Although the use of permanent catheters is discouraged by standards and guidelines (e.g., Kidney Disease Outcomes Quality Initiative), 1 about 17% of prevalent US hemodialysis patients and 8% of prevalent European hemodialysis patients depend on central catheters for blood access (Dialysis Outcomes and Practice Patterns Study). 2 The numbers for incident patients are 60% for the United States and 31% for Europe, respectively. Between treatments, the catheters are filled with a “locking solution,” usually containing 1000 to 5000 units of heparin per lumen. An increase of clotting times has been reported after instillation of such heparin locks. 3 The infection rate for these catheters is far higher than the infection rate of fistulas (3–10 vs. < 1 per 1000 days). Recently, antibiotic or antimicrobial catheter locks have been used successfully for treating or preventing catheter related infections. After injection of a gentamicin-citrate lock into catheters, subtherapeutic levels of gentamicin have been found systemically. 4 Ash et al.5 reported on patients who had a “metallic” taste after the precise catheter volume was instilled (47% citrate).
To date, it has been assumed that instillation of a lock volume equal to the catheter filling volume would just fill the catheter lumen and would therefore be safe and effective because no overspill would occur and because the concentration of the instilled substance would be maintained throughout the lumen. Engineers with some knowledge of hydraulics are, of course, aware that flow in a catheter is laminar and that the Hagen-Poiseuille law can be used for estimating the flow distribution. Application of the Hagen-Poiseuille law predicts overspill of the locking solution even when the catheter volume is precisely matched. The Hagen-Poiseuille law, however, is only a rough approximation and does not take the bends in the catheter and the noncontinuous flow during injection into account. We therefore performed measurements to evaluate the overspill volume during injection and the concentration of the active substance in the tip of the catheter.
Materials and Methods
The Dialock blood access device (Biolink Corp, Mansfield, MA) with catheters attached and including the access cannulas was investigated. The Dialock 6 is a subcutaneous blood access device typically implanted in the chest. It is connected to two catheters that are tunneled to a suitable vein, typically the left jugular vein, with the tips placed in the right atrium or close to it. For blood access, the Dialock is accessed with special cannulas that are inserted through the skin.
In a pilot study, a Dialock with a Canaud catheter 7 of 28 cm length was investigated. The Canaud catheter is a silicon catheter with 2 mm inner diameter and six side holes disposed spirally on the last 5 distal centimeters. The inner diameter is constant up to the tip.
The filling volume of the assembly was measured precisely by flushing it with 10 ml demineralized water and finishing with a fluid filled device. Subsequently, the fluid was flushed out with air, and the collected fluid was weighed. The deviation between six consecutive measurements with the catheter positioned horizontally was less than ±0.5%.
For measuring the overspill, the assembly was first completely filled with demineralized water. A precise amount of saline solution of known sodium concentration was then injected, and the overspill fluid was collected. The sodium concentration of the overspill fluid was measured and the relative concentration calculated by dividing the concentration by the concentration of the fluid injected.
In a subsequent study, the Dialock assembly with the Canaud catheter as described above and two proprietary Dialock catheters with different lengths were investigated (Table 1). The proprietary Dialock catheter is a thin wall silicon catheter without any side holes. The wall of the catheter is reinforced with a spirally wound wire. The inner and outer diameters are 2.7 and 3.5 mm, respectively.
The filling volume of each assembly was measured. The system was filled completely with saline, and predetermined volumes of Neutrolin (Biolink Corp., Mansfield, MA) with 1% red dye (FD&C 40, CAS Reg. No.25956–17–6) added were injected at slow speed. The range of injected volumes was lock volume −0.6 ml to +0.4 ml. The spilled effluent was collected, and the volume was measured. The optical absorbance of the effluent was measured with an optical spectrometer at a wavelength of 496 nm, and the concentration relative to the concentration of the instilled lock was calculated. The procedure was repeated three times for each injection volume. Each time, after the complete volume was injected, the last drop at the distal tip was collected and the dye concentration measured.
The result of the pilot study is shown in Figure 1. The filling volume of the assembly was 1.4 ml. As can be seen from Figure 1, instillation of the filling volume results in an overspill of approximately 14%. The filling volume of the Dialock catheter assemblies used in the main study are shown in Table 1.
The results of the overspill tests of the main study are shown in Table 2 (absolute deviations of injection volumes from filling volumes) and Figure 2 (relative deviations of injection volumes from filling volumes). When the precise filling volume was instilled, approximately 15% of the lock solution escaped the catheter. Overspill begins when approximately 80% of the filling volume is injected. The results of the last drop tests are shown in Table 3 (absolute deviations of injection volumes from filling volumes) and Figure 3 (relative deviations of injection volumes from filling volumes).
When the catheter filling volume is injected, the mean concentration of the tip fluid is approximately 90% of the locking solution’s concentration. To achieve full strength of the locking solution at the catheter tip, the assembly must be overfilled by approximately 20%. At the condition of no escape (underfill −20%), the tip concentration decreases to 40%.
Theoretically, the results of both tests should follow a smooth curve. The scatter seen in the graphs indicates the tolerance of the measurement, which are thought to be introduced by the sampling method.
The Dialock catheter assembly has been specially designed for achieving a smooth flow path, resulting in laminar flow in the catheter. This is the reason why the results for the Canaud catheter did not deviate from the results of other catheters when normalized for filling volumes, although the Canaud catheter has side holes at the end. The results may not be representative for catheters with side holes and conical flow paths at the catheter tip. In these cases, fluid is forced out of the side holes, which is caused by the flow restriction in the catheter tip. The loss of locking solution could be considerably larger in these cases.
The test was performed at room temperature. Because catheter locking solutions are at room temperature when injected and thermal equilibration with body temperature takes at least several seconds, we did not expect any deviation from the clinical situation. From a theoretical standpoint, speed of injection and viscosity of the locking solution should not change the results considerably, although tests remain to be performed to confirm this. Viscosity of the locking solution may, however, influence the escape of catheter locking solution in the few minutes after injection, especially when the catheter contains side holes. 8
The results show that, at the time of injection, approximately 15% of the locking solution escapes the catheter if an amount equal to the filling volume of the catheter is injected. This has consequences for the safety and efficacy of catheter locks. If heparin is used as catheter lock, approximately 2000 IU are injected into the patient when both sides of a two lumen catheter are filled with 5000 IU each. This will cause an increased bleeding risk for the next 12 to 24 hours depending on the half-life of the heparin used. Karaaslan et al.3 have reported an increase of the activated partial thromboplastin time (aPTT) ratio immediately after injecting the heparin lock, although the injection volume was matched to the filling volume. For concentrated sodium citrate solutions, patients have reported sensations when the precise lock solution was injected. 5 Over-injection resulted in the death of one patient. 9 If antibiotics are used prophylactically for lock solutions either alone or in combination with an anticoagulant, the concentration of the lock solution must be carefully chosen to avoid development of bacteria resistance because of systemic infusion.
The mean concentration at the catheter tip is reduced immediately after injection. Because the reason for this effect is assumed to be the Hagen-Poiseuille flow velocity distribution, the concentration of the locking substance will be high at the axis but low at the wall of the catheter. Because biofilm formation as well as clotting occurs at the catheter surface, the reduced concentration at the wall of the tip may increase the likelihood of clotting and infection. This effect can only be avoided by overfilling the catheter, which increases the risks mentioned earlier, unless the locking solution does not cause adverse effects when systemically injected.
The conclusion from this finding is that catheter locking solutions should be used that do not cause adverse systemic effects, and catheters should be overfilled by approximately 20%. Neutrolin, which is currently available only in Europe, complies with this requirement. This solution is safe as shown by risk analysis and clinically effective in reducing catheter related infections. 10 Activity against the most common bacteria that form biofilms in catheters has also been demonstrated in vitro. 11
1. National Kidney Foundation: K/DOQI clinical practice guidelines for vascular access, 2000. Am J Kidney Dis
37: S137–S181 (Suppl 1), 2001.
2. Pisoni RL, Young EW, Dykstra DM, et al: Vascular access use in Europe and the United States: results from the DOPPS. Kidney Int 61: 305–316, 2002.
3. Karaaslan H, Peyronnet P, Benevent D, et al: Risk of heparin lock-related bleeding when using indwelling venous catheter in haemodialysis. Nephrol Dial Transplant 16: 2072–2074, 2001.
4. Lynn RI: Reduced incidence of hemodialysis (HD) catheter related bacteremia (CRB) using a weekly citrate-gentamicin lock (CGL). J Am Soc Nephrol
5. Ash SR, Mankus RA, Sutton JM, et al: Concentrated sodium citrate (23%) for catheter lock. Hemodial Int 4: 22–31, 2000.
6. Levin NW, Yang PM, Hatch DA, et al: New access device for hemodialysis. ASAIO J 44: M529–M531, 1998.
7. Canaud B, Beraud JJ, Joyeux H, Mion C: Internal jugular vein cannulation with two silicone rubber catheters: a new and safe temporary vascular access for hemodialysis. Thirty months’ experience. Artif Organs
10: 397–403, 1986.
8. Polaschegg HD, Sodemann K, Estabrook B: Investigation of the effect of catheter side holes on flow properties and outwash of the locking solution, in Proceedings of the 2nd Symposium on Angioaccess for Hemodialysis
, Maastricht, The Netherlands: Vascular Access Society, 1999. p. 195.
9. Food and Drug Administration: FDA issues warning on triCitrasol dialysis catheter anticoagulant. FDA Talk Paper T00–16, 14 April 2000.
10. Sodemann K, Polaschegg HD, Feldmer B. Two years’ experience with Dialock and CLS TM (a new antimicrobial lock solution). Blood Purif 19: 251–254, 2001.
11. Shah CB, Mittelmann MW, Costerton JW, et al: Antimicrobial activity of a novel catheter locking solution. Antimicrob Agents Chemother 46: 1674–1679, 2002.