The predictors of catheter-related bloodstream infections in patients undergoing hemodialysis: A single center experience

Abstract

Background:

Here we aimed to investigate the predictors of catheter-related bloodstream infections (CRBSI) in patients with acute kidney injury or chronic kidney disease who required renal replacement therapy through a non-tunneled hemodialysis catheter.

Methods:

A total of 111 patients who received non-tunneled hemodialysis catheters were retrospectively evaluated. Patients were divided into two groups; those who developed CRBSI and those who did not. Patient’s demographic data, laboratory results at admission, information regarding catheter infections, and culture results were obtained from electronic medical records.

Results:

The mean age of the patients was 64 ± 16 years, and 51 of them were male. CRBSI occurred in 14 patients (12.6%). Admission serum albumin level (OR: 0.119, 95% CI: 0.019–0.756, p = 0.024), admission mean platelet volume (OR: 2.207, 95% CI: 1.188–4.100, p = 0.012) and catheter duration (OR: 1.580, 95% CI: 1.210–2.064, p = 0.001) were independent predictors for the CRBSI development. ROC curve analysis demonstrated that a catheter duration of 22 days was predictive for presence of CRBSI (78% sensitivity, 76% specificity, AUC: 0.825, 95% CI: 0.724–0.925, p < 0.001).

Conclusions:

Prolonged catheter duration, low serum albumin, and high mean platelet volume independently predict the development of CRBSI in patients undergoing hemodialysis for acute kidney injury or chronic kidney disease.

Keywords

Introduction

The proportion of end-stage renal disease (ESRD) patients requiring renal replacement therapy (RRT) has grown in the last two decades and is believed to continue to increase in the ensuing 10 years. Previous data have shown that relative risk for mortality and hospitalization is greater in patients receiving a catheter compared to patients having an arteriovenous fistula (AVF).¹

From the latest data from the Dialysis Outcome and Practice Patterns Study (DOPPS), it is understood that 15%–50% of patients in Europe and 60% of patients in the USA start their hemodialysis treatment with a catheter as primary access.² Nevertheless, non-tunneled catheters are the most commonly used catheters for RRT in AKI. In fact, the Kidney Disease: Improving Global Outcomes (KDIGO) guidelines suggest using non-tunneled catheters rather than tunneled catheters for vascular access in acute kidney injury (level of evidence, 2D) primarily for logistical reasons namely, ease of insertion and timeliness.³ There are multiple risk factors for CRBSI, including previous episodes of bacteremia, older age, diabetes, malnutrition, iron overload, longer duration of catheter use, and peripheral atherosclerosis.⁴ Catheter-related complications not only lead to increased healthcare costs but also result in serious morbidity and even death. Among ESRD patients, infection-related mortality is the second most frequent cause of death, following cardiovascular events.⁵ Thus, determining the factors associated with catheter-related bloodstream infections is crucial to improve outcomes in patients undergoing urge hemodialysis. Microorganisms that cause infection vary from region to region. Regardless of the organism involved, CRBSI can be complicated by metastatic infections such as endocarditis, osteomyelitis, septic arthritis, and epidural abscess. The risk of metastatic infection is highest with Staphylococcus aureus.⁶

The purpose of this study was to investigate the predictors of CRBSI in patients who required renal replacement therapy through a non-tunneled hemodialysis catheter due to renal failure.

Methods

This retrospective study was conducted among patients with a diagnosis of acute kidney injury who needed hemodialysis or who had chronic kidney disease and had just started dialysis and was treated with non-tunneled hemodialysis catheters between June 2016 and April 2018 in Istanbul Kanuni Sultan Suleyman Training and Research Hospital, Department of Internal Medicine, Türkiye. All non-tunneled catheters were inserted by experienced nephrologists under ultrasonography (USG) guidance except the femoral catheter application. The demographic data (age, gender and comorbid diseases including diabetes mellitus, hypertension, ESRD), laboratory results at admission (complete blood count, creatinine, sodium, potassium, calcium, phosphorus, total protein, serum albumin, and c-reactive protein), site of the catheter (right jugular vein, left jugular vein, or femoral vein), duration with non-tunneled catheter (days), information regarding catheter infections and culture results were obtained from electronic medical records. Patients with any diagnosis of active infectious disease, with a tunneled catheter inserted or with any signs of infection, chronic liver diseases (viral hepatitis, cirrhosis), and patients with incomplete data were excluded (Figure 1). The diagnosis was confirmed by two blood cultures (peripheral vein and hemodialysis catheter) taken before the administration of antimicrobial therapy and after the exclusion of alternative sources of infection among patients with fever that increased with shivering after temporary hemodialysis catheter insertion, with suspected clinical CRBSI. Blood cultures were taken by antiseptic techniques. Also, the incubation was observed within 48 h with the automated blood culture system. The positivity of skin flora pathogens in a single blood culture was considered as contamination (n = 4). If the blood culture taken from the catheter was considered as colonization (n = 18). These blood culture results (n = 22) were excluded from the study. Patients were divided into two groups according to the absence/presence of CRBSI. When signs of CRBSI were observed, blood samples for culture were acquired and empiric antimicrobial treatment (Vancomycin 15 mg/kg IV) was initiated immediately. Also, the non-tunneled catheter was removed and the tip was sent for culture. Written informed consent for the study was obtained from all patients. The study protocol was approved by Kanuni Sultan Suleyman Education and Training Hospital, Türkiye, Clinical Research Ethics Committee in January 07.2019 (protocol number: KAEK/2019.01.07), and the study was performed in accordance with the most recent version of the Helsinki Declaration.

Figure 1.

Patient selection and enrollment diagram.

Statistical analysis

Statistical analyses were carried out using SPSS for Windows, version 17 (SPSS, Chicago, IL, USA). The normality of the distribution of variables was studied with the Kolmogorov-Smirnov test. Continuous variables were presented as median, minimum and maximum, while categorical variables were given as frequency (n) and percentage (%). Comparisons among groups were performed using the Mann–Whitney U test. The chi-square test was used for the univariate analysis of the categorical variables. Logistic regression analysis was performed to identify the independent predictors of CRBSI. The capacity of catheter duration in predicting CRBSI was analyzed using receiver–operating characteristic (ROC) curve analysis. Two-sided p ⩽ 0.05 was interpreted to be statistically significant.

Results

A total of 111 patients (51 males, mean age 64 ± 16 years) electronic data were accessed in this study. CRBSI developed in 14 (12%) of these patients. Sixty (54%) patients were diabetic, 81 (73%) were hypertensive and 62 (54%) were suffering from ESRD. Catheter insertion sites were the right jugular vein in 68 patients (61%), left jugular vein in 35 patients (31%), and femoral vein in eight patients (7%). The pathogenic organisms isolated from blood cultures were: S. Epidermidis (n = 7, 50%), S. Aureus (n = 5, 35%), Enterococci (n = 1, 7%), and Pseudomonas aeroginosa (n = 1, 7%). Blood culture results of 22 of 34 patients excluded from the study were evaluated as contamination (n = 4, 3%) and colonization (n = 18, 12%).

Demographic features and laboratory findings were similar in patients with and without CRBSI (Table 1). Catheter insertion sites were also similar between the two groups. However, catheter duration was significantly higher in patients with CRBSI compared to those without CRBSI (25 ± 4 days vs 16 ± 6, p = 0.001).

Table 1.

Demographic features and laboratory results of the study population.

	Catheter-related bloodstream infection (−) n = 97	Catheter-related bloodstream infection (+) n = 14	p Value
Demographic parameters
Age, years	66 (19–94)	66.5 (48–81)	0.666
Sex (male/female)	46/51	5/9	0.411
Diabetes mellitus, n	50	10	0.163
Hypertension, n	70	11	0.614
ESRD, n	52	10	0.209
Laboratory parameters
Creatinine, mg/dL	5.5 (0.9–14.9)	6.45 (0.7–13)	0.206
Na, mmol/L	137 (119–146)	136.5 (130–143)	0.685
K, mEq/L	4.7 (3–7.4)	4.9 (3.5–8)	0.352
Ca, mg/dL	8.3 (5–15.5)	8.25 (6.7–9.4)	0.909
P, mg/dL	4.5 (1.6–11.2)	3.9 (2.5–8)	0.217
Total protein, g/dL	6.1 (4.1–9.1)	6.45 (4.7–7.7)	0.500
Albumin, g/dL	3.4 (1.6–4.9)	2.6 (1.7–4.4)	0.001
CRP, mg/L	40 (0–330)	25 (3.9–168)	0.428
WBC, 10⁹/L	9.9 (3.4–28.2)	8.7 (6.6–24)	0.611
MPV, fL	9.5 (0.6–14)	10 (5.9–11)	0.897
Catheter-related parameters
Catheter duration, days	24 (13–37)	67 (48–81)	0.001
Catheter site
RJV, n	59	9	0.404
LJV, n	31	4	0.799
FV, n	7	1	0.265

Na: sodium; K: potassium; Ca: calcium; P: phosphate; CRP: C-reactive protein; WBC: white blood cells; MPV: mean platelet volume; ESRD: end-stage renal disease; FV: femoral vein; LJV: left jugular vein; MPV: mean platelet volume; RJV: right jugular vein; WBC: white blood cell.

Logistic regression analysis (including the age, gender, serum levels of creatinine, albumin, total protein, C-reactive protein, white blood cell, mean platelet volume, catheter site, catheter duration and presence of diabetes mellitus, hypertension and ESRD) revealed that admission serum albumin level (OR: 0.119, 95% CI: 0.019–0.756, p = 0.024), admission mean platelet volume (MPV) (OR: 2.207, 95% CI: 1.188–4.100, p = 0.012) and catheter duration (OR: 1.580, 95% CI: 1.210–2.064, p = 0.001) were independent predictors for the development of CRBSI (Table 2). ROC curve analysis demonstrated that a catheter duration of 22 days was predictive for CRBSI development with a sensitivity of 78% and a specificity of 76% (AUC: 0.825, 95% CI: 0.724–0.925, p < 0.001, Figure 2). The odds ratio (OR) for catheter duration in excess of 22 days was 6.196 (95% CI: 1.863–20.212, p < 0.001).

Table 2.

The effects of several variables on CRBSI development.

	OR	95 % CI	p Value
Diabetes mellitus	0.553	0.045–6.745	0.641
Hypertension	3.805	0.3198–45.573	0.292
ESRD	0.523	0.048–5.638	0.593
Male gender	0.313	0.041–2.381	0.262
Age	1.050	0.976–1.130	0.192
Creatinine	1.157	0.793–1.688	0.450
Albumin	0.119	0.019–0.756	0.024
Total protein	0.950	0.304–2.967	0.929
CRP	0.989	0.975–1.004	0.145
WBC	1.073	0.903–1.276	0.424
MPV	2.207	1.188–4.100	0.012
Catheter site	0.902	0.570–1.428	0.659
Catheter duration	1.580	1.210–2.064	0.001

OR: odds ratio; CRP: C-reactive protein; ESRD: end-stage renal disease; CRBSI: catheter-related bloodstream infection; MPV: mean platelet volume; WBC: white blood cell count.

Figure 2.

ROC curve analysis demonstrating the sensitivity and specificity of catheter duration for predicting the development of CRBSI.

Discussion

Results of the present study revealed that non-tunneled catheter duration is significantly higher in patients with CRBSI compared to those without CRBSI. Our results also demonstrate that admission MPV, serum albumin level, and catheter duration are independent predictors for the development of CRBSI in patients with acute kidney injury or chronic kidney disease and undergoing hemodialysis using a non-tunneled catheter.

Several reports have shown that compared to AVFs, hemodialysis performed via central venous catheters is associated with 15–30-fold increased risk for the development of bloodstream infections.^7,8 A meta-analysis including 62 studies found that in the group of patients undergoing hemodialysis, the highest risk for mortality, fatal infections, and cardiovascular events were in those undergoing hemodialysis through hemodialysis catheters, while the risk was lowest in patients undergoing hemodialysis through AVFs.⁹ Several patients- and operator-related factors, including inadequate nutrition and uremia, insertion site, frequent manipulation of the catheter, and lack of aseptic techniques during the insertion, may lead to the development of CRBSI.¹⁰ In addition, performing dialysis through catheters also increases the risk for thrombosis, vascular stenosis, and inefficient dialysis.¹¹ Despite the drawbacks of dialyzing through catheters, a high proportion of patients continue to receive dialysis through central venous catheters, particularly at the initiation of dialysis.¹² The fact that catheters provide easy access, require minimal preparation, and do not usually cause complications are among the reasons for continued use of this approach. Besides, short-term catheter insertion may be the most reliable approach for prompt treatment in those with acute renal injury without access for hemodialysis and also CRF patients with loss of vascular access.¹³

Previous studies have mainly focused on the duration of catheterization as a major contributor to the development of CRBSI.^14–16 A prospective study conducted by Borges and colleagues found that non-tunneled catheter duration longer than 9 days was independently predictive for the development of CRBSI with an odds ratio of 2.634.¹⁷ They also reported that age, insertion site, and the type of renal failure (acute or chronic), were of no value in predicting CRBSI. Available data recommends the use of non-tunneled catheters no longer than 3 weeks as a consequence of the excessive increase in CRBSI frequency when catheters remain for a longer period.¹⁸ Another study recommends the use of tunneled hemodialysis catheters after 14 days.¹⁹ Our results are similar to the findings of a great deal of the previous studies on this matter, indicating that central venous catheters for hemodialysis should be used cautiously and only as a provisional measure. In this study, we found that a catheter duration exceeding 22 days was an independent risk factor for CRBSI development.

Admission serum albumin level, which may reflect the nutritional state of the patient, was also found to be a predictor for the development of CRBSI. A close relationship between inadequate nutrition and altered immunity has been shown in several studies.^20,21 Furthermore, previous scientific data suggests the presence of a bi-directional relationship between poor nutrition and increased susceptibility to infectious diseases; in turn, leading to immunological dysfunction.²² Serum albumin levels are also indicative of liver function, which may also be associated with the general health and nutritional status of the patient.²³ Hypoalbuminemia is more common in chronic liver disorders such as cirrhosis and viral hepatitis. In our study, patients with chronic liver disease were excluded from the study. With this perspective, our results suggest that CRBSI development is likely to be associated with protein malnutrition. Similarly, a study by Adeniyi and Tzamaloukas²⁴ also reported that lower preadmission serum albumin levels were associated with dialysis access site infection in patients with chronic renal failure. Another study, which was performed on children who underwent chronic peritoneal dialysis, concluded that low serum albumin levels could be a risk factor for catheter-related bloodstream infection.²⁵

In addition, we would like to report that there is a relationship between the level of MPV and the development of CRBSI as an independent risk factor. Under normal circumstances, there is an inverse relationship between platelet size and number. Therefore, the total platelet mass is closely regulated. Besides their primary hemostatic function, platelets also act in the pathogenesis of infectious diseases, and MPV is accepted to be associated with inflammatory activity.^26,27 In earlier studies, it was suggested that increased MPV might express either the development of invasive infection or the presence of infection not responding to standard antimicrobial treatment.²⁸ The role of MPV in inflammation and infection is a possible explanation for our results showing the predictive role of MPV in the development of CRBSI. Furthermore, MPV has also been suggested to be a link between inflammation and thrombosis, which may be a crucial underlying factor for complication development in patients with central venous catheters.²⁹ However, further studies are required to draw conclusions on this matter.

The pathogenesis of hemodialysis CRBSI is similar to that of other catheters. The most common source is the colonization of the intracutaneous and intravascular portions of the non-tunneled hemodialysis catheter by microorganisms from the patient’s skin and occasionally the hands of health care workers. In our study, the most common microorganisms isolated from blood cultures of the patients were: s. epidermidis (n = 7, 50%), s. aureus (n = 5, 35%). This finding broadly supports the results of other studies conducted in this field which have demonstrated that S. aureus and S. epidermidis are the most common infectious agents in patients with CRBSI.^30,31 Given the relatively high frequency of S. epidermidis infection in our group of patients, it is important to remind physicians of the importance of aseptic techniques,³² both during and after catheter insertion. However, as the mean duration with the catheter in patients who developed CRBSI was 25 ± 4 days (longer than the suggested 21 days of use), it is highly unlikely that aseptic techniques were disregarded in our group of patients.

There are some limitations to be mentioned in the present study. Firstly, we were looking for a marker that could have prognostic value. The number of our patient population decreased due to the exclusion criteria. Otherwise, our results would be misleading. Secondly, the clinical and laboratory findings of patients after the detection of CRBSI were not assessed in this study, including data concerning the efficiency of antimicrobial treatment and resistance to antibiotics. In addition, catheter-related local infection signs (especially catheter exit-site infection) were not adequately evaluated.

Although the prevalence of comorbid diseases especially diabetes and hypertension did not differ in our study groups, the lack of data about the underlying cause of the ESRD can be considered as a limitation of our study. Another limitation was the non-replacement of catheters with a tunneled catheter in patients with a duration time than 2 weeks because of their unstable clinical condition. Patients with ongoing hemodialysis need were referred for an AV fistula when getting clinically stable. The unavailability of patients’ weight data was our last limitation.

Nevertheless, despite these limitations, our results reflect the significance of catheter duration MPV levels and decreased serum albumin levels in CRBSI development.

Conclusion

Prolonged catheter duration, low serum albumin, and high MPV independently predict the development of CRBSI in patients undergoing hemodialysis with venous catheters for acute kidney injury or chronic kidney disease. Care should be taken not to prolong catheter duration longer than 22 days, as the duration with a catheter has been proved time and time again to be the main predictor for the development of CRBSI. This data should be supported by new studies including a greater number of patients and focused on this association. We also believe that role of MPV in CRBSI development should be studied in the future. Finally, it is important to remind physicians that strict commitment to aseptic techniques during and after the insertion of catheters is crucial for reducing the incidence of CRBSI.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Recep Demirci

References

Dhingra

Young

Hulbert-Shearon

, et al. Type of vascular access and mortality in US hemodialysis patients. Kidney Int 2001; 60(4): 1443–1451.

Pisoni

Young

Dykstra

, et al. Vascular access use in Europe and the United States: results from the DOPPS. Kidney Int 2002; 61: 305–316.

Kidney Disease: Improving Global Outcomes (KDIGO) and Acute Kidney İnjury Work Group. KDIGO clinical practice guideline for acute kidney injury. Kidney Inter Suppl 2012; 2: 1–138.

Garnacho-Montero

Aldabo-Pallás

Palomar-Martinez

, et al. Risk factors and prognosis of catheter-related bloodstream infection in critically ill patients: a multicenter study. Intensive Care Med 2008; 34: 2185–2193.

Patel

Kallen

Arduino

. Epidemiology, surveillance, and prevention of bloodstream infections in hemodialysis patients. Am J Kidney Dis 2010; 56(3): 566–577.

Ashby

Power

Singh

, et al. Bacteremia associated with tunneled hemodialysis catheters: outcome after attempted salvage. Clin J Am Soc Nephrol 2009; 4: 1601–1605.

Stevenson

Hannah

Lowder

, et al. Epidemiology of hemodialysis vascular access infections from longitudinal infection surveillance data: predicting the impact of NKF-DOQI clinical practice guidelines for vascular access. Am J Kidney Dis 2002; 39(3): 549–555.

Taylor

Gravel

Johnston

, et al. Incidence of bloodstream infection in multicenter inception cohorts of hemodialysis patients. Am J Infect Control 2004; 32(3): 155–160.

Ravani

Palmer

Oliver

, et al. Associations between hemodialysis access type and clinical outcomes: a systematic review. J Am Soc Nephrol 2013; 24(3): 465–473.

10.

Haga

Miyanari

Takahashi

, et al. Risk factors for catheter-related bloodstream infections in adult hospitalized patients—multicenter cohort study. Scand J Infect Dis 2013; 45(10): 773–779.

11.

Napalkov

Felici

Chu

, et al. Incidence of catheter-related complications in patients with central venous or hemodialysis catheters: a health care claims database analysis. BMC Cardiovasc Disord 2013; 13: 86.

12.

Pantelias

Grapsa

. Vascular access today. World J Nephrol 2012; 1(3): 69–78.

13.

Quarello

Forneris

Borca

, et al. Do central venous catheters have advantages over arteriovenous fistulas or grafts? J Nephrol 2006; 19(3): 265–279.

14.

Chen

Zhu

Chen

, et al. Clinical analysis of central venous catheter-related infections in patients in the emergency ICU. World J Emerg Med 2013; 4(3): 196–200.

15.

Patil

Ramteerthkar

, et al. Central venous catheter-related bloodstream infections in the intensive care unit. Indian J Crit Care Med 2011; 15(4): 213–223.

16.

Hentrich

Schalk

Schmidt-Hieber

, et al. Central venous catheter-related infections in hematology and oncology: 2012 updated guidelines on diagnosis, management and prevention by the Infectious Diseases Working Party of the German Society of Hematology and Medical Oncology. Ann Oncol 2014; 25(5): 936–947.

17.

Borges

Bedendo

. Rısk factors assocıated wıth temporary catheter-related ınfectıon ın patıents on dıalysıs treatment. Texto Contexto Enfermagem 2015; 24: 680–685.

18.

Bicudo

Batita

Furtado

, et al. Risk factors for catheter-related bloodstream infection: a prospective multicenter study in Brazilian intensive care units. Braz J Infect Dis 2011; 15(4): 328–331.

19.

Weijmer

Vervloet

ter Wee

. Compared to tunnelled cuffed haemodialysis catheters, temporary untunnelled catheters are associated with more complications already within 2 weeks of use. Nephrol Dial Transplant 2004; 19(3): 670–677.

20.

Ibrahim

Zambruni

Melby

, et al. Impact of childhood malnutrition on host defense and infection. Clin Microbiol Rev 2017; 30(4): 919–971.

21.

Keusch

. The history of nutrition: malnutrition, infection and immunity. J Nutr 2003; 133(1): 336–340.

22.

Yeh

. Nutrition, inflammation, and infectious disease. Mediators Inflamm 2012; 2012: 172480.

23.

Yang

, et al. Albumin binding function: the potential earliest indicator for liver function damage. Gastroenterol Res Pract 2016; 2016: 5120760.

24.

Adeniyi

Tzamaloukas

. Relation between access-related Infection and preinfection serum albumin concentration in patients on chronic hemodialysis. Hemodial Int 2003; 7(4): 304–310.

25.

Donmez

Durmaz

Ediz

, et al. Catheter-related complications in children on chronic peritoneal dialysis. Adv Perit Dial 2005; 21: 200–203.

26.

Altun

Paydas

Kaya

, et al. The relationship between inflammation, blood pressure, and mean platelet volume in chronic kidney disease. Saudi J Kidney Dis Transpl 2016; 27(4): 836–838.

27.

Balta

Aparci

Ozturk

, et al. Mean platelet volume as one part of platelet function determining inflammation. Ann Saudi Med 2016; 36(3): 234.

28.

Van der Lelie

Von dem Borne

. Increased mean platelet volume in septicaemia. J Clin Pathol 1983; 36(6): 693–696.

29.

Gasparyan

Ayvazyan

Mikhaillidis

, et al. Mean platelet volume: a link between thrombosis and inflammation? Curr Pharm Des 2011; 17(1): 47–58.

30.

Gahlot

Nigam

Kumar

, et al. Catheter-related bloodstream infections. Int J Crit Illn Inj Sci 2014; 4(2): 162–167.

31.

Zingg

Cartier-Fassler

Walder

. Central venous catheter-associated infections. Best Pract Res Clin Anaesthesiol 2008; 22(3): 407–421.

32.

Young

Commiskey

Wilson

. Translating evidence into practice to prevent central venous catheter-associated bloodstream infections: a systems-based intervention. Am J Infect Control 2006; 34(8): 503–506.