The current prospective, randomized, observational study was approved by the Institutional Review Board of Henan Provincial People's Hospital in China (no. 2020-16; 20 January 2020) and was registered in the Chinese Clinical Trial Registry (ChiCTR200030416; 1 March 2020).
Participants and randomization
Patients of either sex, aged from 40-65 years, ASA status of I-III,with a BMI between 20-35 kg/m-2 and who required continuous arterial pressure monitoring during their scheduled surgery were eligible to participate. Patients who had inflamed skin near the puncture site; patients in whom the skin colour did not return to normal within 10 s after Allen's test; patients who had peripheral arterial disease, abnormal coagulation function, recent arterial puncture <1 month earlier, hypertension and diabetes; and patients requiring emergency surgery were excluded from the study.
The patients were assigned by a randomized block design to three groups. We allocated patients at a 1:1:1 ratio with a computer-generated list of random numbers in blocks of three, with the results accessible to only research nurses. The sealed envelopes were opened immediately before the procedure. To prevent the anaesthesia residents’ level of experience with the techniques from confounding the results, an allocation block was used to assign individual anaesthesia residents to each study procedure. Operators who participated in the study completed from one to three blocks.
The operators selected were anaesthesia residents with no more than one year of experience in blind palpation for radial artery cannulation and who previously performed ultrasound-guided radial artery cannulation fewer than five times in patients.
Training phase
Prior to the start of the study, we provided training to all participants, including a theoretical explanation of the anatomy of the radial artery and the operating processes for the three approaches for ultrasound-guided cannulation of the radial artery, as well as practice on a model of radial artery cannulation for approximately 2 hours. One teacher who was familiar with the three approaches demonstrated the three methods for everybody in steps, and the students practised under the guidance of the teacher to ensure that each resident was familiar with the operation steps. At the end of the training session, all participants had completed radial artery catheterization with the three different approaches, and the first-attempt success rate and satisfaction rate were recorded.
The radial artery puncture model was made of a rectangular, soft silicone material with excellent ultrasonic transmission qualities. The model had an electronic pump to keep the fluid flowing, when the needle penetrated the blood vessel, there was flashback of blood in the tail of the needle (Figure 1).
Operation method
Upon arrival to the operating room, the patients underwent oxygen saturation, electrocardiogram and non-invasive blood pressure monitoring, inhaled oxygen for 2 L/min.
Positioning measurement
In the supine position, the operator selected the side of the hand according to the surgical indication. The arm was abducted 90°, and a sterile pad was placed on the wrist to keep the back of the wrist at approximately 45°[8]. The probe was placed perpendicular to the radial artery, and the radial styloid process was the measuring point. Once the radial artery was identified, the cross-sectional view of the artery was frozen and saved to measure the diameter and depth of the artery. The same ultrasound device was used for the measurements (SONIMAGE HS1, KONICA MINOLTA, China),and the device had a linear transducer (5 to 13 MHz and depth of 1 cm). All the data were measured and recorded by the same person.
Technique of radial artery cannulation
After sterile skin preparation, radial artery cannulation was performed according to the grouping method in the sealed envelope.
In the SAX group, the probe was placed perpendicular to the course of the artery, and the needle was directed perpendicular to the longitudinal axis of the probe. In the LAX group, the probe was placed parallel to the course of the artery, and the needle was directed parallel to the longitudinal axis of the probe. In the OAX group, the probe was positioned transversely perpendicular to the artery as in the SAX group and then rotated clockwise in situ by 60° according to maximum visualization. After the arterial image was obtained, local anaesthesia was performed with 2% lidocaine. A 22-G needle ( Sandy Ut 84070, USA) was inserted at approximately 15-45° into the skin until the tip was seen on the screen. After it penetrated the blood vessel and there was blood confluence in the tail of the needle, the position of the needle was adjusted and gently advanced slightly more distally under ultrasound guidance. Then, the ultrasound probe was placed down, the procedure was continued, and whether there was a waveform on the monitor was observed.
Outcome evaluation
The anaesthesia residents performed radial artery cannulation only once for each patient. Successful arterial cannulation was defined as the appearance of an arterial waveform after the catheter was connected to the transducer.The first attempt success rate was the rate of anaesthesia resident's first attempt on the patient and model using each apprach.The time needed for successful arterial cannulation in all patients was measured. The interval between the probe contacting the skin and the appearance of the arterial waveform was recorded as the total time of arterial cannulation. The interval between the probe contacting the skin and the disposable arterial needle contacting skin was recorded as the imaging time. The time from the needle contacting the skin to the first flashback of blood was recorded as the blood return time. The time from the first flashback of blood to the appearance of the arterial waveform on the monitor was recorded as the cannulation time. A total arterial cannulation time of more than 5 minutes and the presence of a non-arterial waveform were regarded as cannulation failure.
All the procedures were carried out under the guidance of an experienced anaesthesiologist. All failed procedures were completed by an expert anaesthesiologist. The primary outcome was the success rate of cannulation, and the secondary outcomes included the first-attempt success rate of cannulation in the model and patients (the success rate of the first operation for each approach), the imaging time, blood return time, cannulation time, total time and the incidence of adverse events. The adverse events included local haematoma and vasospasm. At the end of the study, the operator's level of satisfaction with the three methods was recorded. All data were recorded by the same anaesthesiologist who was not involved in the trial.
Statistical analysis
Sample size:
To calculate the sample size, 10 anaesthesia residents performed the procedure in 30 patients for a pilot trial. The patients were randomly divided into 10 blocks of three, and every block was assigned to individual anaesthesia residents using each technique. The success rate was 50% for LAX, 80% for SAX, and 60% for OAX. We calculated that a total of 187 patients were needed for the three groups to compare these groups with 80% statistical power at a Bonferroni-adjusted P value of 0.017. Considering a 10% drop-out rate, the final sample size of this study was 207, and 69 patients were needed for each group.
The data were recorded using IBM SPSS 23.0. Only patients in whom the needle could not be seen on ultrasound were excluded. The normally distributed continuous data are presented as the mean ± SD, whereas the nonnormally distributed continuous data are presented as the median (interquartile range). Percentages were used to present the nominal and ordinal data, and the Kruskal–Wallis test was used for subsequent analysis. For the detection of significant differences among the three techniques, 1-way analysis of variance with Tukey’s post hoc analysis or the Kruskal–Wallis H test was used as appropriate. Logistic regression analysis with a stepwise forward selection model was used to identify the independent covariates that were associated with the success rate. A P value of less than 0.05 was considered significant.