- Email: [email protected]
Central Line Infections in Repaired Catheters: A Retrospective Review
O R I G I N A L
A R T I C L E
Central Line Infections in Repaired Catheters: A Retrospective Review Stuart Gordon, MSN, RN, CRNI, VA-BC Legacy Health, Portland, OR Stuart Gardiner, PhD Devers Eye Institute Research Laboratories, Portland, OR
Abstract Background: There have been very few studies conducted to assess the infection risk of repairing a ruptured or broken tunneled central venous access device or a ruptured peripherally inserted central catheter (PICC), a procedure that is fairly common in a certain population of patients. Methods: In a retrospective review of repairs to both tunneled central venous access devices and PICCs in a large metropolitan health system, 258 medical records were reviewed. During a 4-year period there were 258 repairs, 202 to PICC lines and 56 to tunneled catheters. The system-wide infection database was the source queried to provide evidence for and confirmation of a central line infection. This database is maintained by the infection control team using strict guidelines, reducing inter-rater reliability issues. Results: The Fisher exact test for proportions was used to compare infection rates between repaired infected and repaired noninfected lines. The infection rate was 5% in repaired catheters and 5.9% in unrepaired catheters (P ¼ 1.00). On average, repaired catheters were in place longer (mean log [time-days in situ] 2.71 vs 2.31). Despite repairs and longer dwell times the repaired catheters did not have a significantly higher rate of infection when compared with unrepaired catheters. Conclusions: Despite longer dwell times the infection rate for repaired catheters was not statistically significant when compared with unrepaired catheters. Keywords: PICC repair, permanent catheter, tunneled catheter, tunneled catheter repair
Introduction uptures of permanent, tunneled central venous access devices (CVADs) and certain types of peripherally inserted central catheter (PICC) lines are not uncommon and repair of a ruptured tunneled CVAD is often the least invasive and most viable intervention for certain populations of patients. Close to 5,000 Groshong catheter repair kits and more than 3,500 permanent (tunneled) catheter repair kits are sold in the United States annually; however, the number of repairs is most likely fewer than the number of kits sold because organizations keep several repair kits in inventory. There is a scarcity of studies determining the safety of repairs
R
Correspondence concerning this article should be addressed to [email protected] http://dx.doi.org/10.1016/j.java.2013.03.004 Copyright Ó 2013, ASSOCIATION FOR VASCULAR ACCESS. Published by Elsevier Inc. All rights reserved.
164
j
JAVA
j
Vol 18 No 3
j
to either tunneled catheters or to a leg of a temporary catheter such as a PICC. Alternatives to repair are over-the-wire exchange or removal and replacement in a different site. The latter exposes the patient to the complications associated with initial tunneled CAVD placement; that is, bleeding, infection, hemothorax, pneumothorax, deep vein thrombosis, intravascular stenosis, and a potential loss of vascular access. When faced with PICC repair one has to weigh options; exchange attempts often fail resulting in a loss of PICC access altogether, necessitating attempting access in a different site, or insertion of an internal jugular, subclavian, or femoral line. Methods This was a retrospective review of all CVAD repairs and central line-associated blood stream infections (CLABSIs) covering a 4-year span. To eliminate the possibility of overlooking a line repair the hospital billing database was queried for all tunneled catheter repairs and PICC repairs from January 1, 2008, through March 30, 2012. Of the 258 repairs billed
2013
during this time 202 were PICC (Groshong; Bard Access, Salt Lake City, UT) repairs and 56 were permanent CVAD repairs. During the same period, 8,584 repairable PICCs were placed; Groshong PICCs are the only type of PICCs that have a repairable legdany repair to a power-injectable PICC would result in a weaker device that is more prone to failure under pressure injection. There were 3 repaired and 35 unrepaired PICCs associated with an infection. The surgical implanted device database was searched using keywords Groshong, Hickman, Leonard, and Broviac to provide 241 tunneled catheters placed in surgery during the review period. This number excludes those placed outside the surgical suite, for instance those placed by interventional radiologists. The IV Resource database for central lines was queried to provide the dwell time for each repaired catheter and to provide an average dwell time for unrepaired catheters. Statit (Midasþ Statit Solutions Group, Tucson, AZ) was used to provide all infection data to eliminate any controversy over the interpretation of a central line-associated (or related) infection. Infection information entered into Statit is vetted by the infection control practitioners following the National Healthcare Safety Network definition for CLABSI. Any questionable infections are reviewed by a board-certified infectious disease physician. The Statit database was queried for all CVAD infections between January 1, 2008, and March 30, 2012. Patient medical record numbers, infection capture date, as well as the species of pathogen for all CLABSIs were entered into an Excel (Microsoft Corp, Redmond, WA) spreadsheet where the medical record numbers were cross-referenced with medical record numbers of any patient with a line repair charge using the find function. If a match was found, the date of the repair was checked against the date the infection was captured and both dates were referenced in the medical record as a verification of line type, placement date, and dwell time. The find duplicate function was used to find 51 catheters having had >1 repair. The admitting diagnosis for patients counted as having a PICC-associated blood stream infection reflected the admitting diagnosis for all patients with PICC lines; 30% of lines are placed for antibiotic therapy, followed by heart disease, burns, trauma, gastrointestinal disease (short gut), and cancer. The diagnosis for patients with repairs and a subsequent infection was 66% (n ¼ 2) had an infection actively being treated and 33% (n ¼ 1) had a PICC placed for administration of chemotherapy. Within the tunneled CVAD patients group the predominant admitting diagnosis was cancer followed by pediatric short gut syndrome. Again this was reflected in the diagnoses of patients who had a line repair and a subsequent infection; 50% (n ¼ 6) had a cancer diagnosis, 42% (n ¼ 5) had short gut syndrome, and 8% (n ¼ 1) were dependent on total parenteral nutrition due to a fistula. The probability of infection increases with the duration of implantation; therefore, within the group that acquired infections the durations of dwell were compared between the subjects with repaired and unrepaired catheters by performing a t test on log (duration of dwell in days), with use of a natural logarithmic transform of time (dwell days) to ensure an
approximately Gaussian distribution making a t test valid. Infection rates were compared between repaired and nonrepaired lines using the Fisher exact test for proportions. All repairs were performed by the intravenous team registered nurses following the institution’s policy. The segment of the catheter to be repaired is closed with a padded bulldog clamp and cleaned with alcohol swabs and friction followed by chlorhexadine, allowed to air dry, and then placed on a sterile drape. Using a sterile scissors the cleaned segment would be cut at a 90 angle between the clamp and the rupture; if the catheter being repaired is a PICC, the replacement hub would then be placed into the leg of the catheter and an oversleeve used to secure the hub in place. In the case of a tunneled CVAD, the correct French size replacement segment would be flushed with saline, clamped, and the splice tube inserted into the catheter lumen. A small amount of silicone adhesive is applied to the gap existing between both segments and an oversleeve advanced to cover the spliced segments. A blunt cannula is then used to apply silicone adhesive under the oversleeve and the line splinted until the adhesive is set. Aseptic technique is maintained throughout the repair; the gloves and scissors are sterile. Findings Medical record numbers reflecting repaired lines were redundantly checked against the medical record numbers of all central line-related infections. Of the 258 identified repairs, 202 were made to PICC lines, 56 to permanent catheters. For tunneled catheters, no difference was found in infection rate after catheter repair. There were 3 infections in 56 (5.4%) repaired catheters and 11 infections in 185 (5.9%) unrepaired catheters (P ¼ 1.00 based on Fisher exact test). For Groshong catheters, there were 3 infections in 202 repaired catheters and 35 infections in 8,156 unrepaired catheters. This proportion was higher for the repaired catheters, although the difference was not statistically significant (P ¼ .063 Fisher exact test). Post-hoc power calculation revealed a power of 56% with this sample size and population proportions. Repaired catheters were in place for longer periods. For repaired catheters, log (time [days] in situ) had a mean of 2.71 0.43. For unrepaired catheters, log (time [days] in situ) had mean of 2.31 1.14. Although this difference was not significant (P ¼ .212), it is notable that the repaired catheters did not have a significantly higher infection rate despite longer dwell times. Discussion To repair a CVAD, the external segment of the catheter is cleaned vigorously using chlorhexidine and friction for at least 30 seconds. If the leg of a PICC has ruptured, the entire leg is cleaned vigorously with chorhexidine and friction, again for 30 seconds. The ruptured segment is cut between the patient and the rupture using a sterile scissors. The cleansed and cut portion is then rested on a sterile field and the new segment is spliced to the remaining line. The scissors, the cleaned segment, and the new segment are manipulated using sterile gloves, and aseptic technique is maintained throughout;
2013
j
Vol 18 No 3
j
JAVA
j
165
however, a step-by-step recount of the procedure is not the intent of this article. Both permanent, tunneled catheters and PICCs are repaired in this fashion within the authors’ institutiondthe only difference being the permanent catheter splice is more securely attached using sterile silicone. If both catheters (PICC or tunneled) are handled and repaired in the same manner it is safe to assume the risk of bacterial ingress from the repair would be the same. Other important considerations when determining bacterial ingress would be the length of time the line integrity had been compromised and the manner in which the compromise occurred (eg, 1 child bit a hole through her Hickman catheter). Bacterial ingress into the bloodstream in numbers significant enough to cause a central line-associated infection occurs most frequently through the catheter insertion site or the hub of the catheter, with infrequent modes of entry being contaminated infusate or hematogenous seeding. Insertion of a CVAD has the potential to drag bacteria into the bloodstream, carried on the tools of insertion.1 Even with adequate preparation of the insertion site it is virtually impossible to rid the skin surface of all bacteria. Studies of biofilm reveal Staphylococcus epidermidis biofilm as deep as 5 skin cells in dry areas and 15 to 20 skin cells deep in moist areas.2 The hub of the catheter, another source of bacterial ingress is believed to be the cause of infections that occur >10 days after insertion.3e6 With poor adherence to swabbing needleless connectors, bacteria present on the needleless connector will be injected directly into the bloodstream, whereas other cells will adhere to the intraluminal surface of the catheter and in time, will establish a biofilm. Colonization of invasive lines can occur within hours to days of insertion.7e9 In the case of a repair, it is possible bacteria gain entry through the rupture or during the repair itself through poor adherence to aseptic technique. However, given the results of research indicating early colonization of CVADs and a greater density of colonization in the proximal segment of intravascular devices, it is equally likely that the physical manipulation of the line may result in mechanical disruption and dislodgement of existing biofilm, especially in those CVADs with long dwell times.10 Subsequent flushing of the device may result in the dislodged biofilm entering the bloodstream, serving as an inoculum. If disruption of existing biofilm was a direct result of catheter manipulation one could expect a fairly rapid onset of symptoms. Limitations Limitations to this study include a small sample size (PICC n ¼ 202, tunneled CVAD n ¼ 56), the use of National Healthcare Safety Network guidelines for the definition of CLABSI, a potential inter-rater biasdalthough the latter was minimized
166
j
JAVA
j
Vol 18 No 3
j
by the use of set criteria for CLABSI, and the study itself being retrospective. Conclusions Our study did not find catheter repairs increased the rate of infection for tunneled catheters or PICC lines despite longer dwell times. Given the lack of published studies, larger studies are needed to determine if catheter repair significantly increases the risk of infections, and under which conditions the procedure should be done, if at all. Although many in the vascular access field believe that PICC repair is unwarranted, our study shows the procedure does not significantly increase the rate of infection. References 1. Jeske C, Raedler C, Goedecke A, et al. Early identification of bacteria leading to central venous catheter contamination. Anesthes Analg. 2003;97(4):940-943. 2. Costerton W. The Biofilm Primer. Berlin, Germany: Springer; 2007. 3. Raad I, Costerton W, Sabharwal U, Sacilowski M, Anaissie E, Bodey G. Ultrastructural analysis of indwelling vascular catheters: a quantitative relationship between luminal colonization and duration of placement. J Infect Dis. 1993;168(2):400-407. 4. Crnich C, Maki D. The promise of novel technology for the prevention of intravascular device-related bloodstream infection. I. Pathogenesis and short-term devices. Healthcare Epidemiol. 2002;34(9):1232-1242. 5. Safdar N, Maki D. The pathogenesis of catheter related bloodstream infection with noncuffed short-term central venous catheters. Intensive Care Med. 2004;30(1):62-67. 6. The Joint Commission. Preventing Central LineAssociated Bloodstream Infections: A Global Challenge, a Global Perspective. Oakbrook Terrace, IL: The Joint Commission; 2012. 7. Passerini L, Costerton W, King G. Biofilms on indwelling catheters. Crit Care Med. 1992;20(5):665-673. 8. Elliot T, Moss H, Tebbs S, et al. Novel approach to investigate a source of microbial contamination of central venous catheters. Euro J Clin Microbiol Infect Dis. 1997;16(3):210-213. 9. Lewis R, Raad I. Treatment protocols for infections of vascular catheters. In: Pace J, Rupp M, Finch R, eds. Biofilms, Infection, and Antimicrobial Therapy. Boca Raton, FL: Taylor & Francis; 2006:411-412. 10. Koh D, Robertson I, Watts M, Davies A. Density of microbial colonization on external and internal surfaces of concurrently placed intravascular devices. Am J Crit Care. 2012;21(3):162-171.
2013
A R T I C L E
Central Line Infections in Repaired Catheters: A Retrospective Review Stuart Gordon, MSN, RN, CRNI, VA-BC Legacy Health, Portland, OR Stuart Gardiner, PhD Devers Eye Institute Research Laboratories, Portland, OR
Abstract Background: There have been very few studies conducted to assess the infection risk of repairing a ruptured or broken tunneled central venous access device or a ruptured peripherally inserted central catheter (PICC), a procedure that is fairly common in a certain population of patients. Methods: In a retrospective review of repairs to both tunneled central venous access devices and PICCs in a large metropolitan health system, 258 medical records were reviewed. During a 4-year period there were 258 repairs, 202 to PICC lines and 56 to tunneled catheters. The system-wide infection database was the source queried to provide evidence for and confirmation of a central line infection. This database is maintained by the infection control team using strict guidelines, reducing inter-rater reliability issues. Results: The Fisher exact test for proportions was used to compare infection rates between repaired infected and repaired noninfected lines. The infection rate was 5% in repaired catheters and 5.9% in unrepaired catheters (P ¼ 1.00). On average, repaired catheters were in place longer (mean log [time-days in situ] 2.71 vs 2.31). Despite repairs and longer dwell times the repaired catheters did not have a significantly higher rate of infection when compared with unrepaired catheters. Conclusions: Despite longer dwell times the infection rate for repaired catheters was not statistically significant when compared with unrepaired catheters. Keywords: PICC repair, permanent catheter, tunneled catheter, tunneled catheter repair
Introduction uptures of permanent, tunneled central venous access devices (CVADs) and certain types of peripherally inserted central catheter (PICC) lines are not uncommon and repair of a ruptured tunneled CVAD is often the least invasive and most viable intervention for certain populations of patients. Close to 5,000 Groshong catheter repair kits and more than 3,500 permanent (tunneled) catheter repair kits are sold in the United States annually; however, the number of repairs is most likely fewer than the number of kits sold because organizations keep several repair kits in inventory. There is a scarcity of studies determining the safety of repairs
R
Correspondence concerning this article should be addressed to [email protected] http://dx.doi.org/10.1016/j.java.2013.03.004 Copyright Ó 2013, ASSOCIATION FOR VASCULAR ACCESS. Published by Elsevier Inc. All rights reserved.
164
j
JAVA
j
Vol 18 No 3
j
to either tunneled catheters or to a leg of a temporary catheter such as a PICC. Alternatives to repair are over-the-wire exchange or removal and replacement in a different site. The latter exposes the patient to the complications associated with initial tunneled CAVD placement; that is, bleeding, infection, hemothorax, pneumothorax, deep vein thrombosis, intravascular stenosis, and a potential loss of vascular access. When faced with PICC repair one has to weigh options; exchange attempts often fail resulting in a loss of PICC access altogether, necessitating attempting access in a different site, or insertion of an internal jugular, subclavian, or femoral line. Methods This was a retrospective review of all CVAD repairs and central line-associated blood stream infections (CLABSIs) covering a 4-year span. To eliminate the possibility of overlooking a line repair the hospital billing database was queried for all tunneled catheter repairs and PICC repairs from January 1, 2008, through March 30, 2012. Of the 258 repairs billed
2013
during this time 202 were PICC (Groshong; Bard Access, Salt Lake City, UT) repairs and 56 were permanent CVAD repairs. During the same period, 8,584 repairable PICCs were placed; Groshong PICCs are the only type of PICCs that have a repairable legdany repair to a power-injectable PICC would result in a weaker device that is more prone to failure under pressure injection. There were 3 repaired and 35 unrepaired PICCs associated with an infection. The surgical implanted device database was searched using keywords Groshong, Hickman, Leonard, and Broviac to provide 241 tunneled catheters placed in surgery during the review period. This number excludes those placed outside the surgical suite, for instance those placed by interventional radiologists. The IV Resource database for central lines was queried to provide the dwell time for each repaired catheter and to provide an average dwell time for unrepaired catheters. Statit (Midasþ Statit Solutions Group, Tucson, AZ) was used to provide all infection data to eliminate any controversy over the interpretation of a central line-associated (or related) infection. Infection information entered into Statit is vetted by the infection control practitioners following the National Healthcare Safety Network definition for CLABSI. Any questionable infections are reviewed by a board-certified infectious disease physician. The Statit database was queried for all CVAD infections between January 1, 2008, and March 30, 2012. Patient medical record numbers, infection capture date, as well as the species of pathogen for all CLABSIs were entered into an Excel (Microsoft Corp, Redmond, WA) spreadsheet where the medical record numbers were cross-referenced with medical record numbers of any patient with a line repair charge using the find function. If a match was found, the date of the repair was checked against the date the infection was captured and both dates were referenced in the medical record as a verification of line type, placement date, and dwell time. The find duplicate function was used to find 51 catheters having had >1 repair. The admitting diagnosis for patients counted as having a PICC-associated blood stream infection reflected the admitting diagnosis for all patients with PICC lines; 30% of lines are placed for antibiotic therapy, followed by heart disease, burns, trauma, gastrointestinal disease (short gut), and cancer. The diagnosis for patients with repairs and a subsequent infection was 66% (n ¼ 2) had an infection actively being treated and 33% (n ¼ 1) had a PICC placed for administration of chemotherapy. Within the tunneled CVAD patients group the predominant admitting diagnosis was cancer followed by pediatric short gut syndrome. Again this was reflected in the diagnoses of patients who had a line repair and a subsequent infection; 50% (n ¼ 6) had a cancer diagnosis, 42% (n ¼ 5) had short gut syndrome, and 8% (n ¼ 1) were dependent on total parenteral nutrition due to a fistula. The probability of infection increases with the duration of implantation; therefore, within the group that acquired infections the durations of dwell were compared between the subjects with repaired and unrepaired catheters by performing a t test on log (duration of dwell in days), with use of a natural logarithmic transform of time (dwell days) to ensure an
approximately Gaussian distribution making a t test valid. Infection rates were compared between repaired and nonrepaired lines using the Fisher exact test for proportions. All repairs were performed by the intravenous team registered nurses following the institution’s policy. The segment of the catheter to be repaired is closed with a padded bulldog clamp and cleaned with alcohol swabs and friction followed by chlorhexadine, allowed to air dry, and then placed on a sterile drape. Using a sterile scissors the cleaned segment would be cut at a 90 angle between the clamp and the rupture; if the catheter being repaired is a PICC, the replacement hub would then be placed into the leg of the catheter and an oversleeve used to secure the hub in place. In the case of a tunneled CVAD, the correct French size replacement segment would be flushed with saline, clamped, and the splice tube inserted into the catheter lumen. A small amount of silicone adhesive is applied to the gap existing between both segments and an oversleeve advanced to cover the spliced segments. A blunt cannula is then used to apply silicone adhesive under the oversleeve and the line splinted until the adhesive is set. Aseptic technique is maintained throughout the repair; the gloves and scissors are sterile. Findings Medical record numbers reflecting repaired lines were redundantly checked against the medical record numbers of all central line-related infections. Of the 258 identified repairs, 202 were made to PICC lines, 56 to permanent catheters. For tunneled catheters, no difference was found in infection rate after catheter repair. There were 3 infections in 56 (5.4%) repaired catheters and 11 infections in 185 (5.9%) unrepaired catheters (P ¼ 1.00 based on Fisher exact test). For Groshong catheters, there were 3 infections in 202 repaired catheters and 35 infections in 8,156 unrepaired catheters. This proportion was higher for the repaired catheters, although the difference was not statistically significant (P ¼ .063 Fisher exact test). Post-hoc power calculation revealed a power of 56% with this sample size and population proportions. Repaired catheters were in place for longer periods. For repaired catheters, log (time [days] in situ) had a mean of 2.71 0.43. For unrepaired catheters, log (time [days] in situ) had mean of 2.31 1.14. Although this difference was not significant (P ¼ .212), it is notable that the repaired catheters did not have a significantly higher infection rate despite longer dwell times. Discussion To repair a CVAD, the external segment of the catheter is cleaned vigorously using chlorhexidine and friction for at least 30 seconds. If the leg of a PICC has ruptured, the entire leg is cleaned vigorously with chorhexidine and friction, again for 30 seconds. The ruptured segment is cut between the patient and the rupture using a sterile scissors. The cleansed and cut portion is then rested on a sterile field and the new segment is spliced to the remaining line. The scissors, the cleaned segment, and the new segment are manipulated using sterile gloves, and aseptic technique is maintained throughout;
2013
j
Vol 18 No 3
j
JAVA
j
165
however, a step-by-step recount of the procedure is not the intent of this article. Both permanent, tunneled catheters and PICCs are repaired in this fashion within the authors’ institutiondthe only difference being the permanent catheter splice is more securely attached using sterile silicone. If both catheters (PICC or tunneled) are handled and repaired in the same manner it is safe to assume the risk of bacterial ingress from the repair would be the same. Other important considerations when determining bacterial ingress would be the length of time the line integrity had been compromised and the manner in which the compromise occurred (eg, 1 child bit a hole through her Hickman catheter). Bacterial ingress into the bloodstream in numbers significant enough to cause a central line-associated infection occurs most frequently through the catheter insertion site or the hub of the catheter, with infrequent modes of entry being contaminated infusate or hematogenous seeding. Insertion of a CVAD has the potential to drag bacteria into the bloodstream, carried on the tools of insertion.1 Even with adequate preparation of the insertion site it is virtually impossible to rid the skin surface of all bacteria. Studies of biofilm reveal Staphylococcus epidermidis biofilm as deep as 5 skin cells in dry areas and 15 to 20 skin cells deep in moist areas.2 The hub of the catheter, another source of bacterial ingress is believed to be the cause of infections that occur >10 days after insertion.3e6 With poor adherence to swabbing needleless connectors, bacteria present on the needleless connector will be injected directly into the bloodstream, whereas other cells will adhere to the intraluminal surface of the catheter and in time, will establish a biofilm. Colonization of invasive lines can occur within hours to days of insertion.7e9 In the case of a repair, it is possible bacteria gain entry through the rupture or during the repair itself through poor adherence to aseptic technique. However, given the results of research indicating early colonization of CVADs and a greater density of colonization in the proximal segment of intravascular devices, it is equally likely that the physical manipulation of the line may result in mechanical disruption and dislodgement of existing biofilm, especially in those CVADs with long dwell times.10 Subsequent flushing of the device may result in the dislodged biofilm entering the bloodstream, serving as an inoculum. If disruption of existing biofilm was a direct result of catheter manipulation one could expect a fairly rapid onset of symptoms. Limitations Limitations to this study include a small sample size (PICC n ¼ 202, tunneled CVAD n ¼ 56), the use of National Healthcare Safety Network guidelines for the definition of CLABSI, a potential inter-rater biasdalthough the latter was minimized
166
j
JAVA
j
Vol 18 No 3
j
by the use of set criteria for CLABSI, and the study itself being retrospective. Conclusions Our study did not find catheter repairs increased the rate of infection for tunneled catheters or PICC lines despite longer dwell times. Given the lack of published studies, larger studies are needed to determine if catheter repair significantly increases the risk of infections, and under which conditions the procedure should be done, if at all. Although many in the vascular access field believe that PICC repair is unwarranted, our study shows the procedure does not significantly increase the rate of infection. References 1. Jeske C, Raedler C, Goedecke A, et al. Early identification of bacteria leading to central venous catheter contamination. Anesthes Analg. 2003;97(4):940-943. 2. Costerton W. The Biofilm Primer. Berlin, Germany: Springer; 2007. 3. Raad I, Costerton W, Sabharwal U, Sacilowski M, Anaissie E, Bodey G. Ultrastructural analysis of indwelling vascular catheters: a quantitative relationship between luminal colonization and duration of placement. J Infect Dis. 1993;168(2):400-407. 4. Crnich C, Maki D. The promise of novel technology for the prevention of intravascular device-related bloodstream infection. I. Pathogenesis and short-term devices. Healthcare Epidemiol. 2002;34(9):1232-1242. 5. Safdar N, Maki D. The pathogenesis of catheter related bloodstream infection with noncuffed short-term central venous catheters. Intensive Care Med. 2004;30(1):62-67. 6. The Joint Commission. Preventing Central LineAssociated Bloodstream Infections: A Global Challenge, a Global Perspective. Oakbrook Terrace, IL: The Joint Commission; 2012. 7. Passerini L, Costerton W, King G. Biofilms on indwelling catheters. Crit Care Med. 1992;20(5):665-673. 8. Elliot T, Moss H, Tebbs S, et al. Novel approach to investigate a source of microbial contamination of central venous catheters. Euro J Clin Microbiol Infect Dis. 1997;16(3):210-213. 9. Lewis R, Raad I. Treatment protocols for infections of vascular catheters. In: Pace J, Rupp M, Finch R, eds. Biofilms, Infection, and Antimicrobial Therapy. Boca Raton, FL: Taylor & Francis; 2006:411-412. 10. Koh D, Robertson I, Watts M, Davies A. Density of microbial colonization on external and internal surfaces of concurrently placed intravascular devices. Am J Crit Care. 2012;21(3):162-171.
2013