Accuracy of invasive arterial pressure monitoring in cardiovascular patients: an observational study.

INTRODUCTION

Critically ill patients and patients undergoing high-risk and major surgery, are instrumented with intra-arterial catheters and invasive blood pressure is considered the "gold standard" for arterial pressure monitoring. Nonetheless, artifacts due to inappropriate dynamic response of the fluid-filled monitoring systems may lead to clinically relevant differences between actual and displayed pressure values. We sought to analyze the incidence and causes of resonance/underdamping phenomena in patients undergoing major vascular and cardiac surgery.

METHODS

Arterial pressures were measured invasively and, according to the fast-flush Gardner's test, each patient was attributed to one of two groups depending on the presence (R-group) or absence (NR-group) of resonance/underdamping. Invasive pressure values were then compared with the non-invasive ones.

RESULTS

A total of 11,610 pulses and 1,200 non-invasive blood pressure measurements were analyzed in 300 patients. Ninety-two out of 300 (30.7%) underdamping/resonance arterial signals were found. In these cases (R-group) systolic invasive blood pressure (IBP) average overestimation of non-invasive blood pressure (NIBP) was 28.5 (15.9) mmHg (P <0.0001) while in the NR-group the overestimation was 4.1(5.3) mmHg (P < 0.0001). The mean IBP-NIBP difference in diastolic pressure in the R-group was -2.2 (10.6) mmHg and, in the NR-group -1.1 (5.8) mmHg. The mean arterial pressure difference was 7.4 (11.2) mmHg in the R-group and 2.3 (6.4) mmHg in the NR-group. A multivariate logistic regression identified five parameters independently associated with underdamping/resonance: polydistrectual arteriopathy (P = 0.0023; OR = 2.82), history of arterial hypertension (P = 0.0214; OR = 2.09), chronic obstructive pulmonary disease (P = 0.198; OR = 2.61), arterial catheter diameter (20 vs. 18 gauge) (P < 0.0001; OR = 0.35) and sedation (P = 0.0131; OR = 0.5). The ROC curve for the maximal pressure-time ratio, showed an optimum selected cut-off point of 1.67 mmHg/msec with a specificity of 97% (95% CI: 95.13 to 99.47%) and a sensitivity of 77% (95% CI: 67.25 to 85.28%) and an area under the ROC curve by extended trapezoidal rule of 0.88.

CONCLUSION

Physicians should be aware of the possibility that IBP can be inaccurate in a consistent number of patients due to underdamping/resonance phenomena. NIBP measurement may help to confirm/exclude the presence of this artifact avoiding inappropriate treatments.