Table 2 Efficacy and safety outcomes
From: Blinded, randomized trial of sonographer versus AI cardiac function assessment
Outcome | AI (n = 1,740) | Sonographer (n = 1,755) | Mean difference (95% confidence interval) | P value |
|---|---|---|---|---|
Primary efficacy outcome: initial versus final assessment | ||||
Substantial change | 292 (16.8%) | 478 (27.2%) | −10.5% (−13.2% to −7.7%) | <0.001a |
Mean absolute difference | 2.79 ± 5.53 | 3.77 ± 5.22 | −0.97 (−1.31 to −0.61) | <0.001 |
Key secondary safety outcome: final versus previous cardiologist assessment | ||||
Substantial change | 871 (50.1%) | 957 (54.5%) | −4.5% (−7.8% to −1.2%) | 0.008 |
Mean absolute difference | 6.29 ± 5.94 | 7.23 ± 6.18 | −0.94 (−1.34 to −0.54) | <0.001 |
Other secondary outcomes | ||||
Sonographer time (s), median (IQR) | 0 (0–0) | 119 (77–173) | −131 (−134 to −127) | <0.001 |
Cardiologist time (s), median (IQR) | 54 (31–95) | 64 (36–108) | −8 (−12 to −4) | <0.001 |
Any change | 1,100 (63.2%) | 1,218 (69.4%) | −6.2% (−9.3% to −3.1%) | <0.001 |
