Directory & Direct Tab Links

• Main Workbook
  • Tab: HIV Test Baseline — Baseline sensitivity/specificity calculation from the essay.
  • Tab: PPV & NNS Calculator — Positive predictive value and number needed to screen formulas.
  • Tab: Cost to Screen 1 True Positive — Cost-effectiveness model incorporating follow-up confirmatory workups.
  • Tab: Comparative Utility — Side-by-side comparative health economics framework.
  • Tab: Cologuard vs. FIT (2014) — Clinical comparison from the 2014 NEJM study.
  • Tab: Cologuard Plus vs. FIT (2024) — Clinical comparison from the 2024 NEJM study.
  • Tab: Exact Sciences 1997–2023 Financials — 27-year longitudinal analysis of R&D, S&M, and net losses pulled from SEC 10-K filings.

Suggested Citation

If you use these calculators or spreadsheets in preparing a research paper, grant proposal, or investment memo, please cite this resource:

Tseng, B. (2024). Calculators and Economic Models for Diagnostic Figures of Merit [Software & Spreadsheet Models]. Retrieved from https://benjamintseng.com/portfolio/diagnostic-math/