Derivation and Validation of D-RISK: An Electronic Health Record–Driven Risk Score to Detect Undiagnosed Dysglycemia in Clinical Practice
Authors' conclusion
Designed for use in EHR, D-RISK performs better than commonly used screening guidelines and risk scores and may help detect undiagnosed cases of dysglycemia in clinical practice.
Study context
- Population
- Patients in clinical practice screened for undiagnosed dysglycemia (prediabetes plus diabetes; HbA1c ≥5.7%); derivation cohort mean age 48 years, 65% female, 42% Hispanic, 32% non-Hispanic Black, mean BMI 32, 29% with hypertensionvalidation n = 519
- Participants
- 11,387 patients
- Comparison
- D-RISK vs American Diabetes Association (ADA) risk test, ADA and United States Preventive Services Task Force (USPSTF) screening guidelines
- Outcome
- Discrimination (AUC), sensitivity and specificity for dysglycemia (HbA1c ≥5.7%)
Methodological detailsderivation and validation, retrospective electronic health record (EHR) derivation sample, prospective screening validation sample, logistic regression with backward selection, points-based risk score
- Derivation and validation
- Retrospective EHR derivation sample
- Prospective screening validation sample
- Logistic regression with backward selection
- Points-based risk score
Performance
As reported in the abstract
AUC for dysglycemia, validation screening study
| Test or model | Value | AUC | Sensitivity |
|---|---|---|---|
| D-RISK | 0.71 (95% CI 0.66–0.75) | 0.71 | 75% |
| ADA risk test | 0.67 | 0.67 | 61% |
| ADA guidelines | 0.52 | 0.52 | |
| USPSTF guidelines | 0.58 | 0.58 |
Value
D-RISK0.71 (95% CI 0.66–0.75)
ADA risk test0.67
ADA guidelines0.52
USPSTF guidelines0.58
AUC
D-RISK0.71
ADA risk test0.67
ADA guidelines0.52
USPSTF guidelines0.58
Sensitivity
D-RISK75%
ADA risk test61%
ADA guidelines— not stated in the abstract
USPSTF guidelines— not stated in the abstract
- D-RISK vs ADA guidelines: P < 0.001 — D-RISK better
- D-RISK vs USPSTF guidelines: P < 0.001 — D-RISK better
Further findings
| Outcome | D-RISK |
|---|---|
| AUC of risk score, derivation cohort | |
| AUC | 0.75 |
Original publication
Derivation and Validation of D-RISK: An Electronic Health Record–Driven Risk Score to Detect Undiagnosed Dysglycemia in Clinical Practice. Diabetes Care. Published 17 January 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →