J Neurosurg Pediatrics 9:000 000, 9:191 197, 2012 Flow characteristics of cerebrospinal fluid shunt tubing Laboratory investigation Joseph T. Cheatle, M.D., Alexis N. Bowder, B.S., B.A., Sandeep K. Agrawal, Ph.D., Michael D. Sather, M.D., and Leslie C. Hellbusch, M.D. Division of Neurosurgery, Department of Surgery, University of Nebraska Medical Center, Omaha, Nebraska Object. Cerebrospinal fluid shunt systems malfunction for a multitude of reasons, including malpostitioning, obstruction of the ventricular or distal catheter, obstruction of the shunt valve, and catheter disruptions or disconnections. The goal of this study was to examine the hydrodynamic resistance and flow in new and explanted catheters and also in catheters with 1 or 2 straight connectors. Methods. Explanted catheters of multiple lengths, 2-piece catheters, 3-piece catheters, and new catheters were attached to a proximal and distal manometer. A flask with artificial CSF attached to the proximal end provided flow. The flow was allowed to stabilize over 1 hour; then the change in pressure between the proximal and distal end of the catheter was measured. Results. The resistance to flow was calculated for new, never-implanted catheters and compared with the resistance of explanted distal shunt catheters. The resistance of the new catheters was examined after the addition of 1 and 2 straight connectors. Explanted catheters exhibited a slight increase in the resistance to flow of artificial CSF compared with new catheters. Two-piece and 3-piece catheters had a significant increase in resistance to flow compared with new catheters. For all catheters, resistance to flow increased as length increased (new, p = 0.01; explanted, p = 0.009; 1 connector, p = 0.01; 2 connectors, p = 0.03). In this paper, effective diameter is defined as the available crosssectional area of catheter contacted by the artificial CSF. For new and explanted catheters, a decrease in the effective diameter of the catheter was associated with an increase in the resistance to flow of artificial CSF (new, p = 0.1083; explanted, p = 0.0091). However, after the addition of 1 or 2 connectors, an inverse trend was observed: resistance to flow increased with effective diameter. Conclusions. There appears to be some increase in resistance of CSF shunt catheters as they age, altering flow dynamics. In addition, the use of straight connectors within a CSF shunt system increases the resistance to flow of artificial CSF within the shunt system. The increase in resistance appears to be related to the duration of implantation and the length of the catheter and inversely related to the diameter of the catheter. This increase in resistance may be related to sterile shunt malfunction. The addition of straight connectors is associated with a significant increase in resistance in comparison with catheters without connectors (p = 0.005). (http://thejns.org/doi/abs/10.3171/2011.11.peds11255) Key Words cerebrospinal fluid shunt shunt malfunction cerebrospinal fluid dynamics hydrocephalus Cerebrospinal fluid shunting is required when endogenous pathways of circulation and reabsorption of CSF are disrupted. Cerebrospinal fluid shunt system failure is a significant cause of morbidity in patients with hydrocephalus. Shunt system malfunctions can occur due to catheter breakage or disconnection, obstruction of either the ventricular or distal catheter, obstruction of the shunt valve, or malpositioning. As biomechanical devices, CSF shunt systems are inherently prone to failure and infection. In the 1st year after implantation alone, the average failure rate of a CSF shunt is 30%, and reports show that up to 50% of CSF shunt systems fail over time. 6,16 The flow of CSF within a shunt system depends on intraventricular pressure, the diameter and length of the catheter, and the resistance of the shunt system, including the internal resistance in the shunt tubing and the distal pressure at the site of CSF drainage. Distal catheter malfunction is well documented in the literature. Infection, bowel or bladder perforation, pseudocyst formation, volvulus, abdominal wall or diaphragm penetration, umbilical or inguinal hernia formation, scrotal extrusion, and unexplained failure are all described. Late shunt malfunction secondary to smoldering infections, embolization of choroid plexus and leptomeninges, or cellular invasion may contribute to late-term shunt malfunction and changes in resistance to CSF flow. 10 This article contains some figures that are displayed in color on line but in black and white in the print edition. 191
J. T. Cheatle et al. Sterile shunt malfunction is a known cause of shunt failure. Previous studies outline reaction to silicone within the body via a delayed hypersensitivity reaction. 2,3,11 Upon electron microscopy examination, shunts with sterile malfunction have been found to have increased proteinaceous and cellular debris. 10 This reaction may change the resistance to flow of CSF and result in less than optimal shunt function or outright failure. Calcification of the external surface of catheters and microstructure alteration of silicone catheters has been demonstrated in in vitro samples of explanted silicone catheters. 5 The resistance of different shunt valves alone and in series with their catheters has been well documented in the literature. 7 Valves have low intrinsic hydrodynamic resistance. A valve in series with a catheter increases the resistance in the system by 100% 200%, 10 but little is known about how the addition of straight connectors to the distal catheter can change the resistance to flow of the CSF shunt system. To date, no formal studies have been performed on explanted shunt catheters to determine if the resistance changes over the duration of implantation. The distal catheter is responsible for a significant portion of the total resistance, and a significant increase in resistance over the life of a shunt can alter the function of the CSF shunt system. In this study we compare the hydrodynamic resistance properties of explanted distal shunt catheters to identical new distal shunt catheters. In addition, we examine the effects of straight connectors on CSF shunt systems in vitro. Methods New Tubing and Straight Connectors New CSF shunt tubing and straight connectors were donated by Codman. All of the tubing was suitable for implantation without having any defects or being past the expiration date. The shunt tubing from Codman had an inner diameter of 1.0 mm and outer diameter of 2.2 mm and was originally 120 cm in length. The Codman straight connectors with beveled tips were 11.1 mm in length and had an inner diameter of 1.0 mm and outer diameter of 1.9 mm. Used Tubing Under the approval of the institutional review board, distal CSF shunt tubing was obtained from patients in the course of noninfected CSF shunt revision surgery. The tubing was used to study the flow and resistance of used tubing. The manufacturer and model number of the retrieved shunt tubing were identified by visual inspection and confirmed by reviewing operative reports. Subsequent chart review allowed us to determine the length of time the shunt tubing was implanted, as well as the original diagnosis necessitating placement of a CSF shunt system. If an infection was known or suspected at the time of surgery, the shunt tubing was excluded from the study. Retrieved shunt specimens were tested while this study was active between 2004 and 2009. Artificial CSF Artificial CSF was made to closely resemble natural CSF: 17 100-ml aliquots of distilled water were mixed with 126 mm sodium chloride, 3 mm potassium chloride, and 1.25 mm monosodium phosphate until dissolved; 2 mm magnesium chloride and 2 mm of calcium chloride were subsequently added and allowed to dissolve; and 26 mm sodium bicarbonate and 10 mm of dextrose were added last and allowed to dissolve. The artificial CSF was allowed to equilibrate to the temperature of the water bath to closely resemble in vivo conditions. Measurement of CSF Shunt Tubing Resistance A water bath system as pictured in Fig. 1 was maintained at 37 ± 2 C. Manometers were placed at each end of the water bath with a constant height ± 2 mm. A flask with a redundant catheter containing 350 ml of artificial CSF was connected to the proximal CSF drainage catheter and placed in the water bath. The flask and catheter were constant in all experiments. The catheter was of adequate length to ensure that the artificial CSF was equilibrated to the temperature of the water bath. The proximal end of the catheter was attached to the proximal manometer. The distal end of the catheter was then attached to the distal manometer. All air was purged from the system. The pump was set to provide flow of 15 ml per hour. After 1 hour the water column heights of the proximal and distal manometer were read and recorded, as was the total volume of artificial CSF remaining in the beaker. To examine the effects of straight connectors in vitro, the new distal catheter tubing was cut to lengths of 18, 35, 52, 85.7, 93, 96, and 98 cm. These specific lengths were chosen to match lengths of explanted catheters. For each length the resistance was measured using the above procedure for a control catheter (with no connector), a catheter with 1 connector (2-piece catheter), and a catheter with 2 connectors (3-piece catheter). Resistance was measured 5 times for condition (control, 1 connector, and 2 connectors) with catheters that were 18, 35, 52, 85.7, 93, 96, and 98 cm in length. Therefore, for each length 15 trials were performed. Determination of Resistance to Flow The following variables were recorded during the measurement of resistance: starting volume of artificial CSF, ending volume of artificial CSF, total volume of artificial CSF used, height of the water column in proximal Fig. 1. Schematic of shunt flow system within a water bath maintained at 37.0 ± 2.0 C. acsf = artificial CSF. 192
Flow characteristics of CSF shunt tubing manometer, and height of water column in distal manometer. The Poiseuille law states that the volume flow of an incompressible fluid through a circular tube is equal to p/8 times the pressure differences between the ends of the tube, times the fourth power of the tube s radius divided by the product of the tube s length and the dynamic viscosity of the fluid. 9 Using the data obtained and the Poiseuille law, shown in Equation 1, the resistance to flow was calculated for each trial. Equation 1. Using the Poiseuille equation assuming a constant velocity, the resistance to flow for CSF (a Newtonian fluid) through the CSF shunt system was calculated, where r = radius of the catheter, l = length of distal catheter, and h = viscosity of the fluid. A predetermined viscosity of CSF at 0.79 mpa sec 4 was used for the viscosity in each resistance to flow calculation. Q = (P proximal P distal ) π r 4 / 8lη Eq. 1 Effective diameter was calculated by solving the Poiseuille equation for effective diameter, and the resulting equation (Equation 2) was used to calculate the effective diameter for each trial. Equation 2. The following equation was derived from the Poiseuille equation and used to calculate effective diameter (D) in each catheter, where l = length of tube, h = viscosity of the fluid in mpa sec, and Q = resistance to flow of CSF in Pa sec: D = 2 ( η 8l/Q/π) 1/4 Eq. 2 Finally, for each trial the Reynolds number was calculated (Equation 3), and the flow through the shunt system was characterized as either laminar or turbulent. Averages for resistance to flow, effective diameter, and Reynolds numbers and their corresponding standard deviations were calculated for each condition. Equation 3. The following equation was used to calculate the Reynolds number for each trial, where h = viscosity (in centistokes), Q = resistance to flow of CSF (in gallons/minute), and D = effective diameter (in millimeters) 9 : R e = (3160 Q)/(η D) Eq. 3 Results New Versus Explanted Shunts Seven different lengths of shunt tubing were examined in the laboratory (Table 1). The mean resistances of these catheters are illustrated in Fig. 2. The increasing length of catheters is associated with an arithmetic increase in the resistance (p = 0.028) seen in Fig. 2 lower. Another factor regarding resistance is the diameter of the tubing, as seen in Fig. 2 upper (p = 0.028). Here the resistance of the catheter is inversely related to the diameter of the shunt tubing. Table 2 summarizes the data obtained in explanted catheters from 23 patients (age range 1 35 years) who underwent ventriculoperitoneal shunt revisions without any signs of infection. The resistance to flow seen in Fig. TABLE 1: Reynolds number values obtained in new catheters* Shunt Length (cm) Reynolds Number controls (no connectors) New 1 18 3.269977 New 2 35 2.457557 New 3 52 1.575367 New 4 85.7 1.643664 New 5 93 1.022275 New 6 96 0.522794 New 7 98 0.832408 2-piece catheters (1 connector) New 1 18 2.081586 New 2 35 1.74467 New 3 52 1.290886 New 4 85.7 1.150562 New 5 93 0.892396 New 6 96 0.78847 New 7 98 0.800425 3-piece catheters (2 connectors) New 1 18 1.918633 New 2 35 1.460546 New 3 52 1.17843 New 4 85.7 1.109235 New 5 93 0.815843 New 6 96 0.768941 New 7 98 0.620795 * For all measurements involving new catheters, the tubing used was Codman Bactiseal (Ref. 82-3072; SN 980384). 3, upper, demonstrates that length is much more important in the explanted catheters than in new catheters (p = 0.0009), suggesting the internal diameter becomes stenotic with age. The change in resistance with an increase in internal effective diameter is seen in Fig. 3, lower (p = 0.009). Comparison of new shunt tubing and older shunt tubing showed that new tubing has a slightly lower resistance than older tubing (Fig. 4). New Shunts With the Addition of Straight Connectors Seven different lengths of shunt tubing with 1 or 2 connectors were examined in the laboratory. The mean resistances to flow for each length and condition are shown in Fig. 5. The mean resistance to flow increased with length (1 connector, p = 0.01; 2 connectors, p = 0.03). More importantly, the mean resistance to flow increased significantly after the addition of 1 connector or 2 connectors to the shunt system (Fig. 5). A t-test revealed that there was also a significant difference in resistance between 2- and 3-piece catheters (p = 0.005). After the calculations of mean effective diameter were completed for each condition and at each length, the mean resistance to flow for the mean effective diameter was examined. In the control condition, increased effective diameter was associated with decreased resistance to 193
J. T. Cheatle et al. Fig. 2. Line graphs demonstrating how average resistance per trial (average of 5 trials) is affected by the effective diameter and length of new catheters. Upper: The average resistance to flow plotted as a function of effective diameter. In new catheters as length increases and effective diameter decreases, resistance to flow increases. The average resistance to flow measured in Pa sec/m 3 is displayed on the y-axis where E+10 denotes scientific notation (1.00E+10 = 1.00 10 10 ). Lower: The average resistance to flow plotted as a function of length. flow as shown in Fig. 3. In contrast, when effective diameter increased in catheters containing either 1 or 2 straight connectors, resistance to flow increased as well (Fig. 6). To characterize the flow through shunts, Reynolds numbers were used to determine whether flow was laminar or turbulent. Laminar flow indicates smooth, constant flow of a fluid through a system and correlates to Reynolds number values less than 2,300. The calculations of Reynolds numbers were completed for each condition at each length, and the results are displayed in Fig. 7 and Table 1. For all 3 conditions, the Reynolds number indicated laminar flow with values that were well below the cutoff for turbulent flow. In addition, the Reynolds numbers appeared to decrease as the length of the catheters increased. Discussion Cerebrospinal fluid shunts are the mainstay of hydrocephalus treatment. However, dysfunction of CSF shunt systems continues to be a common problem. 1 As with all mechanical implanted devices, shunts are prone to failure, infection, and infiltration by host tissue. Little is known about the changes in resistance associated with aging of shunts or the effect of such changes on shunt function, although it appears that shunt tubing predictably becomes more resistive to flow, possibly contributing to shunt failure. In this study we examined new shunt tubing Fig. 3. Line graphs demonstrating how average resistance for the 23 explanted catheters is affected by the effective diameter and length of each catheter. Upper: The average resistance to flow plotted as a function of length. Lower: The average resistance to flow plotted as a function of effective diameter. As seen with the new catheters in Fig. 2, an increase in the length of explanted catheters corresponds to an increase in resistance to flow. In addition, smaller effective diameters lead to an increase in resistance to flow in explanted catheters. and compared this to explanted tubing removed because of CSF shunt malfunction. Explanted shunt tubing has a slight increase in resistance to flow compared with new tubing. This gradual increase in resistance is of unknown physiological significance, but it may contribute to sterile shunt malfunction. It is possible that during shunt replacements the shunt tubing may be stretched or damaged upon removal. Such damage could possibly have altered the resistance to flow in the explanted shunts that we used in our study, although we very carefully examined the entire visible surface of the shunts before testing. Nevertheless, in situ studies of resistance to flow would not be practical because it would be difficult to obtain multiple trials for each patient and would potentially increase the risk of shunt revision surgery. For incompressible flow within a viscid fluid, the ki- Fig. 4. Bar graph comparing the change in average resistance to flow between new and explanted catheters. 194
Flow characteristics of CSF shunt tubing TABLE 2: Resistance and effective diameter in explanted catheters* Shunt Mean Resistance (Pa sec/m 3 ) SD Mean Effective Diameter (mm) SD Explanted 1 1.88E+10 1.90E+09 1.94 4.65E 02 Explanted 2 7.29E+10 1.65E+10 1.39 7.47E 02 Explanted 3 5.68E+10 4.76E+09 1.50 3.23E 02 Explanted 4 2.49E+10 1.25E+10 1.49 7.93E 02 Explanted 5 1.56E+10 7.75E+09 1.87 2.64E 01 Explanted 6 1.84E+10 2.47E+09 1.90 6.46E 02 Explanted 7 2.80E+10 4.80E+09 1.54 6.35E 02 Explanted 8 4.06E+10 3.75E+10 1.63 3.16E 01 Explanted 9 2.30E+10 1.76E+10 1.34 1.83E 01 Explanted 10 4.16E+10 1.91E+10 1.66 1.92E 01 Explanted 11 1.98E+10 5.84E+09 1.55 1.01E 01 Explanted 12 1.88E+10 1.21E+10 1.81 2.81E 01 Explanted 13 2.62E+10 7.50E+09 1.57 1.21E 01 Explanted 14 3.63E+10 1.16E+10 1.76 1.25E 01 Explanted 15 2.01E+10 1.73E+10 1.50 2.84E 01 Explanted 16 1.82E+10 4.58E+09 1.76 1.06E 01 Explanted 17 5.06E+10 1.74E+10 1.67 1.29E 01 Explanted 18 2.54E+10 3.18E+09 1.66 4.98E 02 Explanted 19 5.76E+10 1.52E+10 1.47 9.40E 02 Explanted 20 2.98E+10 5.05E+09 1.65 6.80E 02 Explanted 21 1.11E+10 6.07E+09 1.57 2.12E 01 Explanted 22 4.18E+10 7.04E+10 1.81 4.62E 01 Explanted 23 8.79E+10 4.44E+10 1.19 2.19E 01 * Values are presented in scientific E notation: 1.00E+10 = 1.00 10 10 ; 1.00E 01 = 1.00 10 1. Fig. 5. Bar graph depicting the average resistance to flow for the control, 2-piece, and 3-piece catheters at 7 different lengths. None, one, and two refer to the number of connectors. *Significant increase in average resistance to flow (p = 0.05). netic and potential energy of a fluid must remain constant. Therefore, when a fluid moves from a region of low pressure to a region of higher pressure, the speed at which it flows must decrease. Cerebrospinal fluid is a Newtonian fluid, and its viscosity is nearly that of water ranging from 0.7 to 1 mpa sec at 37 C. The viscosity of our artificial CSF was found to be 0.79 mpa sec. 4,12 In nonimplanted CSF shunts, CSF flow is laminar. The Reynolds number for each of the explanted catheters remains low, well below the threshold of turbulent or even transitional flow, and therefore the characteristic nature of the flow remains laminar. The change in the resistance to flow, although significant, does not cause turbulent flow. This suggests that a relatively uniform stenosis along the length of the tube occurs rather than a stenotic area with a relatively spared diameter in other areas, as occurs in blood vessels in vivo. 13 The standard shunt system consists of tubing with fixed resistance and a 1-way differential pressure valve. 15,16 The flow of fluid through this system is reliant on both the differential pressure across the shunt and the resistance of the shunt itself. This resistance can be reproducibly set by the manufacturers of the shunt tubing. However, within a shunt system there exists a series of resistors: the opening of the ventricular catheter (Rc), the valve (Rv), any shunt accessories (Ra), and finally, the distal tubing (Rt). Therefore the total resistance of a shunt system can be represented in the following equation: 15 Rs = Rc + Rv + Ra + Rt. In some shunts the values of Rv and Ra are designed to change with position. 14 The changes in resistance to flow after the addition of various shunt valves has been well documented. A valve in series with a catheter increases the resistance in the system by 100% 200%. 10 In 195
J. T. Cheatle et al. Fig. 7. Bar graph depicting the Reynolds number for control, 2-, and 3-piece catheters at 7 different lengths of catheter. As expected all trials had Reynolds numbers indicating laminar flow with values well below the threshold for turbulent flow of 2,300. *Significant difference in Reynolds numbers (p = 0.05). Fig. 6. Line graphs showing how effective diameter affects the average resistance to flow in 2- and 3-piece catheters. Upper: The average resistance for 2-piece catheters plotted as a function of effective diameter. Lower: The average resistance to flow for 3-piece catheters plotted as a function of effective diameter. In contrast to the trends displayed by the control or new catheters in Fig. 3, as internal diameter increased resistance to flow also increased. This suggests that the connectors themselves create an additional source of resistance within a CSF shunt system. addition, a study examining the effect of the use of Luer connector devices in arterial and ventricular catheters showed that these devices exerted an increased resistance to flow within the catheter system. 8 Therefore, our findings that the resistance to flow within CSF shunt systems increases with the addition of 1 or 2 straight connectors is consistent with previous research. If the resistance of the connectors (Rsc 1 and Rsc 2 ) were added to the above equation, one would expect the resistance to flow to be significantly different between shunts of similar length containing either 1 or 2 connectors. This assumption is supported by our findings: Rc + Rv + Ra + Rt + Rsc 1 Rc + Rv + Ra + Rt + Rsc 1 + Rsc 2. Connectors are added to a shunt system for 2 reasons: to repair a break and to lengthen the shunt as the patient grows. 15 Not only must shunts sometimes be lengthened using connectors, but shunt valves must also be adjusted for each patient based on activity level, resistance to flow within their system, and their production of CSF. An increased resistance to flow may contribute to shunt malfunction by causing a slower flow through the entire system. A higher resistance to flow could also result in mildly increased intracranial pressure in shunt-treated patients. Knowing that the addition of connectors to a shunt system results in a significant increase in resistance to flow within a CSF shunt system will enable physicians to consider adjusting the CSF shunt valves accordingly. It should be noted that all of the new shunts and straight connectors used in this experiment were from Codman. The different trends between the control conditions and connector conditions in respect to the effect of diameter on resistance to flow support the conclusion that the connectors create additional resistance within the shunt system. As diameter increases within a viscous fluid system, the resistance to flow naturally decreases because there is more area through which the fluid can flow. 9 This trend was exhibited in the control group. When connectors were added to the CSF shunt system we saw an inverse trend. As effective diameter increased, resistance to flow increased as well. This increase in resistance is due to the presence of a connector within the shunt system. Resistance to flow increased as the viscous CSF fluid was forced through the opening of the straight connectors within the CSF shunt systems. As expected, the Reynolds numbers for all runs were well under the thresholds of turbulent or even transitional flow, and therefore the characteristic nature of the flow remained laminar. Therefore, the change in resistance to flow with the addition of connectors, although significant, does not cause turbulent flow. The changes in the resistance to flow observed in this study represent a preliminary look at the dynamics of a CSF shunt system. In addition, it might be worthwhile to include histological examination using electron microscopy of the explanted shunt tubing to see if proteinaceous debris or other components are contributing to stenosis over time. Conclusions An increase in resistance within a CSF shunt system appears to be directly related to the length of the catheter and inversely related to the effective diameter of the catheter. This increase in resistance may be related to sterile shunt malfunction. The addition of straight connectors shows a significant increase in resistance in comparison with catheters without connectors (p = 0.005). There is also a significant difference between resistances to flow in catheters containing 1 connector versus the resistance to flow in catheters containing 2 connectors (p = 0.005). 196
Flow characteristics of CSF shunt tubing Disclosure New shunt tubing and connectors for this study were donated by Codman. Author contributions to the study and manuscript preparation include the following. Conception and design: Bowder, Agrawal, Sather, Hellbusch. Acquisition of data: Bowder, Agrawal, Satire. Analysis and interpretation of data: all authors. Drafting the article: Bowder, Cheatle, Agrawal, Hellbusch. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Bowder. Statistical analysis: Bowder, Cheatle. Administrative/ technical/material support: Agrawal, Hellbusch. Study supervision: Agrawal, Hellbusch. References 1. Aschoff A, Kremer P, Hashemi B, Kunze S: The scientific history of hydrocephalus and its treatment. Neurosurg Rev 22:67 95, 1999 2. Baldwin CM Jr, Kaplan EN: Silicone-induced human adjuvant disease? Ann Plast Surg 10:270 273, 1983 3. Bass SJ, Gastwirth CM, Green R, Knights EM, Weinstock RE: Phagocytosis of silastic material following silastic great toe implant. J Foot Surg 17:70 72, 1978 4. Bloomfield IG, Johnston IH, Bilston LE: Effects of proteins, blood cells and glucose on the viscosity of cerebrospinal fluid. Pediatr Neurosurg 28:246 251, 1998 5. Boch AL, Hermelin E, Sainte-Rose C, Sgouros S: Mechanical dysfunction of ventriculoperitoneal shunts caused by calcification of the silicone rubber catheter. J Neurosurg 88:975 982, 1998 6. Borgbjerg BM, Gjerris F, Albeck MJ, Hauerberg J, Børgesen SE: Frequency and causes of shunt revisions in different cerebrospinal fluid shunt types. Acta Neurochir (Wien) 136:189 194, 1995 7. Czosnyka Z, Czosnyka M, Richards H, Pickard JD: Hydrodynamic properties of hydrocephalus shunts. Acta Neurochir Suppl 71:334 339, 1998 8. Eloot S, De Vos JY, Hombrouckx R, Verdonck P: How much is catheter flow influenced by the use of closed luer lock access devices? Nephrol Dial Transplant 22:3061 3064, 2007 9. Fox RW, Pritchard PJ, McDonald AT: Introduction to Fluid Mechanics, ed 7. Hoboken JN: John Wiley and Sons, 2005, p 348 10. Gower DJ, Lewis JC, Kelly DL Jr: Sterile shunt malfunction. A scanning electron microscopic perspective. J Neurosurg 61: 1079 1084, 1984 11. Heggers JP, Kossovsky N, Parsons RW, Robson MC, Pelley RP, Raine TJ: Biocompatibility of silicone implants. Ann Plast Surg 11:38 45, 1983 12. Howden L, Giddings D, Power H, Aroussi A, Vloeberghs M, Garnett M, et al: Three-dimensional cerebrospinal flow within the human ventricular system. Comput Methods Biomech Biomed Engin 11:123 133, 2008 13. Ku DN: Blood flow in arteries. Annu Rev Fluid Mech 29: 399 434, 1997 14. Magram G: Cerebrospinal fluid shunt management: Part 1. Neurologist 2:274 287, 1996 15. Magram G, Liakos AM: Cerebrospinal fluid flow through an implanted shunt. Neurol Res 22:43 50, 2000 16. Malm J, Lundkvist B, Eklund A, Koskinen LO, Kristensen B: CSF outflow resistance as predictor of shunt function. A longterm study. Acta Neurol Scand 110:154 160, 2004 17. Sendelbeck L: Recipe for preparation of artificial CSF. Special Delivery 8:4, 1987 Manuscript submitted June 29, 2011. Accepted November 7, 2011. Please include this information when citing this paper: DOI: 10.3171/2011.11.PEDS11255. Address correspondence to: Alexis Bowder, B.S., B.A., Department of Surgery, Division of Neurosurgery, Nebraska Medical Center, Omaha, Nebraska 68198-6250. email: alexis.bowder@unmc. edu. 197