ABSTRACT
Using a new in vitro model and samples of human dura, a number of factors related to spinal needle design and use were examined with respect to their effects on the rate of transdural fluid leak. These included needle size, bevel design, bevel orientation, and angle of approach. Using 25-gauge Quincke needles, a 30 degree approach caused a rate of leak across the dura significantly less than those following 60 degree and 90 degree approaches. A significant increase in leak rate was found with 22-gauge Quincke needles when the bevels were oriented so as to be perpendicular rather than parallel to the long axis of the dura. Also, 22-gauge Whitacre needles caused significantly less leak than did 22-gauge Quincke needles, and 25-gauge Quincke needles produced significantly less leak than 22-gauge Quincke needles. If human dura behaves in vivo as it does in this in vitro model, it would be advantageous to perform lumbar puncture using oblique approaches and small needles with conical tips.
