Updating a Clinical Prediction Model for Identifying Monogenic Diabetes to Include Both Clinical Features and Biomarkers
Authors' conclusion
We developed an updated monogenic diabetes probability calculator that integrates both clinical features and biomarkers, providing greater discrimination than using clinical features or biomarkers alone and providing appropriate measures for selecting individuals for monogenic diabetes diagnostic testing. This is now available as an online calculator and has immediate clinical utility for White European individuals diagnosed with diabetes ≤35 years.
Study context
- Population
- Individuals with diabetes; early-insulin-treated (proxy for T1D) and not-early-insulin-treated (proxy for T2D); case-control data (monogenic diabetes = 594, non-monogenic diabetes = 597), UNITED population data, external validation cohortUNITED population data 1,299 · external validation cohort 1,025
- Comparison
- Clinical features + biomarkers model vs clinical features only vs biomarkers alone
- Outcome
- Discrimination (ROCAUC) for monogenic diabetes
Methodological detailsBayesian recalibration mixture model, case-control development, recalibrated to population data, external validation in independent population-based cohort
- Bayesian recalibration mixture model
- Case-control development
- Recalibrated to population data
- External validation in independent population-based cohort
Performance
As reported in the abstract
Clinical features + biomarkers
0.98
95% CrI 0.95–0.98
Clinical features only
0.80
95% CrI 0.71–0.82
Biomarkers alone
0.96
95% CrI 0.95–0.97
P < 0.001
Further findings
ROCAUC, not-early-insulin-treated participants
- ROCAUC
- 0.86 (95% CrI 0.85-0.88)
External validation ROCAUC
- Early-insulin-treated
- 0.98 (95% CrI 0.98-0.98)
- Not-early-insulin-treated
- 0.92 (95% CrI 0.90-0.93)
- Both models calibrated well
Positive test rate for monogenic diabetes using ≥5% probability threshold
- Positive test rates
- 16-19%
Original publication
Composed by AI from the published abstract and checked against it. How Briefs are made →