Accuracy of injection pressure measurement at peripheral nerves using high-resolution 40 MHz ultrasound in an anesthetized porcine model

Background Fluid injection pressure measurement is promoted as a marker of needle tip position that discriminates between tissue layers. However, clinical ultrasound has insufficient resolution to identify the exact position of the needle tip. Our primary objective was to use 40 MHz ultrasound in anesthetized pigs in order to precisely locate the tip of the needle and measure opening injection pressure in muscle, at epineurium and in subepineurium. Methods We surgically exposed the axillae of four anesthetized pigs. Two operators placed a 40 MHz ultrasound transducer over the pectoral muscle and imaged axillary, median and radial nerves. Injections (0.5 mL) were randomized to in-plane and out-of-plane needle trajectories and flow rates of 1, 6 and 12 mL/min. Results We identified 541 fascicles in 23 nerves. The ratio of fascicle area to nerve area remained constant at similar to 0.30 for all nerves. Axillary nerves were smaller than median and radial nerves, difference in diameter (95% CI) 1.61 (0.87 to 2.36) mm, p<0.001 and 1.59 (0.82 to 2.36) mm, p=0.001, respectively. Axillary nerves had less fascicles per nerve than median nerves, difference 7.63 (2.43 to 12.83) and radial nerves, difference 9.02 (3.64 to 14.40). We visualized the circumneurium and injection within the subcircumneural compartment. Intraneural injection increased nerve area (SD) from 5.7 (2.2) mm(2) to 13.7 (5.5) mm(2), difference 8.0 (5.4-10.6) mm(2), p<0.001. Mean injection pressure was greater in subepineurium compared with muscle, geometric ratio 2.29 (1.30 to 4.10), p<0.001; and greater on epineurium compared with muscle, geometric ratio 1.73 (1.03 to 3.00), p=0.01. Twenty-two out of 23 injections in muscle, 14 out of 23 injections at epineurium and 11 out of 22 injections in subepineurium were <138 kPa (20 psi). Conclusion Needle tip position was not discernible using pressure monitoring. The circumneurium and subcircumneural injection compartment were observed but not intrafascicular injection.