Background: In this study, we tested the reliability of a non-invasive finger-cuff-based continuous arterial blood pressure monitoring device (Nexfin, BMEYE, Amsterdam, NL) in critically ill surgical patients.
Methods: Invasive intra-arterial and non-invasive arterial pressure measurements from 25 patients during a 4-h period were compared at five time points. Correlation and linear regression analysis were used and mean bias, precision [sd of bias] and limits of agreement (LOA) [bias (2.0 sd)] were calculated using the Bland-Altman method.
Results: Eight data pairs were excluded because of error message from the non-invasive technique, and thus a total of 117 data pairs were analysed. Overall, correlation between mean arterial pressure (MAP) was r(2)=0.50. Bias, precision, and LOA between invasive and non-invasive MAP were 6 (12) and -18 to +30 mm Hg. In patients requiring norepinephrine (83 data pairs), correlation was r(2)=0.28 and bias, precision, and LOA were 6 (13) and -20 to +32 mm Hg, whereas in patients not receiving norepinephrine (34 data pairs) r(2) was 0.80 and mean bias, precision, and LOA were 6 (11) and -16 to +28 mm Hg. In patients with peripheral oedema (49 data pairs), r(2) was 0.40 and mean bias, precision and LOA were 7 (15) and -23 to +37 mm Hg. In patients without oedema (64 data pairs), r(2) was 0.66 and mean bias, precision, and LOA were 5 (9) and -13 to +23 mm Hg.
Conclusions: Non-invasive blood pressure monitoring with Nexfin does not seem to be sufficiently accurate to replace intra-arterial invasive blood pressure measurements in critically ill patients.
Keywords: equipment-Finapres; intensive care; monitoring-arterial pressure.