Ultrasound guidance for arterial (other than femoral) catheterisation in adults

Background Arterial vascular access is a frequently performed procedure, with a high possibility for adverse events (e.g. pneumothorax, haemothorax, haematoma, amputation, death), and additional techniques such as ultrasound may be useful for improving outcomes. However, ultrasound guidance for arterial access in adults is still under debate. Objectives To assess the effects of ultrasound guidance for arterial (other than femoral) catheterisation in adults. Search methods We searched CENTRAL, MEDLINE, Embase, LILACS, and CINAHL on 21 May 2021. We also searched IBECS, WHO ICTRP, and ClinicalTrials.gov on 16 June 2021, and we checked the reference lists of retrieved articles. Selection criteria Randomised controlled trials (RCTs), including cross-over trials and cluster-RCTs, comparing ultrasound guidance, alone or associated with other forms of guidance, versus other interventions or palpation and landmarks for arterial (other than femoral) guidance in adults. Data collection and analysis Two review authors independently performed study selection, extracted data, assessed risk of bias, and assessed the certainty of evidence using GRADE. Main results We included 48 studies (7997 participants) that tested palpation and landmarks, Doppler auditory ultrasound assistance (DUA), direct ultrasound guidance with B-mode, or any other modified ultrasound technique for arterial (axillary, dorsalis pedis, and radial) catheterisation in adults. Radial artery Real-time B-mode ultrasound versus palpation and landmarks Real-time B-mode ultrasound guidance may improve first attempt success rate (risk ratio (RR) 1.44, 95% confidence interval (CI) 1.29 to 1.61; 4708 participants, 27 studies; low-certainty evidence) and overall success rate (RR 1.11, 95% CI 1.06 to 1.16; 4955 participants, 28 studies; low-certainty evidence), and may decrease time needed for a successful procedure (mean difference (MD) -0.33 minutes, 95% CI -0.54 to -0.13; 4902 participants, 26 studies; low-certainty evidence) up to one hour compared to palpation and landmarks. Real-time B-mode ultrasound guidance probably decreases major haematomas (RR 0.35, 95% CI 0.23 to 0.56; 2504 participants, 16 studies; moderate-certainty evidence). It is uncertain whether real-time B-mode ultrasound guidance has any effect on pseudoaneurysm, pain, and quality of life (QoL) compared to palpation and landmarks (very low-certainty evidence). Real-time B-mode ultrasound versus DUA One study (493 participants) showed that real-time B-mode ultrasound guidance probably improves first attempt success rate (RR 1.35, 95% CI 1.11 to 1.64; moderate-certainty evidence) and time needed for a successful procedure (MD -1.57 minutes, 95% CI -1.78 to -1.36; moderate-certainty evidence) up to 72 hours compared to DUA. Real-time B-mode ultrasound guidance may improve overall success rate (RR 1.13, 95% CI 0.99 to 1.29; low-certainty evidence) up to 72 hours compared to DUA. Pseudoaneurysm, major haematomas, pain, and QoL were not reported. Real-time B-mode ultrasound versus modified real-time B-mode ultrasound Real-time B-mode ultrasound guidance may decrease first attempt success rate (RR 0.68, 95% CI 0.55 to 0.84; 153 participants, 2 studies; low-certainty evidence), may decrease overall success rate (RR 0.93, 95% CI 0.86 to 1.01; 153 participants, 2 studies; low-certainty evidence), and may lead to no difference in time needed for a successful procedure (MD 0.04 minutes, 95% CI -0.01 to 0.09; 153 participants, 2 studies; low-certainty evidence) up to one hour compared to modified real-time B-mode ultrasound guidance. It is uncertain whether real-time B-mode ultrasound guidance has any effect on major haematomas compared to modified real-time B-mode ultrasound (very low-certainty evidence). Pseudoaneurysm, pain, and QoL were not reported. In-plane versus out-of-plane B-mode ultrasound In-plane real-time B-mode ultrasound guidance may lead to no difference in overall success rate (RR 1.00, 95% CI 0.96 to 1.05; 1051 participants, 8 studies; low-certainty evidence) and in time needed for a successful procedure (MD -0.06 minutes, 95% CI -0.16 to 0.05; 1134 participants, 9 studies; low-certainty evidence) compared to out-of-plane B-mode ultrasound up to one hour. It is uncertain whether in-plane real-time B-mode ultrasound guidance has any effect on first attempt success rate or major haematomas compared to out-of-plane B-mode ultrasound (very low-certainty evidence). Pseudoaneurysm, pain, and QoL were not reported. DUA versus palpation and landmarks DUA may lead to no difference in first attempt success rate (RR 1.01, 95% CI 0.90 to 1.14; 666 participants, 2 studies; low-certainty evidence) or overall success rate (RR 0.99, 95% CI 0.92 to 1.07; 666 participants, 2 studies; low-certainty evidence) and probably increases time needed for a successful procedure (MD 0.45 minutes, 95% CI 0.20 to 0.70; 500 participants, 1 study; moderate-certainty evidence) up to 72 hours compared to palpation and landmarks. Pseudoaneurysm, major haematomas, pain, and QoL were not reported. Oblique-axis versus long-axis in-plane B-mode ultrasound Oblique-axis in-plane B-mode ultrasound guidance may increase overall success rate (RR 1.27, 95% CI 1.05 to 1.53; 215 participants, 2 studies; low-certainty evidence) up to 72 hours compared to long-axis in-plane B-mode ultrasound. It is uncertain whether oblique-axis in-plane B-mode ultrasound guidance has any effect on first attempt success rate, time needed for a successful procedure, and major haematomas compared to long-axis in-plane B-mode ultrasound. Pseudoaneurysm, pain, and QoL were not reported. We are uncertain about effects in the following comparisons due to very low-certainty evidence and unreported outcomes: real-time B-mode ultrasound versus palpation and landmarks (axillary and dorsalis pedis arteries), real-time B-mode ultrasound versus near-infrared laser (radial artery), and dynamic versus static out-of-plane B-mode ultrasound (radial artery). Authors' conclusions Real-time B-mode ultrasound guidance may improve first attempt success rate, overall success rate, and time needed for a successful procedure for radial artery catheterisation compared to palpation, or DUA. In addition, real-time B-mode ultrasound guidance probably decreases major haematomas compared to palpation. However, we are uncertain about the evidence on major haematomas and pain for other comparisons due to very low-certainty evidence and unreported outcomes. We are also uncertain about the effects on pseudoaneurysm and QoL for axillary and dorsalis pedis arteries catheterisation. Given that first attempt success rate and pseudoaneurysm are the most relevant outcomes for people who underwent arterial catheterisation, future studies must measure both. Future trials must be large enough to detect effects, use validated scales, and report longer-term follow-up.