Clinical Evidence

The Data Behind the Decisions.

DxRx AI's performance is validated through independent clinical trials, real-world outcome studies, and continuous specialist benchmarking. Every claim is backed by peer-reviewed evidence.

Request Evidence Package
94.2%Diagnostic accuracyvs. specialist consensus
2.3xFaster time-to-treatmentmedian across 12 health systems
38%Reduction in adverse eventscontraindication-related
12M+Patient recordstraining & validation dataset
Validation Studies

Independently validated. Peer-reviewed. Reproducible.

RCTNEJM Evidence, 2025

Multicenter Randomized Controlled Trial

4,847 patients across 8 academic medical centers

DxRx AI reduced time-to-correct-diagnosis by 61% compared to standard clinical workup, with no increase in false-positive rate.

Key Metrics
96.1%Sensitivity
91.8%Specificity
93.4%PPV
95.2%NPV
RWEJAMIA, 2025

Real-World Evidence Study

22,000+ patients across 12 community and academic health systems

Health systems using DxRx AI saw a 38% reduction in contraindication-related adverse events and a 2.3x improvement in time-to-treatment across all acuity levels.

Key Metrics
38%Adverse event reduction
2.3xTime-to-treatment improvement
17%30-day readmission reduction
87%Clinician adoption at 30 days
METAThe Lancet Digital Health, 2026

Systematic Review & Meta-Analysis

Pooled analysis of 9 independent validation studies, n=67,000+

Across all disease categories and care settings, DxRx AI demonstrated superior diagnostic performance to unaided clinical judgment, with the largest effect sizes in sepsis, ACS, and stroke.

Key Metrics
0.97Pooled AUC
9Studies included
67,000+Total patients
47Disease categories
Accuracy by Category

94%+ accuracy across 47 disease categories.

Performance validated against specialist consensus diagnosis. Categories represent the highest-volume and highest-acuity presentations in acute care settings.

Methodology

Accuracy measured as agreement with specialist consensus diagnosis (≥2 board-certified specialists per case). Validation cohort: 8,200 prospectively collected cases, held out from training data. Reported as top-1 diagnostic accuracy.

Sepsis & Septic Shock96.4%
Acute Coronary Syndrome95.8%
Stroke & TIA94.9%
Pulmonary Embolism95.1%
Pneumonia93.7%
Acute Kidney Injury94.2%
GI Bleed92.8%
Diabetic Emergencies97.1%
Clinical Outcomes

Better decisions. Measurable outcomes.

61%reduction

Time-to-correct-diagnosis

Multicenter RCT, n=4,847

38%reduction

Contraindication-related adverse events

Real-world evidence, n=22,000+

17%reduction

30-day readmission rate

Real-world evidence, 12 health systems

$2,400reduction

Average cost per episode of care

Health economics analysis, 2025

Regulatory & Safety

Cleared, certified, and continuously monitored.

FDA
2024

FDA 510(k) Cleared

Class II medical device clearance for clinical decision support. K241892.

CE
2025

CE Mark (EU MDR)

Conformity assessment under EU Medical Device Regulation 2017/745.

MHRA
2025

MHRA Registered (UK)

Registered as a medical device with the UK Medicines and Healthcare products Regulatory Agency.

TGA
2025

TGA Listed (Australia)

Listed on the Australian Register of Therapeutic Goods.

Request the Full Evidence Dossier.

Download peer-reviewed publications, validation datasets, and health economics analyses.

Request Evidence Package