Continuous noninvasive pulse wave analysis using finger cuff technologies for arterial blood pressure and cardiac output monitoring in perioperative and intensive care medicine: a systematic review and meta-analysis.

Abstract

Finger cuff technologies allow continuous noninvasive arterial blood pressure (AP) and cardiac output/index (CO/CI) monitoring. We performed a meta-analysis of studies comparing finger cuff-derived AP and CO/CI measurements with invasive measurements in surgical or critically ill patients. We calculated overall random effects model-derived pooled estimates of the mean of the differences and of the percentage error (PE; CO/CI studies) with 95%-confidence intervals (95%-CI), pooled 95%-limits of agreement (95%-LOA), Cochran's Q and I² (for heterogeneity). The pooled mean of the differences (95%-CI) was 4.2 (2.8 to 5.62) mm Hg with pooled 95%-LOA of -14.0 to 22.5 mm Hg for mean AP (Q=230.4 [P<0.001], I²=91%). For mean AP, the mean of the differences between finger cuff technologies and the reference method was ≤5±8 mm Hg in 9/27 data sets (33%). The pooled mean of the differences (95%-CI) was -0.13 (-0.43 to 0.18) L min^-1 with pooled 95%-LOA of -2.56 to 2.23 L min^-1 for CO (Q=66.7 [P<0.001], I²=90%) and 0.07 (0.01 to 0.13) L min^-1 m^-2 with pooled 95%-LOA of -1.20 to 1.15 L min^-1 m^-2 for CI (Q=5.8 [P=0.326], I²=0%). The overall random effects model-derived pooled estimate of the PE (95%-CI) was 43 (37 to 49)% (Q=48.6 [P<0.001], I²=63%). In 4/19 data sets (21%) the PE was ≤30%, and in 10/19 data sets (53%) it was ≤45%. Study heterogeneity was high. Several studies showed interchangeability between AP and CO/CI measurements using finger cuff technologies and reference methods. However, the pooled results of this meta-analysis indicate that AP and CO/CI measurements using finger cuff technologies and reference methods are not interchangeable in surgical or critically ill patients. PROSPERO registration number: CRD42019119266.