DDX-BRO
Trial question
What is the role of computerized diagnostic decision support systems in patients presenting to the emergency department?
Study design
Multi-center
Double blinded
RCT
Population
Characteristics of study participants
51.0% female
49.0% male
N = 1204
1204 patients (613 female, 591 male).
Inclusion criteria: patients presenting to the emergency department with abdominal pain, fever of unknown origin, syncope, or non-specific symptoms.
Key exclusion criteria: admitted with an acute life-threatening condition; presenting with a worsening pre-existing condition; pregnancy; medical referral with a definite diagnosis.
Interventions
N=559 CDDSS (diagnoses with the aid of computerized diagnostic decision support systems [Isabel Pro]).
N=645 usual care (diagnosis without the aid of computerized diagnostic decision support systems).
Primary outcome
Diagnostic quality risk
18%
18%
18.0 %
13.5 %
9.0 %
4.5 %
0.0 %
CDDSS
Usual
care
No significant
difference ↔
No significant difference in diagnostic quality risk (18% vs. 18%; OR 0.96, 96% CI 0.71 to 1.3).
Secondary outcomes
No significant difference in diagnostic discrepancy (7% vs. 9%; OR 0.73, 95% CI 0.47 to 1.13).
No significant difference in the rate of unscheduled medical care within 14 days (12% vs. 11%; OR 1.21, 95% CI 0.84 to 1.74).
No significant difference in length of emergency department stay (4.52 hours vs. 4.7 hours; IRR 0.96, 96% CI 0.91 to 1.01).
Safety outcomes
No significant difference in serious adverse events.
Conclusion
In patients presenting to the emergency department with abdominal pain, fever of unknown origin, syncope, or non-specific symptoms, CDDSS was not superior to usual care with respect to diagnostic quality risk.
Reference
Wolf E Hautz, Thimo Marcin, Stefanie C Hautz et al. Diagnoses supported by a computerised diagnostic decision support system versus conventional diagnoses in emergency patients (DDX-BRO): a multicentre, multiple-period, double-blind, cluster-randomised, crossover superiority trial. Lancet Digit Health. 2025 Feb;7(2):e136-e144.
Open reference URL