The central venous catheter (CVC) is commonly used in intensive care unit (ICU). The primary complications associated with CVCs especially with prolonged use include thrombosis and infections. Hence, it is essential to remove the CVC as soon as it becomes unnecessary.

Peripheral intravenous cannulation (PIVC) on a critically ill patient can be a significant challenge for nurses. After several days in ICU, patients may develop significant edema in the upper limbs, complicating the PIVC.

Near-infrared light devices (NILD) are medical devices that use near-infrared light to highlight the patient's peripheral venous network directly on their skin. The advantage of this device is its minimal training and ease of use for effective application. Nurses can use this medical device without specific conditions once they have received training on its use. The vein illuminator has not been extensively studied in ICU.

This study aims to compare two techniques for PIVC in critically ill patients with existing CVC for whom maintaining the deep venous access is no longer indicated.

The investigators hypothesize that the use of the NILD would increase the success rate of first-attempt PIVC insertion compared to a landmark approach (traditional method) for PIVC in ICU.

This is a multicenter randomized controlled study. 380 patients will be randomly assigned with a 1:1 ratio to landmark approach (standard group) or near-infrared light device (interventional group).

Randomization will be done immediately after the enrollment of the patient. Randomization will be stratified by center and according to anticipated difficult venous access defined by:

• BMI > 30 Kg/m2
• Increase in body weight between admission and day of inclusion > 10%.
• Clinically evident edema of the upper limbs (with pitting)
• Absence of visible or palpable veins on arms and forearms If one or more items are present, the patient is classified as difficult to venous access.

Once the patient has been enrolled and randomized, the nurse in charge of the patient performs the PIVC according to the randomization group as soon as possible after randomization.

The nurse will have a maximum of 3 attempts before handing over to another nurse. The total number of attempts will be 5 with the allocated method (traditional method or NILD).

In the case of failure after 5 initial attempts, the intervention will be considered as a failure. However, a maximum of 5 additional attempts will be allowed with any method on the day of randomization. The success or failure of PIVC will be recorded, as well as the device used. The attempts should be performed within 6 hours after randomization.

If one of the 5 first initial attempts is successful, the study continues until the PIVC is removed (max 7 days according to recommendations) or patient discharge, whichever comes first, in order to evaluate the occurrence of local complications.

Inclusion Criteria:

• ≥18 years old
• Hospitalized in ICU
• with a deep venous catheter (jugular, subclavian or femoral)
• In whom the use of a deep venous line is no longer justified (absence of: vasopressor amines, chemotherapy, parenteral nutrition, hypertonic solutions)

Exclusion Criteria:

• Absolute contraindication or anatomical impossibility to perform a PIVC on the upper limbs
• Patient already included in the study
• Tattoo covering most of both forearms
• Adult subject to a legal protection measure (guardianship, curators, person under court protection)
• Persons deprived of their liberty by a judicial or administrative decision, persons hospitalised without consent and persons admitted to a health or social establishment for purposes other than research.
• Pregnant or breast-feeding woman