Intraoperative coagulopathy can have various causes, for example, loss, dilution or consumption of coagulation factors, chronic hepatic or renal insufficiency as well as platelet dysfunction. The present study was based on two clinical cases in which patients with acute renal failure who received two Shaldon catheters each for dialysis developed a severe intraoperative coagulation disorder. The Shaldon catheter used in these cases has an internal volume of 1.07 mL specified by the manufacturer and contains seven perforations at its distal part, spaced at a distance of 1 cm (Fig. 1). To avoid intraluminal thrombosis of the catheter it is filled with 1 mL of heparin solution in the dialysis-free interval, based on the manufacturer's recommendation. However, it is tempting to speculate that the procedure might bear the risk of iatrogenic heparin administration subsequently triggering a coagulation disorder. In order to investigate this hypothesis, we measured in an experimental model the overspilled quantity of the recommended volume of a heparin lock. The two cases which stimulated this investigation are briefly described.
A 51-year-old male who had sustained 82% burns in a car accident was admitted to the intensive care unit. Due to severe sepsis, acute renal failure occurred thus requiring renal replacement. This was performed via two single-lumen Shaldon catheters (GamCath®; Joka Kathetertechnik, Hechingen, Germany) placed in the internal jugular vein and the femoral vein. Surgery was required and he received an episode of dialysis. Routine laboratory findings prior to dialysis revealed a platelet count of 35 000 μL−1, APTT 43 s (normal <35 s) and INR 1.1. Following dialysis, both catheters were locked with heparin. The Shaldon catheter located in the internal jugular vein was used for administration of drugs and fluids; the heparin injected to lock the catheter having been withdrawn via aspiration of 20 mL of blood. During the operation unexpected and substantial bleeding occurred. Thrombelastography and clotting studies showed that both the intrinsic and extrinsic pathways of coagulation were markedly disturbed and the APTT exceeded 200 s. Adding heparinase (hep-TEM® reagent; Pentapharm, Munich, Germany) to the patient's blood sample completely reversed the state proving that the coagulopathy was the result of the presence of heparin. Protamine 10 mg (Protamin ICN 1000; ICN Pharmaceuticals, Frankfurt, Germany) was administered and the operation completed without problems.
In the second case, a 58-year-old healthy patient developed severe peritonitis, multi-organ failure and acute renal failure following a laparoscopic cholecystectomy. Two Shaldon catheters were inserted. The platelet count was 27 000 μL−1 and there was a coagulopathy with abdominal bleeding. The patient received 4 units of platelets and 500 000 IU aprotinin (Trasylol®; Bayer HealthCare, Leverkusen, Germany). After dialysis both Shaldon catheters were locked with 1 mL heparin (5000 IU mL−1) each and the patient transferred to the operating theatre for an explorative laparotomy. Soon after starting the operation substantial bleeding occurred. Thromboelastography and coagulation studies now revealed a clotting disorder (APTT > 200 s). Adding heparinase to the blood sample reversed this. Protamine 10 mg was given and the operation continued without problem.
The results of thromboelastography (Rotem®; Pentapharm, Munich, Germany) together with the fact that the coagulation disorder occurred in a relative small time frame suggested an overdose of heparin as cause for the bleeding. It was initially suspected that the anaesthetist has not correctly withdrawn the heparin from the Shaldon catheter before using it but this was excluded. It was therefore suggested that iatrogenic administration of heparin to ‘lock’ the Shaldon catheter might be the cause. This hypothesis was substantiated by recent reports describing similar problems although there was no evidence for the presence of heparin or determination of the quantity of administered heparin [1,2]. The objective of this investigation was to measure the volume which escapes from the Shaldon catheter when filled with 1 mL of a solution according to the manufacturer's specifications. The hypothesis was that a greater amount of an instilled volume and thus more heparin than commonly assumed could escape from the catheter. Additionally, the effect of two arbitrarily chosen instillation velocities was measured to see if there was any effect of injection speed on overspilled volume.
Material and method
A Shaldon catheter was hung up in a small container filled with NaCl solution 0.9% 5 mL (Fig. 2). All seven perforations had to be located beneath the fluid level of the solution. Instead of using a heparin solution, KCl solution 7.46% (1 mol L−1) was injected in the catheter, as it is easier and faster to measure potassium than heparin in a solution. After filling, the Shaldon catheter was closed with a clamp and immediately carefully removed from the container. During this procedure it was very important neither to shake the catheter nor to touch the side of the container to prevent an additional escape of volume. The potassium content of the mixture was measured twice using an ion-selective method . In each case the overspilled volume was calculated as follows:
where E = added volume of KCl solution (1 mol L−1) to 5 mL NaCl solution (0.9%) and ▵ = measured potassium in mmol L−1. The figure of 1000 is a correction factor.
This procedure was repeated 10 times at two injection speeds; 0.5 and 3 s. Speeds were arbitrarily chosen and injection was performed by hand as routinely done in clinical practice. The content of the container of each measurement was diluted 1:3 and potassium concentration reassessed to exclude the theoretical possibility of an incorrect measurement because the produced potassium concentrations were much greater than physiological level.
Analysis was performed using Systat® 9.0 (Systat Software GmbH, Erkrath, Germany). The experiment was repeated 10 times at each injection. Data are presented as mean ± SD. The potassium concentrations were compared with U-test to look for a statistical significant difference between the two injection velocities. P < 0.05 was considered as significant.
We showed that the KCl solution injected into the catheter escaped into the surrounding NaCl solution. A speed dependent pattern was noted. After fast injection 0.38 ± 0.03 mL of the KCl solution escaped (final concentration 69.86 ± 5.26 mmol L−1) and after slow injection 0.51 ± 0.02 mL of the KCl solution (final concentration 92.30 ± 3.55 mmol L−1) escaped. These equate to relative escaped amounts of 37.6 ± 2.9% of rapidly and 50.9 ± 2.0% of slowly injected volume. Projecting these data to the concentration of heparin used for a ‘heparin lock’ (5000 IU mL−1) the amount of heparin administered would be 1879 ± 144 IU during fast injection and 2543 ± 102 IU during slow injection. The difference between the two injection velocities was statistically significant (P = 0.002). Considering the fact that the patients were provided with two Shaldon catheters the total overspilled amount of heparin would have been twice that; 3758 ± 288 IU (rapid injection) or 5086 ± 204 IU (slow injection).
Patients requiring dialysis via a Shaldon catheter may suffer from a substantial coagulation disorder due to heparin lock. This has been demonstrated by two case reports as well as by the experimental data. In clinical practice, heparin lock helps to prevent an occlusion of the Shaldon catheter and thus dysfunction as well as the need to change the catheter frequently. However, it could be clearly shown that a clinically significant amount of heparin may escape from the Shaldon catheter into the circulation leading to relevant coagulopathy. Neither the heparin lock procedure recommended by the manufacturer, nor the appropriate heparin concentration has been clinically examined. To date, the NKF-DOQI (National Kidney Foundation, USA; Dialysis Outcomes Quality Initiative) guidelines still do not present evidence-based recommendations how a catheter thrombosis should be sufficiently prevented . Complying with the manufacturer's recommendation could lead to severe complications especially when dialysis is followed by surgery or other invasive interventions associated with a controlled injury of the patient (e.g., renal biopsy or epidural anaesthesia).
Two patients had developed an iatrogenic coagulation disorder compounded by the presence of two Shaldon catheters and thus a calculated injection of approximately 4000-5000 IU heparin. A similar incidence could be produced by using a heparin locked dialysis catheter for drug administration without withdrawal of the residual heparin which has been excluded in the present cases.
The assessment of the presence and the effect of heparin using thromboelastography helped to identify the problem rapidly and avert further hazards, a technique which is routinely used during liver transplantations [5,6]. The principles of thrombelastography are well known [7,8]. The Rotem® is an easy-to-use, computer-based bed-side coagulation monitoring system where results are displayed in figures and graphics so that the findings can be captured within a few minutes. Value tables and a decision tree lead the user to a recommended treatment of different coagulation disorders. The effect of heparin can be demonstrated by adding heparinase to the patient's blood sample. Heparinase does not affect Rotem® variables in contrast to protamine [9,10]. Difficulties concerning the origin of platelet function impairment (e.g., drug effects or problems in differentiation of single clotting factor deficiencies may confound the issue but in this case did not affect the evidence of a heparin-induced coagulopathy which was reversed with 10 mg of protamine). A low dose of protamine was chosen because the administered amount of heparin was unknown at the time and protamine itself increases the risk of bleeding.
We used KCL solution instead of heparin because we were able to measure very small volumes with greater accuracy than if we had used heparin and a heparin assay. In contrast to the determination of heparin, the potassium content in a solution can be easily and precisely defined. This enabled an exact measurement of even small potassium concentrations which could be used together with a known amount of a solution (5 mL NaCl) for precise calculation of even very small volumes. One difficulty is that the different molecular weights and electric charges of potassium and heparin could interfere with a catheter wall in different ways. This might lead to over- or under-estimation. Under clinical conditions, such interactions are impossible to predict. It has to be pointed out that the results obtained from this study showed clearly an overspillage of an injected solution, but not a precise measurement of an unambiguously defined amount of accidentally administered heparin. Nevertheless, a significant amount of heparin must have left the Shaldon catheter in order to produce such a clinical incidence.
One further problem is the almost impossible determination of fluid distribution within the catheter. If up to 50% of injected fluid volume can escape through the proximal holes of the catheter, 50% of the catheter will remain unfilled. In this case heparin locking does not achieve its purpose to prevent a catheter thrombosis and other designs are required (e.g., the bi-lumen dialysis catheter - VMP Catheter; Joline, Hechingen, Germany).
To prove whether a thrombosis of the distal part of the catheter could occur under clinical conditions, 13 Shaldon catheters which had been used for up to 15 days were examined after removal from the patient by injecting methylene blue. It is quite conceivable that due to extensive tip thrombosis even more heparin can pass into the patient's circulation than the calculated volumes found in this study. A large portion of the solution escaped from the catheter through the first opening (Fig. 3), indicating that the catheters could be thrombosed up to the sixth of seven holes.
In conclusion, following a heparin lock of a Shaldon catheter, a considerable amount of the injected solution is accidentally administered to the patient. As little as 49.1% of the injected volume may remain in the Shaldon catheter. This could lead to an increased risk of coagulopathy. A rapid injection of the heparin solution may be better in order to keep the overspilled volume as low as possible.
1. Karaaslan H, Peyronnet P, Benevent D, Lagarde C, Rince M, Leroux-Robert C. Risk of heparin
lock-related bleeding when using indwelling venous catheter in haemodialysis. Nephrol Dial Transplant
2. Moritz ML, Vats A, Ellis D. Systemic anticoagulation and bleeding in children with hemodialysis catheters. Pediatr Nephrol
3. Meier PC, Ammann D, Osswald HF, Simon W. Review: ion-selective electrodes in clinical chemistry. Med Prog Technol
4. NKF-DOQI clinical practice guidelines for vascular access. National Kidney Foundation - Dialysis Outcomes Quality Initiative. Am J Kidney Dis
1997; 30 (Suppl 3)
5. Pivalizza EG. Heparinase and thromboelastography
in liver transplantation for a patient with von Willebrand's disease. Anesthesiology
6. Harding SA, Mallett SV, Peachey TD, Cox DJ. Use of heparinase modified thromboelastography
in liver transplantation. Br J Anaesth
7. Hartert H. Blutgerinnungsstudien mit der Thrombelastographie, einem neuen Untersuchungsverfahren. Klin Wochenschr
8. Calatzis AN, Haas S, Gödje O et al
. Thromboelastographic coagulation monitoring during cardiovascular surgery with the ROTEG coagulation analyzer. In: Pifarre R, ed. Management of Bleeding in Cardiovascular Surgery
. Philadelphia: Hanley & Belfus, 2000: 215-226.
9. Tuman KJ, McCarthy RJ, Djuric M et al
. Evaluation of coagulation during cardiopulmonary bypass with a heparinase modified thromboelastographic assay. J Cardiothorac Vasc Anesth
10. Despotis GJ, Summerfield AL, Joist JH et al
. In vitro
reversal of heparin
effect with heparinase: evaluation with whole blood prothrombin time and activated partial thromboplastin time in cardiac surgical patients. Anesth Analg