<h3>BACKGROUND CONTEXT</h3> Incidental durotomies are a relatively infrequent occurrence during lumbar spine surgery and are generally benign, without significant adverse consequences. Treatment with direct repair, with or without augmentation, is generally very effective and further cerebrospinal fluid (CSF) leak is uncommon. The strength and stability of a dural repair is dependent on several factors, including patient age, quality of dural tissue, the size and location of the durotomy, and the manner in which the repair is performed. There is limited in vitro evidence comparing the strength of different dural repair techniques and the resulting likelihood of recurrent CSF leak. <h3>PURPOSE</h3> To compare the strength and stability of various dural repair techniques in an in vitro human cadaveric durotomy model. <h3>STUDY DESIGN/SETTING</h3> An in vitro testing device was created to apply controlled, hydrostatic pressure to evaluate the integrity of a surgically-created durotomy. <h3>PATIENT SAMPLE</h3> Sixteen human cadaveric specimens. <h3>OUTCOME MEASURES</h3> The hydrostatic pressure required to cause a leak at the site of the dural repair. <h3>METHODS</h3> Complete human cadavers were placed in the prone position and the dura in the thoracic and lumbar spines was exposed through a midline incision and subsequent laminectomies. The dura was clamped 8 cm proximal and distal to the level of the durotomy and clamped with reliable, water-tight plastic rings. A 16-gauge plastic catheter was introduced into the lateral aspect of the dura and sutured in a water-tight fashion to serve as the source for instilling saline into the dura. The catheter was attached to a hydrostatic pressure measuring column to assess the intradural pressure. Linear durotomies of either 1 cm or 6 cm in length were made with a 15 blade in both the thoracic and lumbar spines of each cadaver. A dural repair was performed with either a 6-0 prolene or 2-0 silk suture. Repair technique were either multiple, simple interrupted sutures or a simple running repair with knots at the respective margins of the durotomy. An analysis of variant analysis (ANOVA) was employed to assess differences in leak rates with respect to applied hydrostatic pressure. <h3>RESULTS</h3> Eight repair groups (thoracic vs lumbar / 6-0 prolene vs 2-0 silk / running vs multiple interrupted) were studied for each of the 1 cm and 6 cm durotomies. There were no significant differences noted in the frequency or extent of spinal fluid leak at different hydrostatic pressures with respect to either suture type employed or suture technique performed. There was a slight difference in the mean applied hydrostatic pressure applied to create a fluid leak in the 6 cm durotomy (16 cm H2O) compared to the 1 cm durotomy (12 cm H2O) but this did not reach statistical significance. <h3>CONCLUSIONS</h3> In this in vitro study of the integrity of the repair of experimentally created durotomies, no significant difference was noted with respect to different suture type or technique of repair. Larger durotomies had a slight, not significant trend for recurrent leak at lower applied hydrostatic pressures. In general, expansion of the repaired dura with controlled delivery of intrathecal fluid, simulating an in vivo, intraoperative Valsalva maneuver, did not reveal significant differences in varying repair types. While this in vitro study does not directly replicate an in vivo scenario, the results do reinforce the necessity of a meticulous water tight suture repair to adequately treat incidental durotomies and thus avoid the troublesome consequences of a recurrent CSF leak. <h3>FDA DEVICE/DRUG STATUS</h3> This abstract does not discuss or include any applicable devices or drugs.