Invasive arterial pressure

To the Editor: We read with great interest the paper by Holt et al (1) that compared the arterial blood pressure, monitored with an invasive system (radial line) and noninvasive one (oscillometric), with a “gold standard” Doppler technique. The authors conclude that outside the normotensive range, the invasive arterial blood pressure method was close to the Doppler ultrasound measurement in either hypertensive or hypotensive patients. The paper is interesting because the topic is very important in the pediatric intensive care unit; however, we would like to raise some concerns on the invasive arterial blood pressure monitoring. Direct measurement of arterial blood pressure requires the pressure waveform from the cannulated artery to be reproduced accurately on the monitor. Unfortunately, and not surprisingly, this is not always the case and the displayed pressure signal is often significantly artifacted by the measuring device-cardiovascular system coupling (2). The catheter-transducer systems are connected to the patient's circulation via fluid-filled tubing, which may have other components attached (e.g., stopcocks or extensions). System reliability is best when the catheters and tubing are stiff, the mass of the fluid is small, the number of stopcocks limited, and the length of the connecting tubing is not excessive (3, 4). Even if these conditions are taken into account, dynamic response artifacts are frequently encountered in patients admitted to intensive care units, catheterization laboratories, and operating rooms (2). Consequently, pressure waveforms reproduced on the monitor may not be adequately accurate: they may appear amplified (resonance artifact), with an artifactual increase in systolic blood pressure (overshoot) and decrease in diastolic blood pressure. The signal may also be overdamped, typically due to clots, air bubbles, or kinks in the arterial catheter. Holt et al (1) evaluated the quality of the invasive measurement by using morphologic criteria: the waves were considered as good when they showed the systolic upstroke, systolic peak pressure, systolic decline, dicrotic notch, diastolic run off, and end-diastolic pressure. When artifacts occur, morphologic criteria may be inappropriate since a resonating signal could be very similar to a nonresonating one, especially when a “narrow” systolic decline is present (2). Therefore, the discrimination between the presence and absence of resonance artifacts should be investigated with the fast flush test as described by Gardner et al (3, 4) in 1981. In 2011, we published a study (5) in which we observed that, in >30% of patients undergoing vascular surgery, systolic arterial pressure may significantly be overestimated (up to 80 mm Hg). In conclusion, we agree with Holt et al that “reliance on [arterial pressure monitoring] may seriously underestimate the severity of hypertension and hypotension in pediatric intensive care unit patients, potentially leading to undertreatment”; however, artifacts due to inadequate damping level are frequent and overlooked phenomena that, once identified, should be addressed to definitely improve invasive blood pressure monitoring reliability. The authors have not disclosed any potential conflicts of interest. Stefano Romagnoli, MD, Heart and Vessels Department, Cardio-Vascular Anesthesia and Post-Surgical Intensive Care Unit, Azienda Ospedaliero-Universitaria Careggi, Florence, Italy; Zaccaria Ricci, MD, Department of Pediatric Cardiac Surgery, Bambino Gesù Children's Hospital, Rome, Italy; Angelo Raffaele De Gaudio, MD, Department of Intensive Care, Azienda Ospedaliero-Universitaria Careggi, Florence, Italy