The upper arm is sometimes inaccessible for arterial pressure (AP) measurement. Placing the automatic cuff at the ankle is a common alternative. However, whether AP measurement at the ankle is reliable is uncertain. Furthermore, it is unknown whether it is necessary to place the patient in a horizontal position to improve the reliability of the measurement, or if leaving the patient in default position (often semi-recumbent) is acceptable. The aim of this study is to answer these questions. Additionally, investigators will evaluate whether a smartphone application that allows for the simple measurement of AP (by gently placing a finger on the camera lens) is a valuable alternative to upper arm measurements.

A question that arises relatively frequently is "when the upper arm is inaccessible, where should the investigators place the automatic cuff to measure AP?". Indeed, surgical intervention involving the upper limb, a wound, a fracture, or vascular access (temporary catheter or arteriovenous fistula) often makes the arm unsuitable for monitoring AP with an automatic cuff. The arterial catheter is not devoid of risks (including the loss of precious time for its insertion) and is rarely an appropriate alternative, at least in the first instance. Thus, the inaccessibility of the upper arm legitimates the frequent placement of the automatic cuff at the ankle. However, the automatic cuff has only been validated for placement at the upper arm, not at the ankle. Surprisingly, the reliability of ankle AP measurements obtained by this positioning of the cuff at the ankle has been little studied: only two studies, using arterial catheter as the reference method, are available, one of which was in the specific population of patients with obesity. These studies reported that the measurement error at the ankle was excessive, but confirmation is necessary.

Patients, especially in acute care settings, are rarely in a strictly horizontal supine position. For various reasons, the critically ill patient is rather in a semi-recumbent position, meaning that the angle between the trunk and the horizontal axis is between 30 and 60°. Thus, the cuff placed at the ankle is no longer at the level of the heart and more precisely of the phlebostatic axis. This could modify the hydrostatic pressure and lead to overestimation of AP compared to a measurement of AP at heart level (on the arm or invasively, for example). Indeed, the impact on AP measurement of the height of the upper arm relative to the heart has already been demonstrated: the vertical displacement of the upper arm changes the measured AP value. Does the angle of elevation of the trunk significantly modify the AP measured at the ankle? If so, and since the measurement at the ankle in the horizontal supine position (0°) tends to underestimate AP, do the different sources of measurement error "cancel out" in the semi-recumbent position? In what position of the trunk is it preferable to measure AP at the ankle? The answers to these questions are uncertain. Indeed, the only study that investigators are aware of (a pilot study) and that specifically addressed this issue did not provide clear answers. This is important since AP is a crucial parameter of monitoring and therefore of therapeutic decision-making. It is remarkable that in acute care settings, little to no attention is paid to trunk elevation when measuring BP at the ankle, and of course, there is no automation of patient positioning (e.g., in a horizontal position) prior to each automatic measurement at the ankle.

Therefore, it appears important to evaluate the accuracy and precision of non-invasive ankle measurements of AP and the impact of patient position on it. This will be the main objective of the study. If these ankle measurements prove to be neither accurate nor precise, having an alternative would be desirable. It is noteworthy that when the arm is inaccessible, the finger often is. AP measurement at the finger is now possible with a simple smartphone. This finger measurement requires calibration with one or two automatic cuff measurements, which should be possible in most cases (the arm is frequently inaccessible for prolonged AP monitoring, but still allows for one or two isolated measurements). So, which is the better alternative site for AP measurement, the finger or ankle? AP measurement using a smartphone has never been evaluated from this perspective. This will be a secondary objective of the study.

Inclusion Criteria :

• Patients having a catheter in the radial artery.
• and with stable AP over a 5-min period (no change in invasive mean AP >10% and no change in vasoactive drugs)

Exclusion Criteria :

• Cardiac arrhythmia.
• Dysfunction of the arterial catheter and/or waveform suggesting an over- or an under-damping.
• Arm circumference > 42 cm (measured at mid-arm).
• Contraindication to AP measurement at the arm (fracture, wound, amputated limb, limb ischemia, infection, phlebitis, history of lymph node dissection, venous access used on this limb and contraindicating inflation, other).
• Contraindication to the horizontal supine position or to the semi-recumbent position, even transiently.
• Patient's health condition requiring urgent care that is incompatible with the study protocol.
• Asymmetry of mean AP between the two upper arms (> 5 mmHg) or inability to assess it.
• Pregnancy.
• Age <18 years old.
• Adult known to be under guardianship or curatorship before inclusion.
• Absence of coverage by the French national health insurance (Sécurité sociale).

Patients of a surgical intensive care unit having an arterial catheter.