Risk factors for recurrent catheter-related infections after catheter-related bloodstream infections

Article Outline

• Summary
• Introduction
• Materials and methods
  • Study population and hospital setting
  • Microbiologic criteria
  • Data collection and study design
• Results
• Discussion
• References
• Copyright

Summary 

Objectives

This study was performed to identify the risk factors for recurrent catheter-related infections (CRIs) following non-tunneled central venous catheter (CVC) reinsertion after catheter-related bloodstream infections (CRBSIs).

Methods

A retrospective cohort was constructed from a computer database for patients who underwent reinsertion of a non-tunneled CVC after a CRBSI during the period January 2004 to December 2007. Among these patients, recurrent CRI cases were selected through an electronic chart review, and the risk factors for recurrent CRI were investigated.

Results

Fifty-three patients who had had a reinserted non-tunneled CVC after a CRBSI were analyzed and 22 patients were considered as having recurrent CRIs (41.5%). Recurrent/persistent CRBSI after catheter reinsertion was observed in 16 patients, and six patients with systemic inflammatory response syndrome revealed positive results of an identical organism with the initial CRBSI in semi-quantitative reinsertion-catheter tip cultures. In multivariate analysis, fungal CRBSI compared with bacterial infection (adjusted hazard ratio (HR) 7.77, 95% confidence interval (CI) 1.71–35.36) and CRBSI occurrence during intensive care unit (ICU) care (adjusted HR 5.20, 95% CI 1.41–19.18) were revealed as independent risk factors for recurrent CRIs after catheter reinsertion on account of CRBSIs.

Conclusions

A substantial proportion of the patients with CRBSIs revealed recurrent CRIs after catheter reinsertion. Fungal CRBSIs when compared with bacterial infections and CRBSI occurrence during ICU care were independent risk factors for recurrent CRIs following catheter reinsertion after a CRBSI.

Keywords: Catheter-related bloodstream infection, Recurrence, Catheter-related infection, Risk factors

Back to Article Outline

Introduction 

Vascular catheter-related bloodstream infections (CRBSIs) have become a leading cause of healthcare-associated bloodstream infections, and are associated with substantial morbidity and mortality.¹, ², ³ In the case of a suspected CRBSI, removal of the catheter is generally recommended for non-tunneled central venous catheters (CVCs).⁴, ⁵ However, a substantial proportion of patients with CRBSIs require reinsertion of their CVC for various reasons, such as the administration of fluid therapy, medications, blood products, and parenteral nutrition. During catheter exchange for CRBSIs, guide-wire exchange appears to be associated with a greater risk of catheter colonization and catheter-related infections (CRI), however this finding has not been found to be statistically significant.⁶ In one study, the transfusion of blood products was found to be associated with recurrent CRBSIs.⁷ Both the removal of an ‘old’ catheter prior to a de novo insertion of a new line and antibiotic prophylaxis are also recommended during catheter exchange.⁸ However, data related to the risk factors associated with recurrent CRIs following catheter reinsertion are scarce. In this study, we organized a retrospective cohort of patients who had had a reinserted non-tunneled CVC after a CRBSI, and we investigated the risk factors contributing to their recurrent CRIs.

Back to Article Outline

Materials and methods 

Study population and hospital setting 

We collected clinical data between January 2004 and December 2007 at a single 2000-bed tertiary hospital. Using the computer database, we constructed a retrospective cohort of patients who had revealed CRBSI as defined below. Among the CRBSI patients, eligible cases were selected through an electronic chart review for those who had had a reinserted non-tunneled CVC. The clinical findings and microbiology data were analyzed for the recurrent CRIs in this population. Patients with multiple CVCs during the CRBSI were excluded from the recurrent CRI analysis.

For CVC insertion, a chlorhexidine/silver sulfadiazine surface-treated polyurethane catheter (two-lumen central venous catheterization set with blue Flextip^® ARROWg⁺ardBlue^® catheter, Teleflex, USA) was used for non-tunneled CVC insertion. Maximal sterile barrier precautions (gloves, gown, mask, cap, and large drape) and use of 2% chlorhexidine for skin antisepsis during CVC insertion were taken as standard of care by the institutional infection committee as suggested in the practice guidelines.⁹ In cases of recurrent or persistent CRBSIs, further occult infection foci were evaluated, including infective endocarditis, endophthalmitis, and infected thrombi.

Microbiologic criteria 

CRBSI was defined by positive blood cultures obtained from a peripheral vein and semi-quantitative catheter tip cultures yielding more than 15 colonies for the same organism with an identical antimicrobial susceptibility, without an obvious source of infection apart from the central line.⁴, ⁵ Patients with a single positive blood culture of coagulase-negative staphylococci (CNS) were excluded from the analysis.

Recurrent CRI was defined by the following criteria: (1) recurrent positive blood culture of identical organism after catheter exchange and three negative blood cultures; (2) persistent positive blood cultures after catheter exchange and positive semi-quantitative reinsertion-catheter tip cultures (yielding more than 15 colonies) of an identical organism; (3) positive semi-quantitative reinsertion-catheter tip cultures (yielding more than 15 colonies) of an identical organism with systemic inflammatory response syndrome (SIRS) upon catheter removal, without an obvious source of infection apart from the central line.

Data collection and study design 

Collected data included patient demographics, underlying comorbidities, types of initial catheter, administration of total parenteral nutrition/lipid formula/blood products via the reinsertion catheter, date of the CRBSI, date and time that the appropriate antibiotics were administered, date and time of catheter insertion/removal, isolated microorganisms, methods of catheter change (guide-wire vs. de novo site), and date/type of recurrence. Appropriate antibiotic administration was defined as parenteral administration of antibiotics for which the CRBSI-related microorganism revealed susceptibility, for more than 14 days. We defined three parameters related with catheter reinsertion to identify the risk factors contributing to recurrent CRI during a catheter exchange: (1) catheter removal after CRBSI: time between the initial CRBSI and the catheter removal; (2) catheter exchange interval: time between the initial catheter removal and the reinsertion of a CVC; and (3) pre-reinsertion treatment duration: time between the initiation of the appropriate antibiotics for the CRBSI and the CVC reinsertion.

The primary endpoint was recurrence of the CRI after the reinsertion of a CVC. Mean comparisons for the continuous variables were done using independent Student t-tests. Proportion comparisons for the categorical variables were done using Chi-square tests, and a Fisher's exact test was used when the data were sparse. Multivariate logistic regression analyses with a stepwise model selection procedure were performed to identify independent risk factors contributing to recurrent CRIs after CRBSIs and CVC reinsertion. Predefined variables which have been known to be associated with recurrent CRIs and variables revealing p<0.1 in our study were included in the final model. Kaplan–Meier survival analyses were performed to estimate catheter survival without recurrent CRI according to independent variables identified in the multivariate logistic regression model. Significance was set at p<0.05 using two-sided comparisons, and all of the analyses were performed using SAS version 9.1.3 (SAS Institute, Inc., Cary, NC, USA).

Back to Article Outline

Results 

During the study period, 151 CRBSI cases were identified. Candida albicans was the most common pathogen (n=44, 29.1%), followed by Staphylococcus aureus (n=21, 13.9%) (Table 1). Most cases were associated with non-tunneled CVCs (81.5%); venous ports were the next most common (11.3%). Among 123 non-tunneled CVC CRBSIs, the most frequent insertion site was the femoral vein (43.9%), followed by the subclavian (33.3%) and jugular (22.0%) veins. The median interval between the initial catheter insertion and the CRBSI was 17 days (interquartile range 10–25 days).

Table 1. Microbiologic data of the initial catheter-related bloodstream infections and reinsertion non-tunneled central venous catheters

	Initial CRBSI (n=151)	Eligible cases among non-tunneled CVC reinsertion (n=53)
		Recurrent CRI (n=22)	Non-recurrent CRI (n=31)	p-Value
Fungus (n=85; 56.3%)
Candida albicans	44 (29.1)	7 (31.8)	12 (38.7)	0.606
Candida tropicalis	17 (11.3)	5 (22.7)	2 (6.5)	0.113
Candida parapsilosis	14 (9.3)	4 (18.2)	1 (3.2)	0.147
Candida glabrata	4 (2.6)	0	0
Pichia anomala	4 (2.6)	2 (9.1)	0	0.168
Other	2 (1.3)	1 (4.5)	0	0.415
Bacteria (n=66; 43.7%)
Staphylococcus aureus	21 (13.9)	2 (9.1)	2 (6.5)	1.000
Coagulase-negative staphylococci	19 (12.6)	0	9 (29.0)	0.007
Enterococcus spp	14 (9.3)	0	2 (6.5)	0.505
Acinetobacter baumannii	3 (2.0)	0	0
Klebsiella pneumoniae	2 (1.3)	0	1 (3.2)	1.000
Other Gram-negative organism	7 (4.6)	1 (4.5)	2 (6.5)	1.000

Data are presented as the number of patients (%) unless otherwise indicated.

After the CRBSI, 88 patients underwent the reinsertion of a non-tunneled CVC after the initial catheter removal, excluding patients with multiple CVCs. Among the 88 patients, 35 cases were excluded from the analysis (Figure 1). Nine patients died without either recurrent CRBSI documentation or reinsertion catheter tip culture studies. Sixteen patients met SIRS criteria when the reinsertion catheters were removed, but no evidence of recurrent CRIs was observed. Nine patients revealed persistent BSI more than 48hours after the reinsertion CVC was removed. These patients were considered as having an occult infection focus apart from the CVC. One patient revealed persistent BSI without CRI evidence.

• 
  • View Large Image
  • Download to PowerPoint

• Figure 1. 

  Eligible case selection for recurrent catheter-related infection (CRI) evaluation. Thirty-five cases were excluded among 88 non-tunneled central venous catheter (CVC) reinsertion cases, and 53 eligible cases were selected (BSI, blood stream infection; CRBSI, catheter-related blood stream infection; RC, reinsertion catheter; SIRS, systemic inflammatory response syndrome).

Among the remaining 53 patients, 22 patients (41.5%) were considered as having recurrent CRIs after the initial CRBSI. Recurrent CRBSIs were the most common cause (n=12, 54.5%), followed by recurrent CRIs with SIRS (n=6, 27.3%) and persistent BSIs with CRIs (n=4, 18.2%). C. albicans was the most common pathogen among recurrent CRIs after a CRBSI, followed by Candida tropicalis (Table 1). Among the nine CNS CRBSI cases, no recurrent CRIs were observed after CVC reinsertion and this was statistically significant (p=0.007).

Comparing the recurrent CRI group after CRBSI with the non-recurrent CRI group (Table 2), the mean age was not different between the two groups, but the proportion of males was greater than females (77.3% vs. 45.2%, p=0.019). The mean APACHE II score was higher in the recurrent CRI group (11.8 vs. 9.0, p=0.017), and more patients were under intensive care unit (ICU) care when the initial CRBSI occurred (68.2% vs. 32.3%, p=0.010). Fungal CRBSIs resulted in recurrent CRIs after catheter reinsertion more frequently than bacterial infections (55.9% vs. 15.8%, p=0.005). The Charlson index, underlying comorbidities, and delayed administration of appropriate antibiotics (>48hours after CRBSI) were not associated with recurrent CRIs. Analysis of the mean catheter removal after CRBSI, catheter exchange interval, and pre-reinsertion treatment duration revealed no significant differences between the two groups. All patients received appropriate antibiotic treatment and there was no case of catheter exchange with guide-wire. The mean catheter survival time was significantly longer in the non-recurrent group (21.6 vs. 7.8 days, p<0.001).

Table 2. Comparison between the recurrent and the non-recurrent catheter-related infection groups

	Recurrent CRI (n=22; 41.5%)	Non-recurrent CRI (n=31; 58.5%)	p-Value
Mean age, years	65.9	60.0	0.167
Male sex	17 (77.3)	14 (45.2)	0.019
Charlson index, mean	3.14	3.06	0.913
APACHE II score, mean	11.8	9.0	0.017
Type of initial catheter
Non-tunneled CVC	18 (81.8)	27 (87.1)
Venous port	3 (13.6)	3 (9.7)
Hemodialysis catheter	1 (4.5)	1 (3.2)	0.505
Site of CVC reinsertion
Subclavian	15 (68.2)	12 (38.7)
Jugular	0	6 (19.4)
Femoral	7 (31.8)	13 (41.9)	0.131
Underlying disease
Malignancy	11 (50)	13 (41.9)	0.561
Diabetes mellitus	7 (31.8)	6 (19.4)	0.299
Renal disease	6 (27.3)	6 (19.4)	0.524
Heart failure	2 (9.1)	3 (9.7)	1.000
Liver cirrhosis	0	2 (6.5)	0.505
Neutropenia	3 (13.6)	1 (3.2)	0.295
Initial CRBSI during ICU management	15 (68.2)	10 (32.3)	0.010
Transfusion	11 (50)	19 (61.3)	0.414
TPN	20 (90.9)	28 (90.3)	1.000
Lipid formula	13 (59.1)	16 (51.6)	0.680
Major surgery	6 (27.3)	7 (22.6)	0.696
Microorganism
Bacteria	3 (13.6)	16 (51.6)
Fungus	19 (86.4)	15 (48.4)	0.005
Catheter removal after CRBSI, mean hours	57.8	55.3	0.833
Catheter exchange interval, mean hours	14.8	23.9	0.589
Pre-reinsertion treatment duration, mean hours	23.4	28.9	0.810
Early initiation of appropriate treatment (<48 hours)	9 (40.9)	14 (45.2)	0.758
Reinsertion catheter survival, mean days	7.8	21.6	<0.001

CRI, catheter-related infection; CVC, central venous catheter; CRBSI, catheter-related blood stream infection; ICU, intensive care unit; TPN, total parenteral nutrition.

Data are presented as the number of patients (%) unless otherwise indicated.

A logistic-regression model was used to identify any independent risk factors contributing to recurrent CRIs after a CRBSI. Variables incorporated into the model were age, sex, APACHE II score, Charlson index, neutropenia, malignancy, transfusion,⁷ ICU care during the initial CRBSI, reinsertion site of non-tunneled CVC, interval between initial CRBSI and catheter removal,¹⁰ and fungal CRBSIs when compared with bacterial infections. In the multivariate analysis (binary logistic, forward-conditional), fungal CRBSIs when compared with bacterial infections (adjusted hazard ratio (HR) 7.77, 95% confidence interval (CI) 1.71–35.36) and CRBSI occurrence during ICU care (adjusted HR 5.20, 95% CI 1.41–19.18) were revealed as independent risk factors for recurrent CRIs after CRBSIs and catheter reinsertion.

In the Kaplan–Meier survival analyses estimating catheter survival without recurrent CRI, fungal CRBSIs compared with bacterial infections and CRBSI occurrence during ICU management adversely affected the recurrent CRIs with statistical significance (log rank test, Figure 2).

• 
  • View Large Image
  • Download to PowerPoint

• Figure 2. 

  Kaplan–Meier survival analyses regarding the interval between catheter reinsertion and the occurrence of recurrent catheter-related infections (CRIs) determined by (a) fungal catheter-related blood stream infections (CRBSIs) compared with bacterial infections and (b) CRBSI occurrence during intensive care unit (ICU) management.

Back to Article Outline

Discussion 

It is well known that a prolonged catheter indwelling time is related to an increased risk of CRI. If the indwelling time is more than 7 days, the cumulative risk increases to 5–10%.⁸ However, the practice of routine replacement of CVCs in the absence of any clinical signs of infection, has revealed no decrease in the infection rates and increased morbidity due to mechanical complications.⁹, ¹⁰ In the cases of CRIs, some investigators have reported increases in CRBSIs associated with catheter exchange over the guide-wire.¹¹, ¹² However, other studies have not revealed significant increases in infectious complications with this procedure.¹³ Apart from the reinsertion method related to the guide-wire exchange, data are sparse regarding the risk factors related to recurrent CRIs after catheter exchange or reinsertion. Some authors recommend that the ‘old’ line should be removed prior to the insertion of the ‘new’ line, if possible.⁸ Recently, risk factors and outcomes associated with recurrent CRBSIs have been investigated.⁷ In this study, the authors reported 34% of recurrent CRBSIs after catheter reinsertion, and suggested transfusion of blood products as an independent risk factor for recurrent CRBSIs.

Our study included not only recurrent CRBSIs, but also recurrent CRI cases. For the specificity issues of recurrent CRIs, patients without SIRS or documented exit site infections were not regarded as having recurrent CRIs even if the semi-quantitative reinsertion-catheter tip culture yielded more than 15 colonies of the identical organism isolated from the initial CRBSI. For recurrent/persistent CRBSIs, nine cases revealed persistent bacteremia more than 48hours after the reinsertion catheter was removed. We excluded these cases from the analysis because we could not rule out occult infection foci unrelated to the reinsertion of the CVC.

Regarding risk factors for recurrent CRIs after CRBSIs, fungal CRBSIs compared with bacterial infections and CRBSI occurrence during ICU management were independently associated with recurrent CRIs in the multivariate analysis. It has been accepted as the standard of care to remove the CVC in patients with candidemia to improve clinical outcomes,¹⁰, ¹⁴, ¹⁵ although some authors are not in line with this strategy.¹⁶, ¹⁷ In this study, we found that the fungal CRBSIs were related with recurrent CRIs even if the infected catheters were removed and appropriate antibiotic administration was implemented. In a previous study,¹⁴ CVC retention more than 72hours after candidemia was related with a decreased response to antifungal therapy, as well as increased mortality and morbidity when candidemia was related with the CRI. However, among the 31 CRBSI cases of Candida spp in our study population, the interval between CRBSI and catheter removal was not different between the recurrent group and the non-recurrent group (70.0h vs. 64.3h, p=0.779). In contrast to fungal CRBSIs, no recurrent CRIs were observed among the nine CNS-associated CRBSI cases and this was statistically significant. We suppose that this is related to the lesser virulence of this organism.

We attempted to identify the modifiable risk factors that influence recurrent CRIs after a CRBSI, such as the time lapse before catheter removal after CRBSI, the catheter exchange interval, and the time lapse before catheter reinsertion after appropriate antibiotic administration. However, we failed to find a link between these variables and recurrent CRIs. The mean catheter exchange interval was slightly longer in the non-recurrent group, but this was not statistically significant. This is one of the issues emphasized for further investigation in the previous Infectious Diseases Society of America guidelines,⁵ but there is little correlated data. As previously reported,⁷ a substantial proportion of patients underwent the reinsertion of a new catheter immediately after or even before the initial catheter was removed during their management of the CRBSI in our study, and this might have resulted in a limited significance of the catheter exchange interval for recurrent CRIs.

Initial CRBSI occurrence during ICU management was revealed as an independent risk factor for recurrent CRIs in our study. We suppose that host immunity, which was not appropriately measured in our study model, or the resistance of the microorganism involved in the initial CRBSI among ICU patients, might influence the recurrence of CRIs. However, microbiologic verification regarding ICU preponderance for recurrence was not feasible because all of the isolates of the Candida spp were susceptible to fluconazole and the CNS isolates did not recur irrespective of their methicillin sensitivity (data not shown).

In our study, the proportion of fungal infections was relatively high among the initial CRBSIs (85/151, 56.3%). We suppose this is related to our inclusion criteria for CRBSIs, which required a positive semi-quantitative catheter tip culture. It has become the standard of care to remove the CVC immediately when candidemia occurs. So this policy might have resulted in a higher frequency of fungal CRBSIs in our study group because catheter-sparing management is more frequently tried in bacterial CRBSI cases, such as CNS.

The rate of recurrent CRIs was relatively high (41.5%) in our study. Among the 88 patients who had a reinserted CVC after a CRBSI, we excluded 16 patients from the analysis because they met the SIRS criteria when the reinsertion catheter was removed, but did not reveal a significant catheter tip colonization. Although a substantial proportion of these patients may not have had recurrent CRIs, these cases were excluded from the analysis because we could not rule out the possibility of recurrent CRIs below the sensitivity of the semi-quantitative tip culture methods such as exit site infection. This might have resulted in an increased proportion of recurrent CRI cases. If we limit the analysis to recurrent CRBSIs according to a previous report,⁷ the recurrence rate would be 24.5% (12/49, excluding four persistent bacteremia cases), lower than that of the report (34%).

In conclusion, a substantial proportion of patients with CRBSIs revealed a recurrent CRI after catheter reinsertion (41.5%). Fungal CRBSIs compared with bacterial infections and CRBSI occurrence during ICU care were independent risk factors for recurrent CRIs after CRBSIs and catheter reinsertion.

Our study has several limitations. First, the study number was relatively small and the proportion of fungal CRBSIs was relatively high. However, we confirmed that the microbiologic features, which determine management strategy for the catheter in CRBSIs, also influence the prognosis of a reinserted catheter even after the initial catheter is removed. Second, CRBSIs related to the exit site or tunnel infections were not included in our study. This might have influenced the epidemiologic features of CRBSIs. However, our microbiologic criteria using mainly semi-quantitative catheter tip cultures might be a relatively objective measurement, and this might contribute to reducing any biases in selecting the eligible cases.

Conflict of interest: No conflict of interest to declare.

Back to Article Outline

References 

1. Pittet D, Tarara D, Wenzel RP. Nosocomial bloodstream infection in critically ill patients: excess length of stay, extra costs, and attributable mortality. JAMA. 1994;271:1598–1601
   • View In Article
   • MEDLINE
2. Edgeworth JD, Treacher DF, Eykyn SJ. A 25-year study of nosocomial bacteremia in an adult intensive care unit. Crit Care Med. 1999;27:1421–1428
   • View In Article
   • MEDLINE
   • CrossRef
3. Wenzel RP, Edmond MB. The impact of hospital-acquired bloodstream infections. Emerg Infect Dis. 2001;7:174–177
   • View In Article
   • MEDLINE
4. Raad I, Hanna H, Maki D. Intravascular catheter-related infections: advances in diagnosis, prevention, and management. Lancet Infect Dis. 2007;7:645–657
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (201 KB)
   • CrossRef
5. Mermel LA, Farr BM, Sherertz RJ, Raad II, O’Grady N, Harris JS, et al. Guidelines for the management of intravascular catheter-related infections. Clin Infect Dis. 2001;32:1249–1272
   • View In Article
   • MEDLINE
   • CrossRef
6. Snyder RH, Archer FJ, Endy T, Allen TW, Condon B, Kaiser J, et al. Catheter infection: a comparison of two catheter maintenance techniques. Ann Surg. 1988;208:651–653
   • View In Article
   • MEDLINE
   • CrossRef
7. Erbay A, Ergonul O, Stoddard GJ, Samore MH. Recurrent catheter-related bloodstream infections: risk factors and outcome. Int J Infect Dis. 2006;10:396–400
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (137 KB)
   • CrossRef
8. Polderman KH, Girbes ARJ. Central venous catheter use: part 2: infectious complications. Intensive Care Med. 2002;28:18–28
   • View In Article
   • MEDLINE
   • CrossRef
9. O’Grady NP, Alexander M, Dellinger EP, Gerberding JL, Heard SO, Maki DG, et al. Guidelines for the prevention of intravascular catheter-related infections. Clin Infect Dis. 2002;35:1281–1307
   • View In Article
   • CrossRef
10. Nucci M, Colombo AL, Silveira F, Richtmann R, Salomao R, Branchini ML, et al. Risk factors for death in patients with candidemia. Infect Control Hosp Epidemiol. 1998;19:846–850
    • View In Article
    • MEDLINE
11. Cobb DK, High KP, Sawyer RG, Sable CA, Adams RB, Lindley DA, et al. A controlled trial of scheduled replacement of central venous and pulmonary-artery catheters. N Engl J Med. 1992;327:1062–1068
    • View In Article
    • MEDLINE
    • CrossRef
12. Olson ME, Lam K, Bodey GP, King EG, Costerton JW. Evaluation of strategies for central venous catheter replacement. Crit Care Med. 1992;20:797–804
    • View In Article
    • MEDLINE
    • CrossRef
13. Cook D, Randolph A, Kernerman P, Cupido C, King D, Soukup C, et al. Central venous catheter replacement strategies: a systematic review of the literature. Crit Care Med. 1997;25:1417–1424
    • View In Article
    • MEDLINE
    • CrossRef
14. Raad I, Hanna H, Boktour M, Girgawy E, Danawi H, Mardani M, et al. Management of Central venous catheters in patients with cancer and candidemia. Clin Infect Dis. 2004;38:1119–1127
    • View In Article
    • CrossRef
15. Nguyen MH, Peacock JE, Tanner DC, Morris AJ, Nguyen ML, Snydman DR, et al. Therapeutic approaches in patients with candidemia: evaluation in a multicenter, prospective, observational study. Arch Intern Med. 1995;155:2429–2435
    • View In Article
    • MEDLINE
16. Pasqualotto AC, Severo LC. The importance of central venous catheter removal in patients with candidaemia: time to rethink our practice?. Clin Microbiol Infect. 2008;14:2–4
    • View In Article
    • CrossRef
17. Rodriguez D, Park BJ, Almirante B, Cuenca-Estrella M, Planes AM, Mensa J, et al. Impact of early central venous catheter removal on outcome in patients with candidaemia. Clin Microbiol Infect. 2007;13:788–793
    • View In Article
    • CrossRef