- Email: [email protected]
Bacteremia related with arterial catheter in critically ill patients
Journal of Infection (2011) 63, 139e143
www.elsevierhealth.com/journals/jinf
Bacteremia related with arterial catheter in critically ill patients ´ L. Pe ´rez a, Javier Ariza b, Francisco Esteve a,*, Miquel Pujol b, Xose n b, Ricard Verdaguer c, Rafael Man ~ ez a Francesc Gudiol b, Enric Limo a
Intensive Care Department, IDIBELL, Hospital Universitario Bellvitge, Feixa LLarga S/N 08907 Hospitalet de Llobregat, Barcelona, Spain b Infectious Diseases Department, IDIBELL, Hospital Universitario Bellvitge, Feixa LLarga S/N 08907 Hospitalet de Llobregat, Barcelona, Spain c Microbiology Department, IDIBELL, Hospital Universitario Bellvitge, Feixa LLarga S/N 08907 Hospitalet de Llobregat, Barcelona, Spain Accepted 29 May 2011 Available online 7 June 2011
KEYWORDS Arterial catheter; Nosocomial infection; Intensive care unit; Bacteremia related catheter
Summary Objective: Catheter-related bloodstream infections (CR-BSI) are an increasing problem in the management of critically ill patients. Our objective was to analyze the incidence and epidemiology of CR-BSI in arterial catheters (AC) in a population of critically ill patients. Methods: We conducted a two-year, prospective, non-randomized study of patients admitted for > 24 h in a 24-bed medical-surgical major teaching ICU. We analyzed the arterial catheters and differentiated between femoral and radial locations. Difference testing between groups was performed using the two-tailed t-test and chi-square test as appropriate. Multivariate logistic regression analyses were conducted to identify independent predictors of CR-BSI occurrence and type of micro-organism responsible. Results: The study included 1456 patients requiring AC placement for 24 h. A total of 1543 AC were inserted for 14,437 catheter days. The incidence of AC-related bloodstream infections (ACR-BSI) was 3.53 episodes per 1000 catheter days. In the same period the incidence of central venous catheter (CVC)-related bloodstream infections was 4.98 episodes per 1000 catheter days. Logistic regression analysis showed that days of insertion (OR: 1.118 95% confidence interval (CI) 1.026e1.219) and length of ICU stay (OR: 1.052 95% CI: 1.025e1.079) were associated with a higher risk of ACR-BSI. Comparing 705 arterial catheters in femoral location with 838 in radial location, no significant differences in infection rates were found, although there was a trend toward a higher rate among femoral catheters (4.13 vs. 3.36 episodes per 1000 catheter days) (p Z 0.72). Among patients with ACR-BSI, Gram-negative bacteria were isolated in 16 episodes (61.5%) in the femoral location and seven (28%) in radial location (OR: 2.586; 95% CI: 1.051e6.363).
* Corresponding author. Tel.: þ34 4932607923. E-mail address: [email protected] (F. Esteve). 0163-4453/$36 ª 2011 The British Infection Association. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.jinf.2011.05.020
140
F. Esteve et al. Conclusions: We concluded that as has been reported for venous catheters ACR-BSI plays an important role in critically ill patients. Days of insertion and length of ICU stay increase the risk of ACR-BSI. The femoral site increases the risk for Gram-negative infection. ª 2011 The British Infection Association. Published by Elsevier Ltd. All rights reserved.
Introduction Catheter-related bloodstream infections (CR-BSI) are associated with increases in morbidity, mortality, length of hospital stay and medical costs.1e7 Arterial catheters (AC) are commonly used in patients who need continuous hemodynamic monitoring and multiple blood samples, such as intensive care unit (ICU) admissions. Several recent studies have reported high infection rates related to arterial catheter use in critically ill patients8e14 and others15e17 show that the infection potential of the AC is comparable with that of short term CVC. Therefore, in patients with suspected catheter bacteremia, both types of catheters, arterial and venous, should be analyzed. It is unclear whether the arterial catheter site is associated with an increased risk of infection.12,13,18 The objective of this study was to analyze the incidence and epidemiology of CR-BSI in arterial catheter use in a large cohort of critically ill patients.
Materials and methods Patients and setting We conducted a two-year, prospective, non-randomized study of patients admitted for > 24 h in a 24-bed medicalsurgical major teaching ICU located in a 900-bed university hospital (Bellvitge University Hospital, Barcelona, Spain). The nurse: patient ratio is 1:2. The data were collected from 1 April 2002 to 30 April 2004. We included all patients who required an AC placement for 24 h during their ICU stay. The local ethics committee for clinical research approved the study protocol. Written informed consent was not required as all procedures were routine, but we obtained oral informed consent from patients or their legal representative.
connections are protected with sterile gauze impregnated with chlorhexidine pomade. The dressing change of the insertion point is made every 72 h or if they are deteriorated. The infusion equipments are changed every 72 h. Nurses registered the location of the catheter, date of insertion, the appearance of the insertion site, the date of change of dressings and infusion equipment. Physicians working in the ICU were responsible for the insertion of the catheters. The attending physician decided on the insertion site, the indication for catheterization and the catheter type. The catheters were removed in the following circumstances: if they were no longer needed; if the patient exhibited signs of infection; if the skin surrounding the insertion site showed signs of local infection. Catheter tips were cultured by the Maki roll plate technique.20
Infection control and antibiotic use In our hospital, all the patients admitted in ICU are screened for meticillin-resistant Staphylococcus aureus (MRSA) by nasal swab. Contact isolation precautions are applied for the MRSA nasal carriers. If during the ICU admission detected a patient colonized or infected with multiresistant gram-negative bacteria also apply contact isolation precautions and screening for multidrug-resistant gram-negative colonization in the other patients admitted. The most widely used antibiotic in our ICU during the study period is piperacillin-tazobactam followed by amoxicillin-clavulanate (Table 1). 1194 patients received antibiotic treatment with a mean of 2.39 antibiotics per patient (total antibiotics was 2858). 66.38% of antibiotic treatments are empirical.
Table 1
Antimicrobial therapy in ICU.
Type of catheters The arterial catheters used were Leader Cath arterial (Vygon ) for radial locations and Seldicath (PTFE) (Plastimed Division Prodimed) for femoral locations.
Catheter insertion, care and removal In our center we have developed a protocol for the prevention of CR-BSI.19 The recommendations included in the protocol are for CVC and AC. All catheters were inserted aseptically using the Seldinger technique, after skin preparation with chlorhexidine digluconate 0.5% alcohol solution. The barrier measures followed for catheter insertion are surgical hand washing, use of sterile gown and gloves, use of surgical mask and cap; the puncture site is delimited by sterile drapes. The zone of catheter insertion and the catheter
Piperacillinetazobactam Amoxicillineclavulanate Vancomycin Imipenem Aztreonam Ciprofloxacin Ceftriaxone Tobramycin Levofloxacin Meropenem Others (Including antiviral and antifungal treatments) SD: Standard deviation.
N Z 2858 (%)
Mean days of treatment (SD)
411 300 249 232 160 132 123 109 100 95 947
7.3 5.6 3.5 7.8 2.9 7.3 6.6 4.3 6.1 7.6
(14.4%) (10.5%) (8.7%) (8.1%) (5.6%) (4.6%) (4.3%) (3.8%) (3.50%) (3.3%) (33.2%)
(5.7) (3.6) (3.3) (6.1) (1.5) (4.3) (5.7) (3.7) (4.6) (5.5)
Bacteremia related with arterial catheter
141
Definitions
Catheters
Colonized catheter was defined as a growth of 15 CFU (colony-forming units) in a semi quantitative culture from the tip.20 CR-BSI was defined as the presence of the same micro-organisms in both peripheral blood and the catheter tip cultures, when there were clinical signs of sepsis and no other source of bacteremia except the catheter. The definitions were based on the Centers for Disease Control and Prevention (CDC) criteria.21
One or more AC was used in 1013 of 1456 patients. A total of 1543 arterial catheters were inserted for 14,437 catheter days. We found 51 episodes of arterial catheter-related bloodstream infection (ACR-BSI). The ACR-BSI was 3.53 episodes per 1000 days of AC use. In 1110 of 1456 patients (76.2%) one or more CVCs were used. A total of 1662 catheters (570 in subclavian location, 405 in jugular location and 687 in femoral location) were inserted for 15,854 catheter days. In the same period we found 79 episodes of CVC-related bloodstream infection (CVCR-BSI) (4.98 episodes per 1000 days of CVC use). We studied 705 arterial catheters in femoral location and 838 in radial location. Analyzing the characteristics of patients according to catheter location, SAPS II, ICU stay and days of catheter insertion were significantly higher in the femoral location. We found no differences regarding ACR-BSI incidence in the different locations (4.13 episodes per 1000 catheter days for femoral location vs. 3.36 for radial location (p Z 0.72) (Table 2)). Logistic regression analysis showed that days of line insertion (OR: 1.118 95% CI: 1.026e1.219) and length of ICU stay (OR:1.052 95% CI: 1.025e1.079) were associated with a higher risk of CR-BSI (Tables 3 and 4). Gram-negative bacteria were isolated in 16 episodes (61.5%) in the femoral location and in seven (28%) in the radial location (OR: 2.586 95% CI: 1.051e6.363) (Table 5). Logistic regression analysis showed that the femoral arterial site was associated with a higher proportion of ACRBSI due to Gram-negative bacteria (OR:2.586, 95% CI: 1.051e6.363).
Statistical analysis Data were analyzed using SPSS 18.0 for Windows. Descriptive statistics were computed for all study variables which were presented as mean standard deviation and 95% confidence intervals (CI) or median and range. We analyzed the arterial catheters and differentiated between femoral and radial locations. Difference testing between groups was performed using the two-tailed t-test and chisquare test as appropriate. Variables that were found to be significant risk factors in univariate analyses were entered simultaneously in the multivariate model. Binary logistic regression for the presence of CR-BSI was conducted studying catheter days and ICU stay as main variables (enter method) and adjusting their effect in a second time by age, sex, SAPS II and catheter site (backstep method). Multinomial logistic regression was conducted for type of micro-organism responsible analyzing catheter site, catheter days, ICU stay, age and sex as covariables. The incidence density of CR-BSI (expressed as incidence per 1000 catheter days) and colonization rates (expressed as colonization per 100 catheters) in each catheter were calculated.
Discussion
Results Population During the study period 1456 patients were admitted, 980 (67.30%) of whom were males. The mean age was 57.27 (SD: 16.04) years and the mean SAPS II score was 34.25 (SD: 14.4) The basal disease was medical in 699 patients (48%), surgical in 493 patients (33.85%), polytrauma in 167 patients (11.4%) and transplantation in 97 patients (6.6%). The mean length of ICU stay was 24.7 days (SD: 16.45), and 344 (23.6%) patients died.
Table 2
Population and arterial catheter.
Age (SD) Sex (%male) SAPS II (SD) Mean to stay (days) (SD) Line days Mean line days (SD) Colonization rate (percentage) Catheter-related bacteremia episodes Episodes per 1000 catheter days a
Accordingly our results show, the arterial catheter is an important source of CR-BRSI in critically ill patients. In our study, the rates of ACR-BSI are comparable with the rates of infection related with CVCR-BSI. This observation may be controversial, in fact even in the clinical guidelines of the Centers for Disease Control and Prevention of 2002,21 the AC is considered to be rarely associated with bloodstream infection. However, these guidelines are contradictory and reporting that the rate of ACR-BSI is similar to the rate of CVC-BSI. A recent study16 concluded that the risk
Femoral (n Z 705)
Radial (n Z 838)
p
57.08 (16.6) 66.4 35.09 (14.6) 28.6 (25.1) 7009 9.96 (5.7) 3.97 26 4,13
57.64 (15.3) 69.9 32.9 (13.7) 22.6 (23.2) 7428 8.85 (5.8) 2.98 25 3.36
0.49 0.138 0.009 <0.001
Odds ratio (95% confidence interval) Z 1.10 (0.62e1.97).
<0.001 (t-Test) 0.33 NS (chi-square) 0.72 NS (chi-square)a
142
F. Esteve et al.
Table 3
Risk for ACR-BSI (univariate analysis).
a
Age SAPS IIa ICU stay (days)a Time of catheter insertion (days)a Sex Maleb Femaleb Location Femoralb Radialb
Non ACR-BSI (n Z 1492)
ACR-BSI (n Z 51)
p
57.3 (15.9) 33.8 (14.4) 25.05 (24.2) 9.3 (5.8)
57.7 (16.3) 35.9 (8.2) 35.8 (24.5) 10.07(4.9)
0.8 0.1 0.001 0.3
68.3% 31.7%
66.6% 33.3%
0.7
679 813
26 25
0.4
ACR-BSI: Arterial catheter-related bloodstream infection. a Mean (Standard deviation). b Percentage.
of colonization and catheter-related infection does not differ between AC and CVC. Other review17 concludes that if an intravascular device is suspected of causing sepsis in a critically ill patient, then the AC and CVC should be considered together, as AC sepsis rates are similar to those observed with short term CVCs. Consequently it may be concluded it must use the same preventive strategies in both devices. The literature contains several studies of AC related infection,8e18 and there is some controversy as to whether AC is a major source of infection. The role of AC in infection may be underestimated, as this catheter is usually used for short periods (1 or 2 days) in patients at a low risk of infection and who are not critically ill. As our results show, however, AC is a significant source of infection when arterial lines are fitted for long periods. Each AC manipulation represents a risk for colonization of the device. Multiple extractions, inadequate manipulation of connector systems and log-term device insertion can increase infection rates in AC. No firm policy of scheduled arterial catheter replacement was applied in our hospital and so we do not change routinely arterial catheters in spite of the recommendations of the CDC guidelines.21 However, a recent study suggests that systematic replacement of peripheral arterial catheters might be recommended for preventing colonization when prolonged cannulation is required.22 There is controversy over whether the arterial catheterization site is associated with a higher risk of infection.12,13,18 In our study, we did not find any differences between the infection rates of femoral and radial locations. This result should be treated with caution because the
Table 4
Risk for ACR-BSI (multivariate analysis).
Age SAPS II ICU stay Time of catheter insertion Sex Location
Odds ratio
IC 95% OR
p
1.007 1.008 1.052 1.118
0.988e1.027 0.988e1.029 1.025e1.079 1.026e1.219
0.4 0.4 <0.001 0.011
0.981 0.776
0.521e1.847 0.429e1.402
0.9 0.4
locations of the catheters were not randomized; however, we found no statistically significant differences between the infection rates of femoral and radial locations in AC in spite of the longer stay, days of catheter insertion and severity in patients with femoral catheter location. However, we note that there is a tendency for the rate of infection is higher in the femoral location (although not significant), this trend can be explained by a longer hospital stay of patients with an AC in the femoral location and more insertion time in this location. It is possible that a greater number of patients could be more conclusive data, which reaffirms the need for multicenter studies to the study of catheter-related bacteremia. In agreement with a recent report,18 we found a higher proportion of Gram-negative bacteria in the femoral location. Largest proportion of Gram-negative bacteria in the femoral location may be due to the proximity of the groin to the anal area. We believe that the increased risk of infection (>2.5) for Gram-negative bacteria after adjusting for days of catheter insertion, length of stay in ICU, age and sex of patients is significant. This observation may
Table 5
Bacteremia in AC.
Microorganism Gram-positive cocci Coagulase-negative staphylococci Enterococcus faecalis Gram-negative bacilli Enterobacter cloacae Klebsiella pneumoniae Proteus mirabilis Pseudomonas aeruginosa Pseudomonas spp Enterobacter aerogenes Serratia marcescens Klebsiella oxytoca Providencia rettgeri Citrobacter spp
Femoral (n Z 26)
Radial (n Z 25)
9
16
1
2
6 2 1 4 1 1 1 1 1 0
1 3 1 1 0 0 0 0 0 1
Bacteremia related with arterial catheter have therapeutic implications (for example, for Gramnegative antibiotic coverage) when infection is suspected in the femoral location as concluded Lorente et al.18 In conclusion, the arterial catheter plays an important role in CR-BSI. We found no differences in infection rates in different locations. Days of insertion and length of ICU stay increase the risk of ACR-BSI. Finally, our results suggest that the femoral site increases the risk for Gram-negative infection.
Source of funding The study was financed by the Spanish department of health and consumption (PI 020132, Fondo de Investigaciones Sanitarias, Instituto deSalud Carlos III).
References 1. Pittet D, Tarara D, Wenzel R. Nosocomial bloodstream infection in critically ill patients. Excess length of stay, extra costs, an attributable mortality. JAMA 1994;271:1598e601. 2. Smith RL, Meixler SM, Simberkoff MS. Excess mortality in critically ill patients with nosocomial bloodstream infections. Chest 1991;100:164e7. 3. Dimick JB, Pelz RK, Consunji R, et al. Increased resource use associated with catheter-related bloodstream infection in the surgical intensive care unit. Arch Surg 2001;136:229e34. 4. Arnow PM, Quimosing EM, Beach M. Consequences of intravascular catheter sepsis. Clin Infect Dis 1993;16:778e84. ~ a C, Pujol M, Pallare n del coste atribu5. Pen ´s R, et al. Estimacio n nosocomial: prolongacio n de la estancia hosible a la infeccio lculo de los costes alternativos. Med Clin (Barc) pitalaria y ca 1996;106:441e4. 6. Rello J, Ochagavia A, Sabanes E, et al. Evaluation of outcome of intravenous catheter-related infections in critically ill patients. Am J Respir Crit Care Med 2000;162:1027e30. 7. Soufir L, Timsit JF, Mahe C, et al. Attributable morbidity and mortality of catheter-related septicemia in critically ill patient: a matched, risk-adjusted, cohort study. Infect Control Hosp Epidemiol 1999;20:396e401. 8. Ducharme FM, Gauthier M, Lacroix J, et al. Incidence of infection related to arterial catheterization in children: a prospective study. Crit Care Med 1988;16:272e6.
143 9. El-Hamamsy I, Durrleman N, Stevens LM, et al. Incidence and outcome of radial artery infections following cardiac surgery. Ann Thorac Surg 2003;76:801e4. 10. Raad I, Umphrey J, Khan A, et al. The duration of placement as a predictor of peripheral and pulmonary arterial catheter infections. J Hosp Infect 1993;23:17e26. 11. Rijjndersa BJA. Catheter-related can be prevented if we take the arterial line seriously too! Crit Care Med 2005;33:1437e9. 12. Traore O, Liottier J, Souweine B. Prospective study of arterial and central venous catheter colonization and of arterial and central venous catheter-related bacteremia in intensive care units. Crit Care Med 2005;33:1276e80. 13. Martin C, Saux P, Papazian L, et al. Long-term arterial cannulation in ICU patients using the radial artery or dorsalis pedis artery. Chest 2001;119:901e6. 14. Lorente L, Santacreu R, Martı´n MM, et al. Arterial catheterrelated infection of 2,949 catheters. Crit Care 2006;10(3):R83. 15. Koh DB, Gowardman JR, Rickard CM, et al. Prospective study of peripheral arterial catheter infection and comparison with concurrently sited central venous catheters. Crit Care Med 2008;36(2):397e402. 16. Lucet JC, Bouadma L, Zahar JR, et al. Infectious risk associated with arterial catheters compared with central venous catheters. Crit Care Med 2010;38(4):1030e5. 17. Gowardman JR, Lipman J, Rickard CM. Assesment of peripheral arterial catheters as a source of sepsis in the critically ill: a narrative review. J Hosp Infect 2010;75(1):12e8. 18. Lorente L, Jime ´nez A, Jime ´nez JJ, et al. The catheter site influences in the micro-organism responsable of arterial catheter-related infection. Intensive Care Med 2006;32: 1919e20. 19. Esteve F, Pujol M, Ariza J, et al. Impact of a prevention program for catheter-related bloodstream infection in the intensive care unit of a tertiary hospital. Enferm Infecc Microbiol Clin 2009 Dec;27(10):561e5. 20. Maki DG, Weise CE. Sarafin HW a semiquantitative culture method for identifying intravenous-catheter-related infection. N Engl J Med 1977;296:1305e9. 21. O’Grady NP, Alexander M, Kellinger EP, et al. Guidelines for the prevention of intravascular catheter-related infections. Centers for disease control and prevention. MMWR Recomm Rep 2002 Aug 9;51(RR-10):1e29. 22. Khalifa R, Dahyot-fizelier C, Laksiri L, et al. Indwelling time and risk of colonization of peripheral arterial catheters in critically ill patients. Intensive Care Med 2008;34: 1820e6.
www.elsevierhealth.com/journals/jinf
Bacteremia related with arterial catheter in critically ill patients ´ L. Pe ´rez a, Javier Ariza b, Francisco Esteve a,*, Miquel Pujol b, Xose n b, Ricard Verdaguer c, Rafael Man ~ ez a Francesc Gudiol b, Enric Limo a
Intensive Care Department, IDIBELL, Hospital Universitario Bellvitge, Feixa LLarga S/N 08907 Hospitalet de Llobregat, Barcelona, Spain b Infectious Diseases Department, IDIBELL, Hospital Universitario Bellvitge, Feixa LLarga S/N 08907 Hospitalet de Llobregat, Barcelona, Spain c Microbiology Department, IDIBELL, Hospital Universitario Bellvitge, Feixa LLarga S/N 08907 Hospitalet de Llobregat, Barcelona, Spain Accepted 29 May 2011 Available online 7 June 2011
KEYWORDS Arterial catheter; Nosocomial infection; Intensive care unit; Bacteremia related catheter
Summary Objective: Catheter-related bloodstream infections (CR-BSI) are an increasing problem in the management of critically ill patients. Our objective was to analyze the incidence and epidemiology of CR-BSI in arterial catheters (AC) in a population of critically ill patients. Methods: We conducted a two-year, prospective, non-randomized study of patients admitted for > 24 h in a 24-bed medical-surgical major teaching ICU. We analyzed the arterial catheters and differentiated between femoral and radial locations. Difference testing between groups was performed using the two-tailed t-test and chi-square test as appropriate. Multivariate logistic regression analyses were conducted to identify independent predictors of CR-BSI occurrence and type of micro-organism responsible. Results: The study included 1456 patients requiring AC placement for 24 h. A total of 1543 AC were inserted for 14,437 catheter days. The incidence of AC-related bloodstream infections (ACR-BSI) was 3.53 episodes per 1000 catheter days. In the same period the incidence of central venous catheter (CVC)-related bloodstream infections was 4.98 episodes per 1000 catheter days. Logistic regression analysis showed that days of insertion (OR: 1.118 95% confidence interval (CI) 1.026e1.219) and length of ICU stay (OR: 1.052 95% CI: 1.025e1.079) were associated with a higher risk of ACR-BSI. Comparing 705 arterial catheters in femoral location with 838 in radial location, no significant differences in infection rates were found, although there was a trend toward a higher rate among femoral catheters (4.13 vs. 3.36 episodes per 1000 catheter days) (p Z 0.72). Among patients with ACR-BSI, Gram-negative bacteria were isolated in 16 episodes (61.5%) in the femoral location and seven (28%) in radial location (OR: 2.586; 95% CI: 1.051e6.363).
* Corresponding author. Tel.: þ34 4932607923. E-mail address: [email protected] (F. Esteve). 0163-4453/$36 ª 2011 The British Infection Association. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.jinf.2011.05.020
140
F. Esteve et al. Conclusions: We concluded that as has been reported for venous catheters ACR-BSI plays an important role in critically ill patients. Days of insertion and length of ICU stay increase the risk of ACR-BSI. The femoral site increases the risk for Gram-negative infection. ª 2011 The British Infection Association. Published by Elsevier Ltd. All rights reserved.
Introduction Catheter-related bloodstream infections (CR-BSI) are associated with increases in morbidity, mortality, length of hospital stay and medical costs.1e7 Arterial catheters (AC) are commonly used in patients who need continuous hemodynamic monitoring and multiple blood samples, such as intensive care unit (ICU) admissions. Several recent studies have reported high infection rates related to arterial catheter use in critically ill patients8e14 and others15e17 show that the infection potential of the AC is comparable with that of short term CVC. Therefore, in patients with suspected catheter bacteremia, both types of catheters, arterial and venous, should be analyzed. It is unclear whether the arterial catheter site is associated with an increased risk of infection.12,13,18 The objective of this study was to analyze the incidence and epidemiology of CR-BSI in arterial catheter use in a large cohort of critically ill patients.
Materials and methods Patients and setting We conducted a two-year, prospective, non-randomized study of patients admitted for > 24 h in a 24-bed medicalsurgical major teaching ICU located in a 900-bed university hospital (Bellvitge University Hospital, Barcelona, Spain). The nurse: patient ratio is 1:2. The data were collected from 1 April 2002 to 30 April 2004. We included all patients who required an AC placement for 24 h during their ICU stay. The local ethics committee for clinical research approved the study protocol. Written informed consent was not required as all procedures were routine, but we obtained oral informed consent from patients or their legal representative.
connections are protected with sterile gauze impregnated with chlorhexidine pomade. The dressing change of the insertion point is made every 72 h or if they are deteriorated. The infusion equipments are changed every 72 h. Nurses registered the location of the catheter, date of insertion, the appearance of the insertion site, the date of change of dressings and infusion equipment. Physicians working in the ICU were responsible for the insertion of the catheters. The attending physician decided on the insertion site, the indication for catheterization and the catheter type. The catheters were removed in the following circumstances: if they were no longer needed; if the patient exhibited signs of infection; if the skin surrounding the insertion site showed signs of local infection. Catheter tips were cultured by the Maki roll plate technique.20
Infection control and antibiotic use In our hospital, all the patients admitted in ICU are screened for meticillin-resistant Staphylococcus aureus (MRSA) by nasal swab. Contact isolation precautions are applied for the MRSA nasal carriers. If during the ICU admission detected a patient colonized or infected with multiresistant gram-negative bacteria also apply contact isolation precautions and screening for multidrug-resistant gram-negative colonization in the other patients admitted. The most widely used antibiotic in our ICU during the study period is piperacillin-tazobactam followed by amoxicillin-clavulanate (Table 1). 1194 patients received antibiotic treatment with a mean of 2.39 antibiotics per patient (total antibiotics was 2858). 66.38% of antibiotic treatments are empirical.
Table 1
Antimicrobial therapy in ICU.
Type of catheters The arterial catheters used were Leader Cath arterial (Vygon ) for radial locations and Seldicath (PTFE) (Plastimed Division Prodimed) for femoral locations.
Catheter insertion, care and removal In our center we have developed a protocol for the prevention of CR-BSI.19 The recommendations included in the protocol are for CVC and AC. All catheters were inserted aseptically using the Seldinger technique, after skin preparation with chlorhexidine digluconate 0.5% alcohol solution. The barrier measures followed for catheter insertion are surgical hand washing, use of sterile gown and gloves, use of surgical mask and cap; the puncture site is delimited by sterile drapes. The zone of catheter insertion and the catheter
Piperacillinetazobactam Amoxicillineclavulanate Vancomycin Imipenem Aztreonam Ciprofloxacin Ceftriaxone Tobramycin Levofloxacin Meropenem Others (Including antiviral and antifungal treatments) SD: Standard deviation.
N Z 2858 (%)
Mean days of treatment (SD)
411 300 249 232 160 132 123 109 100 95 947
7.3 5.6 3.5 7.8 2.9 7.3 6.6 4.3 6.1 7.6
(14.4%) (10.5%) (8.7%) (8.1%) (5.6%) (4.6%) (4.3%) (3.8%) (3.50%) (3.3%) (33.2%)
(5.7) (3.6) (3.3) (6.1) (1.5) (4.3) (5.7) (3.7) (4.6) (5.5)
Bacteremia related with arterial catheter
141
Definitions
Catheters
Colonized catheter was defined as a growth of 15 CFU (colony-forming units) in a semi quantitative culture from the tip.20 CR-BSI was defined as the presence of the same micro-organisms in both peripheral blood and the catheter tip cultures, when there were clinical signs of sepsis and no other source of bacteremia except the catheter. The definitions were based on the Centers for Disease Control and Prevention (CDC) criteria.21
One or more AC was used in 1013 of 1456 patients. A total of 1543 arterial catheters were inserted for 14,437 catheter days. We found 51 episodes of arterial catheter-related bloodstream infection (ACR-BSI). The ACR-BSI was 3.53 episodes per 1000 days of AC use. In 1110 of 1456 patients (76.2%) one or more CVCs were used. A total of 1662 catheters (570 in subclavian location, 405 in jugular location and 687 in femoral location) were inserted for 15,854 catheter days. In the same period we found 79 episodes of CVC-related bloodstream infection (CVCR-BSI) (4.98 episodes per 1000 days of CVC use). We studied 705 arterial catheters in femoral location and 838 in radial location. Analyzing the characteristics of patients according to catheter location, SAPS II, ICU stay and days of catheter insertion were significantly higher in the femoral location. We found no differences regarding ACR-BSI incidence in the different locations (4.13 episodes per 1000 catheter days for femoral location vs. 3.36 for radial location (p Z 0.72) (Table 2)). Logistic regression analysis showed that days of line insertion (OR: 1.118 95% CI: 1.026e1.219) and length of ICU stay (OR:1.052 95% CI: 1.025e1.079) were associated with a higher risk of CR-BSI (Tables 3 and 4). Gram-negative bacteria were isolated in 16 episodes (61.5%) in the femoral location and in seven (28%) in the radial location (OR: 2.586 95% CI: 1.051e6.363) (Table 5). Logistic regression analysis showed that the femoral arterial site was associated with a higher proportion of ACRBSI due to Gram-negative bacteria (OR:2.586, 95% CI: 1.051e6.363).
Statistical analysis Data were analyzed using SPSS 18.0 for Windows. Descriptive statistics were computed for all study variables which were presented as mean standard deviation and 95% confidence intervals (CI) or median and range. We analyzed the arterial catheters and differentiated between femoral and radial locations. Difference testing between groups was performed using the two-tailed t-test and chisquare test as appropriate. Variables that were found to be significant risk factors in univariate analyses were entered simultaneously in the multivariate model. Binary logistic regression for the presence of CR-BSI was conducted studying catheter days and ICU stay as main variables (enter method) and adjusting their effect in a second time by age, sex, SAPS II and catheter site (backstep method). Multinomial logistic regression was conducted for type of micro-organism responsible analyzing catheter site, catheter days, ICU stay, age and sex as covariables. The incidence density of CR-BSI (expressed as incidence per 1000 catheter days) and colonization rates (expressed as colonization per 100 catheters) in each catheter were calculated.
Discussion
Results Population During the study period 1456 patients were admitted, 980 (67.30%) of whom were males. The mean age was 57.27 (SD: 16.04) years and the mean SAPS II score was 34.25 (SD: 14.4) The basal disease was medical in 699 patients (48%), surgical in 493 patients (33.85%), polytrauma in 167 patients (11.4%) and transplantation in 97 patients (6.6%). The mean length of ICU stay was 24.7 days (SD: 16.45), and 344 (23.6%) patients died.
Table 2
Population and arterial catheter.
Age (SD) Sex (%male) SAPS II (SD) Mean to stay (days) (SD) Line days Mean line days (SD) Colonization rate (percentage) Catheter-related bacteremia episodes Episodes per 1000 catheter days a
Accordingly our results show, the arterial catheter is an important source of CR-BRSI in critically ill patients. In our study, the rates of ACR-BSI are comparable with the rates of infection related with CVCR-BSI. This observation may be controversial, in fact even in the clinical guidelines of the Centers for Disease Control and Prevention of 2002,21 the AC is considered to be rarely associated with bloodstream infection. However, these guidelines are contradictory and reporting that the rate of ACR-BSI is similar to the rate of CVC-BSI. A recent study16 concluded that the risk
Femoral (n Z 705)
Radial (n Z 838)
p
57.08 (16.6) 66.4 35.09 (14.6) 28.6 (25.1) 7009 9.96 (5.7) 3.97 26 4,13
57.64 (15.3) 69.9 32.9 (13.7) 22.6 (23.2) 7428 8.85 (5.8) 2.98 25 3.36
0.49 0.138 0.009 <0.001
Odds ratio (95% confidence interval) Z 1.10 (0.62e1.97).
<0.001 (t-Test) 0.33 NS (chi-square) 0.72 NS (chi-square)a
142
F. Esteve et al.
Table 3
Risk for ACR-BSI (univariate analysis).
a
Age SAPS IIa ICU stay (days)a Time of catheter insertion (days)a Sex Maleb Femaleb Location Femoralb Radialb
Non ACR-BSI (n Z 1492)
ACR-BSI (n Z 51)
p
57.3 (15.9) 33.8 (14.4) 25.05 (24.2) 9.3 (5.8)
57.7 (16.3) 35.9 (8.2) 35.8 (24.5) 10.07(4.9)
0.8 0.1 0.001 0.3
68.3% 31.7%
66.6% 33.3%
0.7
679 813
26 25
0.4
ACR-BSI: Arterial catheter-related bloodstream infection. a Mean (Standard deviation). b Percentage.
of colonization and catheter-related infection does not differ between AC and CVC. Other review17 concludes that if an intravascular device is suspected of causing sepsis in a critically ill patient, then the AC and CVC should be considered together, as AC sepsis rates are similar to those observed with short term CVCs. Consequently it may be concluded it must use the same preventive strategies in both devices. The literature contains several studies of AC related infection,8e18 and there is some controversy as to whether AC is a major source of infection. The role of AC in infection may be underestimated, as this catheter is usually used for short periods (1 or 2 days) in patients at a low risk of infection and who are not critically ill. As our results show, however, AC is a significant source of infection when arterial lines are fitted for long periods. Each AC manipulation represents a risk for colonization of the device. Multiple extractions, inadequate manipulation of connector systems and log-term device insertion can increase infection rates in AC. No firm policy of scheduled arterial catheter replacement was applied in our hospital and so we do not change routinely arterial catheters in spite of the recommendations of the CDC guidelines.21 However, a recent study suggests that systematic replacement of peripheral arterial catheters might be recommended for preventing colonization when prolonged cannulation is required.22 There is controversy over whether the arterial catheterization site is associated with a higher risk of infection.12,13,18 In our study, we did not find any differences between the infection rates of femoral and radial locations. This result should be treated with caution because the
Table 4
Risk for ACR-BSI (multivariate analysis).
Age SAPS II ICU stay Time of catheter insertion Sex Location
Odds ratio
IC 95% OR
p
1.007 1.008 1.052 1.118
0.988e1.027 0.988e1.029 1.025e1.079 1.026e1.219
0.4 0.4 <0.001 0.011
0.981 0.776
0.521e1.847 0.429e1.402
0.9 0.4
locations of the catheters were not randomized; however, we found no statistically significant differences between the infection rates of femoral and radial locations in AC in spite of the longer stay, days of catheter insertion and severity in patients with femoral catheter location. However, we note that there is a tendency for the rate of infection is higher in the femoral location (although not significant), this trend can be explained by a longer hospital stay of patients with an AC in the femoral location and more insertion time in this location. It is possible that a greater number of patients could be more conclusive data, which reaffirms the need for multicenter studies to the study of catheter-related bacteremia. In agreement with a recent report,18 we found a higher proportion of Gram-negative bacteria in the femoral location. Largest proportion of Gram-negative bacteria in the femoral location may be due to the proximity of the groin to the anal area. We believe that the increased risk of infection (>2.5) for Gram-negative bacteria after adjusting for days of catheter insertion, length of stay in ICU, age and sex of patients is significant. This observation may
Table 5
Bacteremia in AC.
Microorganism Gram-positive cocci Coagulase-negative staphylococci Enterococcus faecalis Gram-negative bacilli Enterobacter cloacae Klebsiella pneumoniae Proteus mirabilis Pseudomonas aeruginosa Pseudomonas spp Enterobacter aerogenes Serratia marcescens Klebsiella oxytoca Providencia rettgeri Citrobacter spp
Femoral (n Z 26)
Radial (n Z 25)
9
16
1
2
6 2 1 4 1 1 1 1 1 0
1 3 1 1 0 0 0 0 0 1
Bacteremia related with arterial catheter have therapeutic implications (for example, for Gramnegative antibiotic coverage) when infection is suspected in the femoral location as concluded Lorente et al.18 In conclusion, the arterial catheter plays an important role in CR-BSI. We found no differences in infection rates in different locations. Days of insertion and length of ICU stay increase the risk of ACR-BSI. Finally, our results suggest that the femoral site increases the risk for Gram-negative infection.
Source of funding The study was financed by the Spanish department of health and consumption (PI 020132, Fondo de Investigaciones Sanitarias, Instituto deSalud Carlos III).
References 1. Pittet D, Tarara D, Wenzel R. Nosocomial bloodstream infection in critically ill patients. Excess length of stay, extra costs, an attributable mortality. JAMA 1994;271:1598e601. 2. Smith RL, Meixler SM, Simberkoff MS. Excess mortality in critically ill patients with nosocomial bloodstream infections. Chest 1991;100:164e7. 3. Dimick JB, Pelz RK, Consunji R, et al. Increased resource use associated with catheter-related bloodstream infection in the surgical intensive care unit. Arch Surg 2001;136:229e34. 4. Arnow PM, Quimosing EM, Beach M. Consequences of intravascular catheter sepsis. Clin Infect Dis 1993;16:778e84. ~ a C, Pujol M, Pallare n del coste atribu5. Pen ´s R, et al. Estimacio n nosocomial: prolongacio n de la estancia hosible a la infeccio lculo de los costes alternativos. Med Clin (Barc) pitalaria y ca 1996;106:441e4. 6. Rello J, Ochagavia A, Sabanes E, et al. Evaluation of outcome of intravenous catheter-related infections in critically ill patients. Am J Respir Crit Care Med 2000;162:1027e30. 7. Soufir L, Timsit JF, Mahe C, et al. Attributable morbidity and mortality of catheter-related septicemia in critically ill patient: a matched, risk-adjusted, cohort study. Infect Control Hosp Epidemiol 1999;20:396e401. 8. Ducharme FM, Gauthier M, Lacroix J, et al. Incidence of infection related to arterial catheterization in children: a prospective study. Crit Care Med 1988;16:272e6.
143 9. El-Hamamsy I, Durrleman N, Stevens LM, et al. Incidence and outcome of radial artery infections following cardiac surgery. Ann Thorac Surg 2003;76:801e4. 10. Raad I, Umphrey J, Khan A, et al. The duration of placement as a predictor of peripheral and pulmonary arterial catheter infections. J Hosp Infect 1993;23:17e26. 11. Rijjndersa BJA. Catheter-related can be prevented if we take the arterial line seriously too! Crit Care Med 2005;33:1437e9. 12. Traore O, Liottier J, Souweine B. Prospective study of arterial and central venous catheter colonization and of arterial and central venous catheter-related bacteremia in intensive care units. Crit Care Med 2005;33:1276e80. 13. Martin C, Saux P, Papazian L, et al. Long-term arterial cannulation in ICU patients using the radial artery or dorsalis pedis artery. Chest 2001;119:901e6. 14. Lorente L, Santacreu R, Martı´n MM, et al. Arterial catheterrelated infection of 2,949 catheters. Crit Care 2006;10(3):R83. 15. Koh DB, Gowardman JR, Rickard CM, et al. Prospective study of peripheral arterial catheter infection and comparison with concurrently sited central venous catheters. Crit Care Med 2008;36(2):397e402. 16. Lucet JC, Bouadma L, Zahar JR, et al. Infectious risk associated with arterial catheters compared with central venous catheters. Crit Care Med 2010;38(4):1030e5. 17. Gowardman JR, Lipman J, Rickard CM. Assesment of peripheral arterial catheters as a source of sepsis in the critically ill: a narrative review. J Hosp Infect 2010;75(1):12e8. 18. Lorente L, Jime ´nez A, Jime ´nez JJ, et al. The catheter site influences in the micro-organism responsable of arterial catheter-related infection. Intensive Care Med 2006;32: 1919e20. 19. Esteve F, Pujol M, Ariza J, et al. Impact of a prevention program for catheter-related bloodstream infection in the intensive care unit of a tertiary hospital. Enferm Infecc Microbiol Clin 2009 Dec;27(10):561e5. 20. Maki DG, Weise CE. Sarafin HW a semiquantitative culture method for identifying intravenous-catheter-related infection. N Engl J Med 1977;296:1305e9. 21. O’Grady NP, Alexander M, Kellinger EP, et al. Guidelines for the prevention of intravascular catheter-related infections. Centers for disease control and prevention. MMWR Recomm Rep 2002 Aug 9;51(RR-10):1e29. 22. Khalifa R, Dahyot-fizelier C, Laksiri L, et al. Indwelling time and risk of colonization of peripheral arterial catheters in critically ill patients. Intensive Care Med 2008;34: 1820e6.