Bacterial Colonization in Hemodialysis Temporary Dual Lumen Catheters: A Prospective Study

Abstract

Aims. The use of hemodialysis temporary dual-lumen catheters is often complicated by infections, which may be a significant cause of death among patients with end stage renal disease (ESRD). The aim of this study was to assess the incidence of bacteremia and bacterial colonization related to non-tunneled, non-cuffed, dual-lumen temporary catheters in patients with ESRD submitted to hemodialysis. Methods. This study included 29 patients with ESRD. After catheter implantation, patients were monitored throughout the period of catheter permanence by means of blood samples collected weekly from a peripheral vein. Bacteria were isolated and identified according to CLSI recommendations. When catheters were removed for any reason, their tips were evaluated microbiologically. Results. A total of 194 blood samples from the 29 patients implanted with 55 catheters were analyzed. Of these, 15.5% (30 samples) demonstrated bacterial growth, principally Staphylococcus epidermidis (64.5%). Twenty patients (68.9%) presented at least one positive blood culture during follow-up. The median time for catheter colonization was 18.5 days (95% CI: 16.8–30.3). Of the 55 catheters implanted, 28 (50.9%) showed bacterial colonization, corresponding to 23.4 episodes/1000 catheter/days and 9.2 episodes of bacteremia /1000 catheter/days. Fifteen of 28 catheter tips analyzed showed bacterial growth (53.5%). In 14 of these (93.3%), there was agreement between the isolates from the catheter tip and blood cultures. Of 24 episodes of positive blood cultures from 20 different patients in 17 episodes (70.8%), the patients showed no clinical signs or symptoms of bacteremia. Conclusions. The high incidence of catheter colonization, the correlation between blood and catheter tip cultures, and the occurrence of frequent cases of asymptomatic bacteremia justify the proposal of routine peripheral blood collections to monitor patients undergoing hemodialysis with temporary dual-lumen catheters.

Keywords:

• bacterial colonization
• bacteremia
• catheter-related infection
• hemodialysis
• temporary catheters

INTRODUCTION

The use of intravascular catheters for vascular access is a frequent and growing practice worldwide, both inside and outside of hospitals. Each year, millions of intravascular catheters are used in healthcare institutions.[1–3]

The use of temporary or semi-permanent hemodialysis catheters remains an essential component of dialysis practice, either for the management of acute renal failure, as a temporary access for a patient with end stage renal disease (ESRD) who will need urgent dialysis therapy, or even in patients who lose their native arteriovenous (AV) fistula.[4] Unfortunately, the use of these catheters often involves mechanical or infectious complications that may result in patient morbidity, especially due to bacteremias that could provoke severe diseases.[5–7] The risk of catheter-associated bacteremia seems to be related to many different independent factors, such as patient profile, type of catheter, and certain characteristics of the healthcare facilities.[4],[8–14]

We conducted a prospective study on patients submitted to hemodialysis with temporary non-cuffed dual-lumen catheters in order to assess the incidence of bacterial colonization, risk factors for catheter-related bacteremia, and clinical outcomes of infectious episodes.

METHODS

Hemodialysis Unit

The study was performed in a hemodialysis unit that followed-up a universe of approximately 240 patients. Most of them had access through arterial-venous native fistulae.

Study Group

Twenty-nine patients (19 males) with ESRD submitted to hemodialysis at Serviço de Nefrologia de Ribeirão Preto (SENERP) with temporary two-lumen indwelling catheters (16 cm, radiopaque polyurethane with Blue Flex Tip? latex-free, Arrows International Inc., Reading, Pennsylvania, USA) were monitored prospectively from January 2004 to December 2004 by culture of blood samples collected weekly from a peripheral vein (10 ml) by an aseptic technique, 15 minutes after the beginning of the hemodialysis session throughout the period of catheter permanence. Only patients who had implanted catheters at SENERP were included.

Patients with positive blood cultures but who were asymptomatic had the catheters removed and did not receive antibiotics.[15] In these cases, the tip of the catheter (5 cm) was removed, collected in a sterile and dry tube, and sent for a semi-quantitative bacterial culture.[11] If the patients were symptomatic, they also had their catheters removed and received cephalexin (2 g p.o./day) until the microbiological displayed results. All patients were followed up to verify their clinical outcomes.

Control Group

Eleven patients (from the same Service) submitted to hemodialysis with native AV fistulas were also monitored prospectively during the same period of time via blood samples (10 ml) collected weekly from a peripheral vein over a one-month period by an antiseptic technique, 15 minutes after the beginning of the hemodialysis session. Patients with positive blood cultures but asymptomatic were not treated.

The study was approved by the Research Ethics Committee of the University Hospital, Faculty of Medicine of Ribeirão Preto, University of São Paulo, and all patients included in the study gave written informed consent to participate.

Microbiological Analysis of Blood and Catheters

Blood cultures were monitored with the continuous automated system BACT/ALERT (Organon Teknica, using a “standard” bottle, under aerobic conditions), and bacteria were isolated and identified according to the equipment specifications. Positive blood cultures were placed on a Muller Hinton agar plate containing 5% defibrinated sheep blood and used to identify bacterial gender and species according to recommendations of the Clinical Laboratory Standards Institute (CLSI).[16]

A segment of approximately 5 cm of the catheter tip was removed under sterile conditions and placed on the surface of a blood agar plate and incubated at 37°C for 24 h. The number of bacterial colonies was quantified, and the results are expressed as colony-forming units (CFU).[17] The positive catheter tip cultures were placed on a Muller Hinton agar plate containing 5% defibrinated sheep blood and used to identify bacterial gender and species according to CLSI recommendations.[16]

Event Definitions According to Blood and Catheter Tip Microbiological Results

Catheter contamination was considered to have occurred in the presence of CFU <15 in the culture of the catheter tip.[1] Bacterial colonization of the catheter was considered to have occurred when either of the following two criteria was met:

1. in the presence of CFU >15 in the catheter tip culture, even if the peripheral blood culture was negative; or

2. when peripheral blood cultures were positive, even if the catheter tip culture was negative or not done.

Bacteremia related to the catheter was considered to be present when >15 CFU of the same microorganism (characterized as belonging to the same genera and species and having the same profile of antibiotic sensitivity) were isolated from the blood culture and catheter tips.[1]

Risk Factors

The following risk factors for catheter related infection were obtained from the patients’ medical records: age, gender, duration of hemodialytic treatment, length of catheter permanence, site of catheter implantation, and presence of signs and symptoms of bacteremia (axillary temperature ≥38.0ºC, chills) on the day of blood culture collection.

Data Analysis

The rates of colonization and bacteremia related to the catheter were corrected as a function of total time of catheter permanence and are reported per 1000 catheter/days. For the analysis of risk factors, the quantitative variables are reported as mean + SD, and the data were analyzed by the Student t test. Qualitative variables were evaluated by univariate analysis using the chi-square test or Fisher exact test. The level of significance was set at 5% in all analyses.

RESULTS

Catheter Group

Over a period of 12 consecutive months, 55 dual-lumen temporary catheters were implanted in 29 patients (19 men and 10 women). A total of 194 blood samples for cultures were collected from a peripheral vein. Of these, 15.5% (30 samples) demonstrated bacterial growth.

Twenty-eight catheters (50.9%) were colonized after a median time of 18.5 days (95% CI: 16.8–30.3). Of these colonized catheters, 25 (89.2%) were associated with positive blood cultures.

Of the 55 catheters implanted, the most frequent reasons for catheter removal were positive blood cultures (20%), signs/symptoms of bacteremia (18.1%), transition to AV fistulae (16.3%), and obstruction (16.3%). Catheter tips were analyzed in 28 of the 55; of these, 15 (53.8%) showed bacterial growth >15 CFU. In 14 of these 15 tips (93.3%), there was agreement between the isolates from the tip and blood cultures, and in 11 cases (73.3%), the patients showed no signs or symptoms of bacteremia.

Of the 29 patients, 20 presented (68.9%) at least one colonized catheter, which represents a rate of 23.4 colonization episodes/1000 catheter/days. Of these 20 patients, 7 (35.0%) presented catheter-related bacteremia, with an incidence rate of 9.2 bacteremia episodes/1000 catheter/days. Staphylococcus epidermidis was the microorganism most frequently isolated from both blood cultures (64.5%) and catheter tips (42.8%), followed by S. aureus (6.45% and 10.7% from blood and catheter tips, respectively).

The risk factors for the 29 patients implanted with 55 catheters according to the results of peripheral blood cultures are presented in Table 1. The occurrence of positive blood cultures, indicating colonization/bacteremia related to the catheter, was higher in older patients and in catheters implanted in the subclavian veins. However, there was no association between positive blood cultures and patient gender, duration of hemodialytic treatment, and length of catheter permanence. In 17 of 24 (70.8%) episodes of positive blood cultures, originating from 20 different patients, the patients showed no clinical signs or symptoms of bacteremia.

Table 1 Risk factors related to 29 patients implanted with 55 catheters, according to the results of peripheral blood cultures

Factors		Positive blood cultures		p (95% CI)
		Yes (n = 25)	No (n = 30)
Sex (n = 55)	M	16	19	0.81^†
	F	9	11	1.03 (0.30–3.58)
Age (mean ± SD; n = 29)		58.6 ± 2.8	41.0 ± 3.9	0.0024
Months on dialysis (mean ± SD; n = 29)		19.9 ± 5.8	16.3 ± 7.1	0.72
Days of catheter permanence (mean ± SD; n = 55)		22.9 ± 24.9	24.1 ± 21.6	0.11
Site of implantation (n = 47)	Jugular	15	25	0.03^†
	Subclavian	6	1
Bacteremia (n = 53)	Yes	7	5	0.48^†
	No	17	24	1.41 (0.77–2.57)

Note. More than one catheter implanted per patient.

*Student t test, ^†chi-square test or Fisher exact test.

Most of the patients with symptomatic bacteremia received oral antibiotics and had a good clinical outcome, although two patients had to be admitted to the hospital for treatment with parenteral antibiotics without further occurrences. Another symptomatic patient (with intermittent episodes of fever) had positive blood cultures with S. epidermidis, and the cultures were persistently positive during an entire week. During this week, no treatment was provided, and the catheter was removed after only seven days of permanence. After catheter removal, he presented with back pain and persistent fever, and magnetic resonance imaging displayed discitis between lumbar vertebrae. He had a good outcome after being hospitalized and treated with oxacillin for six weeks.

Control Group

A total of 54 blood culture samples from 11 patients submitted to hemodialysis by native fistulas were analyzed. Of these, 3.7% (two samples) demonstrated bacterial growth, and the microorganisms isolated were S. aureus and Acinetobacter baumannii. The patients had no symptoms, were not treated, and had a good outcome.

DISCUSSION

Hemodialysis catheters are known to be a major risk factor for bacteremia, particularly when compared to native AV fistulas.[8] The reported incidence of bacteremia varies greatly, depending on the catheter type, length of catheter permanence, the cuff and tunnel used, and the central vein used.[18] In this study, the catheter colonization rate was 23.4/ 1000 catheter/days, with an incidence rate of 9.2 episodes/bacteremia/1000 catheter/days, and 68.9% of the patients presented at least one catheter with colonization evidence during the course of this study. Deshpande et al.[6] observed colonization rates of 5.07 per 1000 catheter/days, with infection rates of 4.01 per 1000 catheter/days. Saxena and Panhotra[19] reported rates of infection of 5.0 episodes/1000 catheter/days. Medina et al.,[13] studying hemodialysis patients in Uruguay, reported colonization rates of 15.7/1000 catheter/days and an infection rate of 7.6/1000 catheter/days.

In previous studies of non-cuffed catheters, as used in this report, the microbiota of the skin at site of implantation were found to be the main source for catheter colonization, and the microorganisms most often associated with catheter infections were S. epidermidis and S. aureus,[8],[12],[13] which was also observed here.

We observed that catheter colonization was more frequent in older patients and those with catheters implanted in the subclavian veins. In the literature, there are some differences in the risk factors for catheter colonization. Allon[20] did not detect an association of patient gender or age with colonization, although Richet et al.[18] reported a higher risk of colonization for subclavian-implanted catheters. In those two reports, the authors demonstrated an increased colonization risk according to the duration of catheterization. Medina et al.[13] also reported an increased risk for catheters implanted for more than 14 days. In the present study, there was no association of bacterial colonization with the length of catheter permanence or even the duration of dialysis treatment. A probable explanation for these differences is that our catheters were of the temporary non-cuffed type and removed quite early, unlike cuffed catheters, which tend to remain inserted for longer periods of time. Butterly and Schwab[21] reported a higher rate of infection risk for catheters implanted in the first 6 months of dialysis, which may reflect the patients’ unstable clinical pictures and ongoing adaptation to treatment.

Dittmer et al.[11] reported a follow-up of 31 tunneled cuffed-catheters in hemodialysis patients and observed that the mean time for their colonization was 27 days, and that 50% of the catheters were colonized five weeks after implantation. All of the catheters were colonized after 16 weeks. In the present study, 28 catheters (50.9%) were colonized after a period of 18.5 days (95% CI: 16.8–30.3). Of these, 24 (85.7%) were associated with positive blood cultures, and although we did not culture all of the catheter tips, there was a high correlation between the catheter tip cultures and the peripheral blood cultures (93.3%). It is important to emphasize that in 70.8% (17/24) of these cases, patients with positive blood cultures did not show any signs or symptoms of bacteremia on the day the blood was collected. Of 15 patients whose catheter tips showed a positive semi-quantitative culture (>15 CFU), 11 (73.3%) also did not present clinical evidence of bacteremia. Our findings agree with those reported by Dittmer et al.,[11] who observed catheter colonization without clinical evidence of infection in patients with positive blood cultures but without signs of septicemia or bacteremia.

Many prospective reports have demonstrated the occurrence of bacteremia-related morbidity secondary to catheter colonization, especially when catheters are not removed from the patients. Reported catheter infection complications include osteomyelitis, septic arthritis, endocarditis, discitis, septic pulmonary emboli, hepatic abscesses, pulmonary abscesses, myocardial abscesses, endophthalmitis, and septicemia.[5–7]

In the present study, the immediate removal of the catheter when colonization or bacteremia was suspected was followed by a prompt resolution of the infection, even when antibiotics were not prescribed in the asymptomatic patients. When patients showed signs and symptoms of bacteremia, such as fever or chills, the catheters were also removed, and the patients had a good outcome with antibiotics. The catheter was not removed promptly in only one case: the patient had a discitis and was treated, with a good outcome after a six-week course of oxacillin (S. epidermidis was the isolate in this case).

It is important to emphasize that the number of positive blood cultures in the control group (patients with native AV fistulas) was very low. This suggests that continuous venipunctures of the fistulas also facilitate bacterial access to the systemic circulation and could be the cause of some of the infections observed in hemodialysis patients with AV fistulas, although infectious complications are, by far, much more common in hemodialysis patients with artificial devices.[8]

Although it is a single-center study with a low number of patients, the high rates of colonization and bacteremia related to the use of temporary catheters observed in the present study, the high rates of asymptomatic bacteremia episodes, and the occurrence of infectious complications if the catheters are not removed justify the proposal of routine blood collections for culture in order to monitor bacterial colonization in hemodialysis patients with temporary dual-lumen catheters. In cases of positive blood cultures, without another apparent infectious site, the catheter should be removed, although there is no need to prescribe antibiotics if the patients remain asymptomatic.

ACKNOWLEDGMENTS

The authors wish to thank Grace R. Bittencourt and Gustavo R. N. Ferreira, trainees of the Laboratory of Microbiology of HCFMRP-USP, for help with the procedures of bacterial identification. The authors are also grateful to Fabiana D. Pereira, Taisa S. Dias and Rosemeire A. P. Louzada, nurses of the Service of Nephrology of Ribeirão Preto—SENERP, for help with the collection of blood samples and of catheter tip samples for culture.