Non-Insertion-Related Complications of Central Venous Catheterization—Temporary Vascular Access for Hemodialysis

Abstract

The authors analyzed 309 central venous catheters (CVC) inserted in 147 hemodialysis patients before the maturation of the first or new arteriovenous fistula. One clinical manifestations of sepsis after CVC insertion was found. In all, 33.7% of the catheters were removed because of early minor complications: CVC occlusion, inadequate blood flow in CVC, shattered suture and malposition of CVC, fever, signs of infection at the site of CVC insertion, and bleeding at the site of CVC insertion. The most frequently isolated pathogenic bacteria at the tips of the catheters were coagulase-negative staphylococci highly sensitive to vancomycin and gentamicin.

Keywords:

• temporary vascular access
• central venous catheters
• renal disease
• hemodialysis
• complications

Introduction

Vascular access remains a major obstacle in the delivery of hemodialysis. After the first surgical arteriovenous (AV) fistula was created for the purpose of hemodialysis in 1965 by Brescia et al., it remained the gold standard for vascular access for hemodialysis patients.[1],[2] Central venous catheters (CVC) are mostly a temporary replacement for the AV fistula, but also serve as a final vascular access for hemodialysis.

Age and cardiovascular pathologies in hemodialysis patients offer increasing difficulties in determining vascular access. This implies the necessity of keeping central venous catheters in place and finding alternative puncturing sites.[3] International data show that 15% of vascular accesses for chronic hemodialysis are CVC.[4] For insertion of vascular access catheters, a modified Seldinger approach is used.[5],[6] The right internal jugular vein is the preferred site for insertion, but it can be placed in the subclavian or femoral vein.[7],[8] The advantages of CVC are easy insertion, relatively low cost, the possibility of immediate vascular access, and easy replacement. The disadvantages are frequent local and systemic infection, venous stenosis or thrombosis, local complication during insertion (arterial puncture, hemothorax, pneumothorax, cardiac arrhythmias, air embolism, loss of insertion guidewires into the vena cava, right atrium), short life of catheters, and easy dislodging of the catheter.[6],[9–11]

Infection is a common complication of CVC insertions. Catheter-related infections are thought to be due to several different mechanisms, such as infection of the exit site, followed by migration of the pathogen along the external catheter surface; contamination of the catheter hub, leading to intraluminal catheter colonization; and hematogenous seeding of the catheter.[12] Infections occur in 15–60% of catheter insertions and increase with the duration of use. They represent a significant cause of morbidity and mortality in hemodialysis patients.[6] Hoen et al.[13] estimated that the relative risk for bacteremia was increased more than seven-fold in patients with catheters than in those with AV fistulas. The incidence of bacteremia ranges from 0.5 to 13 per 1000 patient-days with hemodialysis catheters.[14] In two-thirds of the cases, Staphylococcus spp. are the causative pathogenic agents,[6],[15] though one can also find Streptococcus spp., gram-negative bacilli, and Candida spp. The presence of fever without another focus, a purulent exit site, and positive blood cultures with isolated organism should lead one to consider infection as a complication of CVC insertion. Recommended treatment is vancomycin 20–30 mg/kg/week and gentamicin 1–2 mg/kg after each dialysis session for at least 2–3 weeks.[6] Patients with persistent fever or elevated C-reactive protein (CRP) should be examined carefully for metastatic staphylococcal infections (endocarditis, osteomyelitis, intracerebral abscess).[6]

An important complication of CVC insertion is fibrin sheath formation in the catheter lumen, which can cause a reduction in blood flow. In the total reduction in blood flow, thrombosis of CVC could be found. The most common treatment is with urokinase or tissue-type plasminogen activator (t-PA).[6] Catheter thrombosis can be prevented by instilling heparin or sodium citrate into the lumens between dialysis sessions.[6]

The aim of the present analysis was to compare different sites of insertion of the CVC with their prevalence of complications. The sensitivity of isolated pathogenic bacteria to amoxicillin in combination with clavulanic acid, to cephalosporins, as well as to vancomycin and gentamicin was also studied. Insertion-related complications were not examined in this analysis.

Patients and Methods

Between April 1996 and March 2006, 309 temporary CVC were placed in 147 hemodialysis patients at the Maribor Teaching Hospital Department of Dialysis. CVC were inserted for time to maturation of first or new AV fistula and for chronic dialysis sessions. In all, 188 CVC (60.8%) were placed in men and 121 CVC (39.2%) in women. The age of patients ranged from 20 to 85 years, giving a mean of 58.4 ± 13.3 years. All catheters were inserted via percutaneous puncture in intensive care units by experienced physicians under strict aseptic techniques. After washing hands and forearms with antiseptic soap and drying with a sterile towel, maximal sterile barrier precautions (sterile gloves, long-sleeved sterile gown, mask, cap, and large sterile sheet drape) were used prior to catheter insertion. Antiseptic soap was used for cleaning the CVC insertion site. For insertion of CVC, a modified Seldinger method was used. After catheter insertion, the site was covered with sterile gauze. All told, 165 (53.4%) CVC were done on the internal jugular vein (147 on the right internal jugular vein, 18 on the left internal jugular vein), 109 (35.3%) CVC were done on the subclavian vein (77 on the right and 32 on the left subclavian vein) and 35 (11.3%) CVC were done on the femoral vein (25 on the right and 10 on the left femoral vein) (see Table 1). Heparin was instilled into the lumen between dialysis sessions. After CVC were removed, the tips of catheters were cut off and sent for culture analysis for the presence of bacteria. The growth of at least 15 colony-forming units on semiquantitative culture (roll plate) from catheter tip in the absence of accompanying clinical symptoms was defined as colonized catheter. In the case of isolation of the same organism from semiquantative culture of the catheter and from the blood of a patient with accompanying clinical symptoms of bloodstream infection and no other apparent source of infection, catheter-related bacteremia was confirmed. The manifestation of severe clinical symptoms (high body temperature, elevated CRP, tachycardia, hypotension) associated with organ dysfunction and hypoperfusion was defined as sepsis.[16],[17]

Table 1 Characteristics of the study patients (n = 309)

Parameter	n
Age (yr)	58.4 ± 13.3
Sex (M, F)	188, 121
Catheterization
Right internal jugular vein	147
Left internal jugular vein	18
Right subclavian vein	77
Left subclavian vein	32
Right femoral vein	25
Left femoral vein	10
Average duration of catheter placement (days)	32.5 ± 24.5

In statistical analysis, SPSS for Windows (version 12.0.1) and MedCalc for Windows (version 5.00.020) were used. Mean values, range, and SD were calculated. The significance of the relationship between variables was investigated using ANOVA test and post hoc test with multiple comparisons. The differences were considered statistically significant at p < 0.05.

Results

The catheters were left in place from 1 to 132 days with an average of 32.5 ± 24.5 days per catheter. A statistically significant relationship was found between different insertion sites of catheters and duration of inserted catheter before there removal because of minor complications, appearance, or deaths of patients ( p < 0.043). Further analysis showed that the duration of catheters inserted on the right internal jugular vein was statistically significantly longer compare to the duration of catheters inserted on the right femoral vein ( p < 0.034). The duration of catheters inserted on the right subclavian vein was statistically significantly longer compared to the duration of catheters inserted on right femoral vein ( p < 0.005) and the left subclavian vein ( p < 0.035). Between other insertion sites of catheters, no statistically significant differences was determined regarding the duration of inserted catheters. The reasons for removing the catheters varied. In all, 174 (56.3%) CVC were removed due to the successful puncture of maturated AV fistulae. About one-third of the catheters (33.7%) were removed because of early minor complications: CVC occlusion (n = 32; 10.4%), inadequate blood flow in CVC (n = 29; 9.4%), shattered suture of CVC and malposition of CVC (n = 12; 3.9%), fever of the patient without another focus (body temperature < 38°C), signs of infection at the site of CVC insertion (n = 27; 8.7%), bleeding at the site of CVC insertion (n = 4; 1.3%). Only one catheter (n = 1; 0.3%) was removed because of diagnostic criteria for sepsis. Six (1.9%) CVC were removed after the beginning of the peritoneal dialysis session. Twenty-four (7.8%) CVC were removed after the death of the patients (see Table 2). On 163 CVC (52.7% of CVC placed), one or more bacteria were isolated on the tip of the catheter after its removal. Bacteria isolated on the tips of the catheters were gram-positive (92.6%) and gram-negative (7.4%). The most frequent pathogenic bacteria isolated on the tips of the catheters were coagulase-negative staphylococci (70%), Staphylococcus aureus (12%), and others (Klebsiella pneumoniae, diphtheroids, Escherichia coli, Enterobacter agglomerans, Pseudomonas aeruginosa, Streptococcus viridans, Lactobacillus spp., Enterococcus spp., Citrobacter spp., Bacillus spp.) (18%) (see Table 3). Pathogenic bacteria isolated on 19 CVC were related to fever, but no diagnostic criteria for sepsis was confirmed. On 144 CVC, the colonization of the catheter was found with different bacteria without clinical signs of infection. The results of in vitro sensitivity data showed a low percentage (50.4%) of sensitivity of isolated gram-positive bacteria and a higher percentage (75%) of sensitivity of isolated gram-negative bacteria to amoxicillin in combination with clavulanic acid. In vitro sensitivity of isolated gram-positive bacteria to cephalosporins was low (50.4%). Contrarily, high in vitro sensitivity of isolated gram-negative bacteria to cephalosporins (90%) was found. Gram-positive bacteria resistant to amoxicillin in combination with clavulanic acid and to cephalosporins were highly sensitive to gentamicin (83.1%) and vancomycin (100%). Gram-negative bacteria resistant to amoxicillin in combination with clavulanic acid and also to cephalosporins were highly sensitive to gentamicin (88.9%) and vancomycin (100%). Analysis of in vitro sensitivity of bacteria (Staphylococcus aureus, coagulase-negative staphylococci, Klebsiella pneumoniae) isolated on the tips of CVC removed because of patients’ fever or signs of infection at the site of catheter insertion showed low sensitivity to amoxicillin in combination with clavulanic acid (66.7%) and to cephalosporins (66.7%). Bacteria resistant to these antibiotics were highly sensitive to gentamicin (88.9%) and vancomycin (100%). The insertion sites of CVC were compared with regard to the number of complications. When the number of early minor complications per inserted catheters was calculated separately for different catheter insertion sites, internal jugular vein catheterization was found to be associated with a smaller number of early minor complications (27.3%) compared to subclavian vein catheterization (33%), but the difference was not statistically significant ( p = 0.41).

Table 2 Reasons to remove CVC, and the numbers and frequencies of CVC

Reasons to remove CVC	n	%
Successful puncture of maturated AV fistulae	174	56.3
Beginning of the peritoneal dialysis	6	1.9
Death of the patients	24	7.8
Complication
CVC occlusion	32	10.4
Inadequate blood flow in CVC	29	9.4
Shattered suture of CVC and malposition of CVC	12	3.9
Fever of the patient, signs of infection at the site of CVC insertion	27	8.7
Bleeding at the site of CVC insertion	4	1.3
Sepsis of the patients	1	0.3

Table 3 Most common pathogens isolated from tips of catheters

Pathogen bacteria	n	%
Gram-positive bacteria
Coagulase-negative staphylococci	122	70.0
Staphylococcus aureus	23	11.8
Diphtheroids	12	6.7
Enterococcus spp.	4	2.3
Lactobacillus spp.	1	0.6
Bacillus spp.	1	0.6
Streptococcus viridans	1	0.6
Total	164	92.6
Gram-negative bacteria
Klebsiella pneumoniae	4	2.3
Pseudomonas aeruginosa	3	1.7
Escherichia coli	3	1.7
Enterobacter agglomerans	2	1.1
Citrobacter spp.	1	0.6
Total	13	7.4

Discussion

Catheter placement in the central vein is not an entirely safe procedure. On the contrary, it involves a number of complications connected with both the insertion of the catheter and leaving it in the vessel for a longer time.[18],[19]

After analyzing the data on the catheterization of CVC for the period between April 1996 and March 2006 at the dialysis center, only one CVC was removed due to clinical manifestation of sepsis. Staphylococcus aureus and coagulase-negative staphylococci were isolated on the tip of the CVC. A small part of the catheters were removed immediately after signs of infection at the site of CVC insertion and before clinical signs of sepsis occurred in the patients. On the tips of those catheters, Staphylococcus aureus, coagulase-negative staphylococci, and Klebsiella pneumoniae were isolated. The isolated bacteria were mostly part of the normal flora of the skin.[16] In comparing in vitro sensitivity of isolated bacteria to amoxicillin in combination with clavulanic acid and to cephalosporins, no important differences were found. In vitro sensitivity of isolated gram-negative bacteria to cephalosporins was comparable to in vitro sensitivity to gentamicin, but the number of gram-negative bacteria isolated on the tips of CVC was very small. Bacteria resistant to amoxicillin in combination with clavulanic acid and to cephalosporins were highly sensitive to vancomycin and gentamicin. These results confirmed the recommended treatment of catheter-related infection with vancomycin and gentamicin concerning antibiotic sensitivities.[6]

All 24 patients with catheters removed after death had died from the consequences of heart failure. There was no sign of infection or septic shock. Four bleedings were observed at the site of the CVC insertion in a patient with coagulation disorders. These data confirm previous reports stating that internal jugular vein catheterization was associated with the smallest number of complications when compared to other insertion sites, considering also the number of insertions of CVC into different sites.[3],[7],[19–24] Experience shows that the right internal jugular vein is the preferred site for insertion of a CVC.

In conclusion, central venous catheter placement is the only procedure for attaining immediate dialysis access for the treatment of hemodialysis patients and acute renal failure. Experience shows that central vein catheterization is relatively safe and effective for temporary vascular access for hemodialysis with many, but rarely major, complications. The internal jugular vein catheterization is the preferred site for the insertion of CVC. Concerning the smallest number of complications associated with internal jugular vein catheterization, the right internal jugular vein catheterization is the preferred site for insertion of CVC. Catheter-related infection with or without severe clinical symptoms requires parenteral antibiotics appropriate for the suspected organism, but definitive therapy should be based on the isolated organism. According to an analysis of in vitro sensitivity of isolated pathogenic bacteria and bacterial sensitivity to amoxicillin in combination with clavulanic acid or cephalosporins, it was concluded that those antibiotics aren’t the most appropriate empirical treatment of patients with catheter-related infection. Furthermore, these data proved high bacterial sensitivity to vancomycin and gentamicin for treatment of patients with catheter-related infection.