Trending Ability and Accuracy of Minimally Invasive Pulse Wave Analysis Devices: A Systematic Review and Meta-Analysis.
Borja Barrachina, Cristina Vinuesa, Ibai Iriarte et al.
✦ AI-curated · Sources linked
In 30 seconds
This systematic review and meta-analysis assessed the accuracy and ability to track cardiac-output changes of minimally invasive pulse wave analysis devices compared with reference techniques. It included 50 studies, with 49 contributing accuracy data and all 50 contributing trending data. Across devices and settings, pooled percentage error was 48.3%; pooled four-quadrant concordance was 72.6% (95% CI, 67.9-77.4), and polar plot ±30° agreement was 63.2% (95% CI, 54.9-71.5). The authors concluded that these devices were not sufficiently accurate or consistent in trending to be interchangeable with reference techniques.
Key findings
- Pooled percentage error across devices and settings was 48.3% (unweighted mean 49.0%, 95% CI, 45.8-52.2).
- Pooled four-quadrant concordance was 72.6% (95% CI, 67.9-77.4); polar plot ±30° agreement was 63.2% (95% CI, 54.9-71.5).
- Heterogeneity was extreme (I 2 >95% throughout), and sensitivity analyses limited to high-quality datasets produced similar overall estimates.
- Percentage error was lower in ICU studies (≈42%) than in noncardiac surgery (≈55%); studies declaring conflicts of interest reported ≈44% versus 51% (P < .05).
Why it matters
Cardiac-output monitoring can inform hemodynamic management, but these pooled results indicate that performance varies substantially and that overall accuracy and trending may be inadequate for treating device readings as interchangeable with reference measurements. The context-dependent findings also highlight the importance of transparent reporting and consistent validation methods when interpreting evidence about monitoring technologies.
What to keep in mind
Extreme heterogeneity across studies limits the consistency of pooled estimates. Favorable results for MostCare and Argos were based on few studies.