A comparison between visual, robotic, and intracoronary imaging lesion length assessment during robotic-assisted percutaneous coronary intervention.

OBJECTIVES

Inaccurate lesion length assessment can lead to adverse clinical outcomes after percutaneous coronary intervention (PCI). Robotic-assisted PCI (R-PCI) may provide more accurate lesion assessment. The authors aimed to compare the accuracy of visual, robotic, and intracoronary imaging-derived lesion length assessment.

METHODS

Sixty-three patients (69 lesions) who underwent R-PCI using the CorPath GRX (Siemens Healthineers) system were analyzed. The operator provided a visual estimate of lesion length, which was compared to the measurement using the robotic system. In cases in which intracoronary imaging (ICI) was performed (n=21), the visual and robotic estimates were compared against the ICI measurements, which were considered the gold standard.

RESULTS

The median visual length measurement was 23.0 mm (IQR: 18.0-28.0) and the median robotic visual length measurement was 24.0 mm (IQR: 18.7-33.0 mm). In only 10.1% of lesions did the visual and robotic estimates agree. In 66.7% of the lesions, the visual estimate was shorter than the robotic measurement, whereas the visual estimate was longer than the robotic measurement in 23.2% of lesions. Using a threshold of greater than 5 mm too long/short, 22 lesions (32%) were inaccurately measured. Both visual (r=0.82) and robotic (r=0.97) estimates had good correlation with ICI measurements; however, this difference in correlation was statistically significant (P = .004). Visual assessment consistently underestimated lesion length compared with ICI (mean difference -4.2 ± 6.9 mm).

CONCLUSIONS

Visual assessment of lesion length often underestimates true lesion length. R-PCI may provide more accurate and reproducible lesion measurements, with better correlation and agreement with ICI measurements than visual estimation.