- Email: [email protected]
Peripherally inserted central venous catheters in the acute care setting: A safe alternative to high-risk short-term central venous catheters
Peripherally inserted central venous catheters in the acute care setting: A safe alternative to high-risk short-term central venous catheters Basel Al Raiy, MD,a Mohamad G. Fakih, MD, MPH,a,b Nicole Bryan-Nomides, MT, MS,b Debi Hopfner, RN,b Elizabeth Riegel, RN,c Trudy Nenninger, RN,d Janice Rey, MT(ASCP),b Susan Szpunar, PhD,e Pramodine Kale, PharmD,f and Riad Khatib, MDa Detroit, Michigan
Background: Peripherally inserted central venous catheters (PICCs) serve as an alternative to short-term central venous catheters (CVCs) for providing intravenous (IV) access in the hospital. It is not clear which device has a lower risk of central line–associated bloodstream infection (CLABSI). We compared CVC- and PICC-related CLABSI rates in the setting of an intervention to remove highrisk CVCs. Methods: We prospectively followed patients with CVCs in the non–intensive care units (ICUs) and those with PICCs hospital-wide. A team evaluated the need for the CVC and the risk of infection, recommended the discontinuation of unnecessary or high-risk CVCs, and suggested PICC insertion for patients requiring prolonged access. Data on age, gender, type of catheter, duration of catheter utilization, and the development of CLABSIs were obtained. Results: A total of 638 CVCs were placed for 4917 catheter-days, during which 12 patients had a CLABSI, for a rate of 2.4 per 1000 catheter-days. A total of 622 PICCs were placed for 5703 catheter-days, during which 13 patients had a CLABSI, for a rate of 2.3 per 1000 catheter-days. The median time to development of infection was significantly longer in the patients with a PICC (23 vs 13 days; P 5 .03). Conclusion: In the presence of active surveillance and intervention to remove unnecessary or high-risk CVCs, CVCs and PICCs had similar rates of CLABSIs. Given their longer time to the development of infection, PICCs may be a safe alternative for prolonged inpatient IV access. Key Words: Peripherally inserted central catheters; central venous catheters; central line associated blood stream infection; risk. Copyright ª 2010 by the Association for Professionals in Infection Control and Epidemiology, Inc. Published by Elsevier Inc. All rights reserved. (Am J Infect Control 2010;38:149-53.)
Central venous catheters (CVCs) are widely used in the acute hospital setting and are associated with a significant risk of bloodstream infection (BSI).1-4 Although CVCs are frequently required to initiate treatment in severely ill patients, they are associated with an increased risk of both infectious and mechanical complications depending on the site of insertion.1,5,6 From the Division of Infectious Diseases, Department of Medicine,a Infection Control Department,b Quality Management Department,c Case Management Department,d Medical Education,e and Pharmacy Department,f St John Hospital and Medical Center, Detroit, MI. Address correspondence to Mohamad G Fakih, MD, MPH, Division of Infectious Diseases, Department of Medicine, St John Hospital and Medical Center, 19251 Mack Avenue, Suite 190, Grosse Pointe Woods, MI 48236. E-mail: [email protected]. This study was presented in part at the 18th Annual Scientific Meeting of the Society for Healthcare Epidemiology of America, April 2008 (Abstract 116). Conflicts of interest: None to report. 0196-6553/$36.00 Copyright ª 2010 by the Association for Professionals in Infection Control and Epidemiology, Inc. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.ajic.2009.06.008
In addition, the conditions of placement correlate with the risk of infection.2,7-10 The use of peripherally inserted central catheters (PICCs) in the hospital setting has grown steadily in an attempt to supplement highrisk traditional nontunneled multiple-lumen CVCs in providing prolonged intravenous (IV) access.3,11 Controversy remains regarding which device has the lowest risk of central line–associated BSI (CLABSI). A recent study investigating PICC-associated infections in the hospital setting has questioned the trend of using PICCs instead of CVCs.11 The objective of the present study was to compare CLABSI rates in patients with CVCs and PICCs in the hospital setting in the presence of an intervention to reduce the use of high-risk CVCs.
METHODS Our facility is a 608-bed tertiary care teaching hospital. Our study included 2 components: prospective monitoring of CVCs in the non–intensive care units (ICUs) and prospective monitoring of PICCs hospitalwide. Institutional review board approval was obtained for both components of the study. The prospective monitoring of CVCs was done between May 2006 and 149
150
Al Raiy et al.
September 2007. Information on all CVCs used in the non-ICUs was obtained twice weekly from the different units and entered into a database (MIDAS1 Care Management). All of the CVCs were then followed until discontinuation. CVCs were placed by resident physicians, anesthesiologists, or attending surgeons under maximum barrier precautions unless the patient’s condition was emergent. We also established a ‘‘central line removal team’’ (CLRT) that addressed the need for CVCs in the non-ICUs beginning in May 2006. The team evaluated each CVC twice weekly (Tuesday and Thursday) for need and risk. If a catheter was deemed high risk (ie, placed emergently or placed in the femoral site) or if central access was no longer necessary, then the recommendation was made to remove the catheter and either place a peripheral IV catheter for short-term use or a PICC if the patient was deemed to require access for a longer period. The recommendations were conveyed to the patient’s physician, who made the ultimate decision as to catheter management. All cases were followed and evaluated for infection. Prospective monitoring of PICCs was done between November 2006 and June 2007. All PICCs were placed by either the ‘‘IV team’’ or the ‘‘interventional radiology’’ team using maximum barrier precautions; all were entered into the MIDAS1 Care Management database and followed from insertion to discontinuation or discharge. All PICCs placed and used for more than 24 hours in the hospital setting were included. No intervention (ie, no assessment of need or risk) was done for the patients with a PICC. For both the CVC and PICC groups, data on age, gender, catheter placement site, duration of use, team inserting the PICC (IV team or interventional radiology team), development of CLABSI, and time to development of CLABSI were obtained. All microbiologic data obtained during the targeted catheter use were retrospectively reviewed. For patients with positive blood cultures, a chart review was done to identify the source of bacteremia or candidemia. The CLABSI rate was defined as the number of CLABSIs divided by the total days of catheter use, multiplied by 1000. The CLABSIs were identified using Centers for Disease Control and Prevention criteria.12 Based on our findings from the original analysis of the data, a post hoc analysis was performed to compare the risk for candidemia in patients with a PICC and those with a CVC.
Statistical analysis Statistical analyses were done using SPSS 15.0. The x2 test was used to study the associations among categorical variables. The Student t-test and one-way analysis of variance, followed by pairwise comparisons using the Bonferroni correction, were used to compare means of continuous variables. The nonparametric
American Journal of Infection Control March 2010
Mann-Whitney U-test was used to compare medians. Forward stepwise logistic regression was done to evaluate potential predictors of CLABSIs. The z-test for the comparison of rates was used to compare CLABSI rates for the different catheter types. Finally, Kaplan-Meier analysis was used to study differences in time to infection in the patients with a CVC versus those with a PICC. A P value , .05 was considered to indicate statistical significance.
RESULTS CVC use and infection During the study period, 638 CVCs were placed over 17 months for a total of 4917 catheter-days (subclavian, n 5 242 [38%]; internal jugular, n 5 187 [29%]; femoral, n 5 209 [33%]). Of these, 169 (27%) were placed in the emergency department and 368 (17%) were placed in ICUs. The CVCs were placed by surgical residents (n 5 273; 43%), medical residents (n 5 116; 18%), or emergency medicine residents (n 5 164; 26%). Of the 209 femoral CVCs, 97 (35%) were placed by surgical residents, 63 (38%) were placed by emergency medicine residents, and 37 (32%) were placed by medical residents. Sixty CVCs were placed in the operating room (9%). The mean and median duration of CVC utilization was 7.7 and 7.0 days, respectively. The mean duration of utilization differed significantly by CVC type (P , .001). Pairwise significant differences were found between subclavian and femoral catheters (8.7 days vs 6.5 days; P , .001) and between internal jugular and femoral catheters (7.8 days vs 6.5 days; P 5 .03). The median duration of utilization was 8 days for subclavian catheters, 7 days for internal jugular catheters, and 5 days for femoral catheters. During the study period, 305 of the 638 CVCs (48%) were deemed to be unnecessary or higher risk by the CLRT. Of these, 107 (35%) were subclavian catheters, 94 (31%) were internal jugular catheters, and 104 (34%) were femoral catheters. The CLRT’s initial recommendation for catheter removal was followed in 214 cases (70%). The microbiology of the 12 CLABSIs included coagulasenegative staphylococci (n 5 3), methicillin-resistant Staphylococcus aureus (n 5 2), vancomycin-resistant enterococcus (n 5 1), and gram-negative organisms (n 5 6). The CLABSI rate for CVCs was 2.4 per 1000 catheter-days (Table 1). Cases that were noncompliant with the CLRT’s recommendation for discontinuation had a CLABSI rate of 3.9 per 1000 catheter-days.
PICC utilization and infection A total of 622 patients had PICCs placed over an 8-month period, for a total of 5703 catheter-days (IV team, n 5 379 [61%]; interventional radiology team,
Al Raiy et al.
www.ajicjournal.org Vol. 38 No. 2
Table 1. Rates of CLABSI for the different devices used
Catheter type Subclavian Internal jugular Femoral PICC
P value CLABSI (2-sided; compared Infected rate (cases/1000 with subclavian catheters, % catheter-days) CLABSI rate) 1.7 (4/242) 2.7 (5/187) 1.4 (3/209) 2.1 (13/638)
1.9 (4/2109) 3.4 (5/1454) 2.2 (3/1354) 2.3 (13/5703)
.28 .56 .50
n 5 243 [39%]). Of the PICCs, 108 (17%) were singlelumen lines and 514 (83%) were double-lumen lines. A total of 122 PICCs (20%) were placed in the ICUs. The mean and median duration of PICC utilization was 9.2 days and 6.0 days, respectively. A total of 119 patients were discharged with a PICC. The microbiology of the 13 CLABSIs identified included Candida species (n 5 5), coagulase-negative staphylococci (n 5 5), methicillin-resistant S aureus (n 5 2), and Serratia species (n 5 1). The CLABSI rate for PICCs was 2.3 per 1000 catheter-days (Table 1). All infected PICCs were doublelumen lines. There were no significant differences in infection between PICCs placed by the IV team and those placed by the interventional radiology team.
Comparing CVCs and PICCs The mean duration of CVC utilization was significantly lower than that of PICC utilization (7.7 6 5.0 days vs 9.2 6 9.9 days; P , .001). Of the 638 patients with a CVC, 253 (40%) had at least one blood culture drawn while the catheter was in use. Of these patients, 55 (22%) had a positive blood culture. In contrast, 242 patients (39%) with an indwelling PICC had blood cultures drawn while the catheter was in use; 68 of these patients (28%) had a positive blood culture. Catheter tip cultures were obtained from 52 CVCs (8.2%) and 22 PICCs (3.5%) (P , .001). Of the catheters cultured, 33 CVCs (63%) and 17 PICCs (77%) grew organisms using the semiquantitative roll method. A forward stepwise logistic regression that included the potential predictors of age, gender, catheter type, and duration of catheter use failed to show any significant association with CLABSI other than for the duration of catheter use (odds ratio 5 1.08; 95% confidence interval 5 1.05 to 1.12). When comparing patients with infection, the mean and median time to development of infection was significantly longer in those with a PICC (mean, 27.9 vs 13.3 days [P 5 .04]; median, 23 vs 13 days [P 5 .03]) (Table 2). The CLABSI rates also were compared in patients with PICCs and those with CVCs and noncompliant with the CLRT’s recommendation for discontinuation; these rates
151
Table 2. Relationship between the duration of catheter utilization and development of CLABSIs
Mean duration of catheter utilization, days 6 SD CVCs PICCs Median duration of catheter utilization, days CVCs PICCs
CLABSIs
No CLABSIs
P value (2-sided)
13.3 6 4.6 27.9 6 22.7
7.6 6 4.9 8.8 6 9.1
,.001 ,.001
13 23
7 6
,.001 ,.001
were 2.3 and 3.9 per 1000 catheter-days, respectively (P 5 .25). The time to development of infection was compared in patients with CVCs and those with PICCs using Kaplan-Meier analysis. As shown in Figure 1, time to infection was significantly longer in the patients with PICCs (log-rank test, P 5 .03). Censored cases refer to patients who contributed patient time to the KaplanMeier curve but did not develop an infection.
Candidemia risk and type of catheter used We evaluated the risk of candidemia in the 25 patients with CLABSIs. The duration of catheter utilization before the development of infection was significantly longer in patients with candidemia compared with those with bacteremia (mean, 30.2 days vs 18.6 days; P 5 .006); the mean number of daily antibiotic doses was higher as well (26.4 vs 13.8; P 5 .02). All of the patients with candidemia were female and had an indwelling PICC. The use of total parenteral nutrition was not a significant factor in univariate analysis. Multivariate analysis was not performed due to the multicollinearity between catheter type and gender.
DISCUSSION PICCs are being increasingly used to provide IV access in inpatient settings. These catheters are usually placed by either a specialized nurse or an interventional radiologist. In the setting of a large teaching institution, resident physicians place short-term CVCs. Many of the CVCs are used in non-ICUs.13 Unfortunately, the adherence of resident physicians to the guidelines of CVC insertion is poor.14 In addition, an experienced operator is essential to reduce the risk.15 Moreover, unjustified CVC utilization seems to occur more often in the non-ICUs than in ICUs.3,16 About 1/ 3 of CVCs used in the non-ICUs in our facility were placed in the femoral area. We implemented a process to evaluate CVCs in non-ICUs twice weekly and attempted to remove all those that were deemed unnecessary
152
Al Raiy et al.
American Journal of Infection Control March 2010
Fig 1. Kaplan-Meier survival curve of time to infection by type of catheter. or high risk. Our success rate reached 70% with the initial attempt. This practice was combined with promoting the utilization of PICCs based on the potential lower risk compared with femoral or emergently placed catheters. This process led to a decrease in the duration of femoral CVC utilization and a similar infection rate in PICCs and CVCs. Our CLABSI rate for both CVCs and PICCs was about 2 per 1000 catheter-days, lower than the rates reported in previous studies of CVC-associated BSIs acquired in non-ICUs.6,17 Moreover, in those patients with CLABSIs, the median and mean time to infection was 10 and 14 days longer, respectively, in the patients with PICCs than in those with CVCs. It is noteworthy that the cases with indwelling CVCs that were not compliant with the CLRT’s recommendation for removal had a higher CLABSI rate, 3.9 per 1000 catheterdays. Although this difference did not reach statistical significance, it may indicate that PICC utilization carries less risk. With the longer time to infection, PICCs are an attractive alternative to CVCs, especially in patients with a femoral catheter or those requiring prolonged inpatient IV access. It is important to note that the gold standard for CLABSI reporting is the rate of bacteremia- or candidemia-infected catheters per 1000 catheter-days.4 This does not completely adjust for risk, however; for example, it does not account for the number of times that the
device was accessed per day, which may better reflect the risk related to catheter hub and lumen contamination for catheters used for longer than 10 days.18,19 Stratifying catheter exposure per patient over a specified time period may improve our understanding of the risk. In one study, the rate of BSI was 2.1 per 1000 catheter-days in patients with catheters in place for 1 to 5 days, increasing to 10.2 per 1000 catheterdays in patients with catheters in place for 16 to 30 days.20 Our survival analysis showed a significantly lower time to infection for CVCs compared with PICCs. For hospitalized patients requiring prolonged IV access, PICCs may carry less risk than CVCs. The recently published Society for Healthcare Epidemiology of America/Infectious Diseases Society of America practice recommendations for prevention of CLABSIs in acute care hospitals list PICCs as a non– evidence-based strategy to reduce the CLABSI rate.21 This recommendation was based on a study comparing PICC infections and historical controls for CVC infections, however.11 We believe that hospital characteristics play an important role in infection outcome. Comparing PICC and CVC infections in the same setting will provide a better assessment of the risk. We also support the recommendations highlighting the importance of leadership support and accountability to ensure that health care personnel are competent
www.ajicjournal.org Vol. 38 No. 2
to perform their work.21 Through partnering by infection control, case management, quality management, and nursing supported by the hospital leadership, we were able to implement our process. References 1. O’Grady NP, Alexander M, Dellinger P, Gerberding JL, Heard SO Maki DG, et al. Guidelines for the prevention of intravascular catheter–related infections. Infect Control Hosp Epidemiol 2002;23:759-69. 2. Braun BI, Kritchevsky SB, Wong ES, Solomon SL, Steele L, Richards CL, et al. Preventing central venous catheter–associated primary bloodstream infections: characteristics of practices among hospitals participating in the Evaluation of Processes and Indicators in Infection Control (EPIC) Study. Infect Control Hosp Epidemiol 2003;24:926-35. 3. Climo M, Diekema D, Warren DK, Herwaldt LA, Perl TM, Peterson L, et al. Prevalence of the use of central venous access devices within and outside of the intensive care unit: results of a survey among hospitals in the Prevention Epicenter Program of the Centers for Disease Control and Prevention. Infect Control Hosp Epidemiol 2003;24:942-5. 4. Maki DG, Kluger DM, Crnich CJ. The risk of bloodstream infection in adults with different intravascular devices: a systemic review of 200 published prospective studies. Mayo Clin Proc 2006;81:1159-71. 5. Merrer J, De Jonghe B, Golliot F, Lefrant JY, Raffy B, Barre E, et al. Complications of femoral and subclavian venous catheterization in critically ill patients: a randomized controlled trial. JAMA 2001;286:700-7. 6. Vonberg RP, Behnke M, Geffers C, Sohr D, Ruden H, Dettenkofer M, et al. Device-associated infection rates for non–intensive care unit patients. Infect Control Hosp Epidemiol 2006;27:357-61. 7. Warren DK, Cosgrove SE, Diekema DJ, Zuccotti G, Climo MW Bolon MK, et al. A multicenter intervention to prevent catheter-associated bloodstream infections. Infect Control Hosp Epidemiol 2006; 27:662-9. 8. Gnass SA, Barbosa L, Bilicich D, Angeloro P, Treiyer W, Grenovero S, et al. Prevention of central venous catheter–related bloodstream infections using non-technologic strategies. Infect Control Hosp Epidemiol 2004;25:675-7. 9. Berenholtz SM, Pronovost PJ, Lipsett PA, Hobson D, Earsing K, Farley JE, et al. Eliminating catheter-related bloodstream infections in the intensive care unit. Crit Care Med 2004;32:2014-20.
Al Raiy et al.
153
10. Lee DH, Jung KY, Choi YH. Use of maximal sterile barrier precautions and/or antimicrobial-coated catheters to reduce the risk of central venous catheter–related bloodstream infection. Infect Control Hosp Epidemiol 2008;29:947-50. 11. Safdar N, Maki DG. Risk of catheter-related bloodstream infection with peripherally inserted central venous catheters used in hospitalized patients. Chest 2005;128:489-95. 12. Horan TC, Gaynes RP. Surveillance of nosocomial infections. In: Mayhall CG, editor. Hospital epidemiology and infection control. 3rd ed Philadelphia: Lippincott Williams & Wilkins; 2004. p. 1659-702. 13. Weber DJ, Sickbert-Bennett EE, Brown V, Rutala WA. Comparison of hospital-wide surveillance and targeted intensive care unit surveillance of healthcare-associated infections. Infect Control Hosp Epidemiol 2007;28:1361-6. 14. Aloul B, Adabag AS, Goussous TY, Weinert CR. Medical resident’s self-reported adherence to guidelines during placement of central venous catheters. Infect Control Hosp Epidemiol 2008;9:86-8. 15. Taylor RW, Palagiri AV. Central venous catheterization. Crit Care Med 2007;35:1390-6. 16. Trick WE, Vernon MO, Welbel SF, Wisniewski MF, Jernigan JA, Weinstein RA. Unnecessary use of central venous catheters: the need to look outside the intensive care unit. Infect Control Hosp Epidemiol 2004;25:266-8. 17. Marschall J, Leone C, Jones M, Nihill D, Fraser VJ, Warren DK. Catheter-associated bloodstream infections in general medical patients outside the intensive care unit: a surveillance study. Infect Control Hosp Epidemiol 2007;28:905-9. 18. Crnich CJ, Maki DG. The promise of novel technology for the prevention of intravascular device–related bloodstream infection, I: pathogenesis and short-term devices. Clin Infect Dis 2002;34:1232-42. 19. Tokars JI, Klevens M, Edwards JR, Horan TC. Measurement of the impact of risk adjustment for central line-days on interpretation of central line–associated bloodstream infection rates. Infect Control Hosp Epidemiol 2007;28:1025-9. 20. Mclaws ML, Berry G. Nonuniform risk of bloodstream infection with increasing central venous catheter-days. Infect Control Hosp Epidemiol 2005;26:715-9. 21. Marschall J, Mermel LA, Classen D, Arias KM, Podgorny K, Anderson DJ, et al. Strategies to prevent central line–associated bloodstream infections in acute care hospitals. Infect Control Hosp Epidemiol 2008;29: S22-30.
Background: Peripherally inserted central venous catheters (PICCs) serve as an alternative to short-term central venous catheters (CVCs) for providing intravenous (IV) access in the hospital. It is not clear which device has a lower risk of central line–associated bloodstream infection (CLABSI). We compared CVC- and PICC-related CLABSI rates in the setting of an intervention to remove highrisk CVCs. Methods: We prospectively followed patients with CVCs in the non–intensive care units (ICUs) and those with PICCs hospital-wide. A team evaluated the need for the CVC and the risk of infection, recommended the discontinuation of unnecessary or high-risk CVCs, and suggested PICC insertion for patients requiring prolonged access. Data on age, gender, type of catheter, duration of catheter utilization, and the development of CLABSIs were obtained. Results: A total of 638 CVCs were placed for 4917 catheter-days, during which 12 patients had a CLABSI, for a rate of 2.4 per 1000 catheter-days. A total of 622 PICCs were placed for 5703 catheter-days, during which 13 patients had a CLABSI, for a rate of 2.3 per 1000 catheter-days. The median time to development of infection was significantly longer in the patients with a PICC (23 vs 13 days; P 5 .03). Conclusion: In the presence of active surveillance and intervention to remove unnecessary or high-risk CVCs, CVCs and PICCs had similar rates of CLABSIs. Given their longer time to the development of infection, PICCs may be a safe alternative for prolonged inpatient IV access. Key Words: Peripherally inserted central catheters; central venous catheters; central line associated blood stream infection; risk. Copyright ª 2010 by the Association for Professionals in Infection Control and Epidemiology, Inc. Published by Elsevier Inc. All rights reserved. (Am J Infect Control 2010;38:149-53.)
Central venous catheters (CVCs) are widely used in the acute hospital setting and are associated with a significant risk of bloodstream infection (BSI).1-4 Although CVCs are frequently required to initiate treatment in severely ill patients, they are associated with an increased risk of both infectious and mechanical complications depending on the site of insertion.1,5,6 From the Division of Infectious Diseases, Department of Medicine,a Infection Control Department,b Quality Management Department,c Case Management Department,d Medical Education,e and Pharmacy Department,f St John Hospital and Medical Center, Detroit, MI. Address correspondence to Mohamad G Fakih, MD, MPH, Division of Infectious Diseases, Department of Medicine, St John Hospital and Medical Center, 19251 Mack Avenue, Suite 190, Grosse Pointe Woods, MI 48236. E-mail: [email protected]. This study was presented in part at the 18th Annual Scientific Meeting of the Society for Healthcare Epidemiology of America, April 2008 (Abstract 116). Conflicts of interest: None to report. 0196-6553/$36.00 Copyright ª 2010 by the Association for Professionals in Infection Control and Epidemiology, Inc. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.ajic.2009.06.008
In addition, the conditions of placement correlate with the risk of infection.2,7-10 The use of peripherally inserted central catheters (PICCs) in the hospital setting has grown steadily in an attempt to supplement highrisk traditional nontunneled multiple-lumen CVCs in providing prolonged intravenous (IV) access.3,11 Controversy remains regarding which device has the lowest risk of central line–associated BSI (CLABSI). A recent study investigating PICC-associated infections in the hospital setting has questioned the trend of using PICCs instead of CVCs.11 The objective of the present study was to compare CLABSI rates in patients with CVCs and PICCs in the hospital setting in the presence of an intervention to reduce the use of high-risk CVCs.
METHODS Our facility is a 608-bed tertiary care teaching hospital. Our study included 2 components: prospective monitoring of CVCs in the non–intensive care units (ICUs) and prospective monitoring of PICCs hospitalwide. Institutional review board approval was obtained for both components of the study. The prospective monitoring of CVCs was done between May 2006 and 149
150
Al Raiy et al.
September 2007. Information on all CVCs used in the non-ICUs was obtained twice weekly from the different units and entered into a database (MIDAS1 Care Management). All of the CVCs were then followed until discontinuation. CVCs were placed by resident physicians, anesthesiologists, or attending surgeons under maximum barrier precautions unless the patient’s condition was emergent. We also established a ‘‘central line removal team’’ (CLRT) that addressed the need for CVCs in the non-ICUs beginning in May 2006. The team evaluated each CVC twice weekly (Tuesday and Thursday) for need and risk. If a catheter was deemed high risk (ie, placed emergently or placed in the femoral site) or if central access was no longer necessary, then the recommendation was made to remove the catheter and either place a peripheral IV catheter for short-term use or a PICC if the patient was deemed to require access for a longer period. The recommendations were conveyed to the patient’s physician, who made the ultimate decision as to catheter management. All cases were followed and evaluated for infection. Prospective monitoring of PICCs was done between November 2006 and June 2007. All PICCs were placed by either the ‘‘IV team’’ or the ‘‘interventional radiology’’ team using maximum barrier precautions; all were entered into the MIDAS1 Care Management database and followed from insertion to discontinuation or discharge. All PICCs placed and used for more than 24 hours in the hospital setting were included. No intervention (ie, no assessment of need or risk) was done for the patients with a PICC. For both the CVC and PICC groups, data on age, gender, catheter placement site, duration of use, team inserting the PICC (IV team or interventional radiology team), development of CLABSI, and time to development of CLABSI were obtained. All microbiologic data obtained during the targeted catheter use were retrospectively reviewed. For patients with positive blood cultures, a chart review was done to identify the source of bacteremia or candidemia. The CLABSI rate was defined as the number of CLABSIs divided by the total days of catheter use, multiplied by 1000. The CLABSIs were identified using Centers for Disease Control and Prevention criteria.12 Based on our findings from the original analysis of the data, a post hoc analysis was performed to compare the risk for candidemia in patients with a PICC and those with a CVC.
Statistical analysis Statistical analyses were done using SPSS 15.0. The x2 test was used to study the associations among categorical variables. The Student t-test and one-way analysis of variance, followed by pairwise comparisons using the Bonferroni correction, were used to compare means of continuous variables. The nonparametric
American Journal of Infection Control March 2010
Mann-Whitney U-test was used to compare medians. Forward stepwise logistic regression was done to evaluate potential predictors of CLABSIs. The z-test for the comparison of rates was used to compare CLABSI rates for the different catheter types. Finally, Kaplan-Meier analysis was used to study differences in time to infection in the patients with a CVC versus those with a PICC. A P value , .05 was considered to indicate statistical significance.
RESULTS CVC use and infection During the study period, 638 CVCs were placed over 17 months for a total of 4917 catheter-days (subclavian, n 5 242 [38%]; internal jugular, n 5 187 [29%]; femoral, n 5 209 [33%]). Of these, 169 (27%) were placed in the emergency department and 368 (17%) were placed in ICUs. The CVCs were placed by surgical residents (n 5 273; 43%), medical residents (n 5 116; 18%), or emergency medicine residents (n 5 164; 26%). Of the 209 femoral CVCs, 97 (35%) were placed by surgical residents, 63 (38%) were placed by emergency medicine residents, and 37 (32%) were placed by medical residents. Sixty CVCs were placed in the operating room (9%). The mean and median duration of CVC utilization was 7.7 and 7.0 days, respectively. The mean duration of utilization differed significantly by CVC type (P , .001). Pairwise significant differences were found between subclavian and femoral catheters (8.7 days vs 6.5 days; P , .001) and between internal jugular and femoral catheters (7.8 days vs 6.5 days; P 5 .03). The median duration of utilization was 8 days for subclavian catheters, 7 days for internal jugular catheters, and 5 days for femoral catheters. During the study period, 305 of the 638 CVCs (48%) were deemed to be unnecessary or higher risk by the CLRT. Of these, 107 (35%) were subclavian catheters, 94 (31%) were internal jugular catheters, and 104 (34%) were femoral catheters. The CLRT’s initial recommendation for catheter removal was followed in 214 cases (70%). The microbiology of the 12 CLABSIs included coagulasenegative staphylococci (n 5 3), methicillin-resistant Staphylococcus aureus (n 5 2), vancomycin-resistant enterococcus (n 5 1), and gram-negative organisms (n 5 6). The CLABSI rate for CVCs was 2.4 per 1000 catheter-days (Table 1). Cases that were noncompliant with the CLRT’s recommendation for discontinuation had a CLABSI rate of 3.9 per 1000 catheter-days.
PICC utilization and infection A total of 622 patients had PICCs placed over an 8-month period, for a total of 5703 catheter-days (IV team, n 5 379 [61%]; interventional radiology team,
Al Raiy et al.
www.ajicjournal.org Vol. 38 No. 2
Table 1. Rates of CLABSI for the different devices used
Catheter type Subclavian Internal jugular Femoral PICC
P value CLABSI (2-sided; compared Infected rate (cases/1000 with subclavian catheters, % catheter-days) CLABSI rate) 1.7 (4/242) 2.7 (5/187) 1.4 (3/209) 2.1 (13/638)
1.9 (4/2109) 3.4 (5/1454) 2.2 (3/1354) 2.3 (13/5703)
.28 .56 .50
n 5 243 [39%]). Of the PICCs, 108 (17%) were singlelumen lines and 514 (83%) were double-lumen lines. A total of 122 PICCs (20%) were placed in the ICUs. The mean and median duration of PICC utilization was 9.2 days and 6.0 days, respectively. A total of 119 patients were discharged with a PICC. The microbiology of the 13 CLABSIs identified included Candida species (n 5 5), coagulase-negative staphylococci (n 5 5), methicillin-resistant S aureus (n 5 2), and Serratia species (n 5 1). The CLABSI rate for PICCs was 2.3 per 1000 catheter-days (Table 1). All infected PICCs were doublelumen lines. There were no significant differences in infection between PICCs placed by the IV team and those placed by the interventional radiology team.
Comparing CVCs and PICCs The mean duration of CVC utilization was significantly lower than that of PICC utilization (7.7 6 5.0 days vs 9.2 6 9.9 days; P , .001). Of the 638 patients with a CVC, 253 (40%) had at least one blood culture drawn while the catheter was in use. Of these patients, 55 (22%) had a positive blood culture. In contrast, 242 patients (39%) with an indwelling PICC had blood cultures drawn while the catheter was in use; 68 of these patients (28%) had a positive blood culture. Catheter tip cultures were obtained from 52 CVCs (8.2%) and 22 PICCs (3.5%) (P , .001). Of the catheters cultured, 33 CVCs (63%) and 17 PICCs (77%) grew organisms using the semiquantitative roll method. A forward stepwise logistic regression that included the potential predictors of age, gender, catheter type, and duration of catheter use failed to show any significant association with CLABSI other than for the duration of catheter use (odds ratio 5 1.08; 95% confidence interval 5 1.05 to 1.12). When comparing patients with infection, the mean and median time to development of infection was significantly longer in those with a PICC (mean, 27.9 vs 13.3 days [P 5 .04]; median, 23 vs 13 days [P 5 .03]) (Table 2). The CLABSI rates also were compared in patients with PICCs and those with CVCs and noncompliant with the CLRT’s recommendation for discontinuation; these rates
151
Table 2. Relationship between the duration of catheter utilization and development of CLABSIs
Mean duration of catheter utilization, days 6 SD CVCs PICCs Median duration of catheter utilization, days CVCs PICCs
CLABSIs
No CLABSIs
P value (2-sided)
13.3 6 4.6 27.9 6 22.7
7.6 6 4.9 8.8 6 9.1
,.001 ,.001
13 23
7 6
,.001 ,.001
were 2.3 and 3.9 per 1000 catheter-days, respectively (P 5 .25). The time to development of infection was compared in patients with CVCs and those with PICCs using Kaplan-Meier analysis. As shown in Figure 1, time to infection was significantly longer in the patients with PICCs (log-rank test, P 5 .03). Censored cases refer to patients who contributed patient time to the KaplanMeier curve but did not develop an infection.
Candidemia risk and type of catheter used We evaluated the risk of candidemia in the 25 patients with CLABSIs. The duration of catheter utilization before the development of infection was significantly longer in patients with candidemia compared with those with bacteremia (mean, 30.2 days vs 18.6 days; P 5 .006); the mean number of daily antibiotic doses was higher as well (26.4 vs 13.8; P 5 .02). All of the patients with candidemia were female and had an indwelling PICC. The use of total parenteral nutrition was not a significant factor in univariate analysis. Multivariate analysis was not performed due to the multicollinearity between catheter type and gender.
DISCUSSION PICCs are being increasingly used to provide IV access in inpatient settings. These catheters are usually placed by either a specialized nurse or an interventional radiologist. In the setting of a large teaching institution, resident physicians place short-term CVCs. Many of the CVCs are used in non-ICUs.13 Unfortunately, the adherence of resident physicians to the guidelines of CVC insertion is poor.14 In addition, an experienced operator is essential to reduce the risk.15 Moreover, unjustified CVC utilization seems to occur more often in the non-ICUs than in ICUs.3,16 About 1/ 3 of CVCs used in the non-ICUs in our facility were placed in the femoral area. We implemented a process to evaluate CVCs in non-ICUs twice weekly and attempted to remove all those that were deemed unnecessary
152
Al Raiy et al.
American Journal of Infection Control March 2010
Fig 1. Kaplan-Meier survival curve of time to infection by type of catheter. or high risk. Our success rate reached 70% with the initial attempt. This practice was combined with promoting the utilization of PICCs based on the potential lower risk compared with femoral or emergently placed catheters. This process led to a decrease in the duration of femoral CVC utilization and a similar infection rate in PICCs and CVCs. Our CLABSI rate for both CVCs and PICCs was about 2 per 1000 catheter-days, lower than the rates reported in previous studies of CVC-associated BSIs acquired in non-ICUs.6,17 Moreover, in those patients with CLABSIs, the median and mean time to infection was 10 and 14 days longer, respectively, in the patients with PICCs than in those with CVCs. It is noteworthy that the cases with indwelling CVCs that were not compliant with the CLRT’s recommendation for removal had a higher CLABSI rate, 3.9 per 1000 catheterdays. Although this difference did not reach statistical significance, it may indicate that PICC utilization carries less risk. With the longer time to infection, PICCs are an attractive alternative to CVCs, especially in patients with a femoral catheter or those requiring prolonged inpatient IV access. It is important to note that the gold standard for CLABSI reporting is the rate of bacteremia- or candidemia-infected catheters per 1000 catheter-days.4 This does not completely adjust for risk, however; for example, it does not account for the number of times that the
device was accessed per day, which may better reflect the risk related to catheter hub and lumen contamination for catheters used for longer than 10 days.18,19 Stratifying catheter exposure per patient over a specified time period may improve our understanding of the risk. In one study, the rate of BSI was 2.1 per 1000 catheter-days in patients with catheters in place for 1 to 5 days, increasing to 10.2 per 1000 catheterdays in patients with catheters in place for 16 to 30 days.20 Our survival analysis showed a significantly lower time to infection for CVCs compared with PICCs. For hospitalized patients requiring prolonged IV access, PICCs may carry less risk than CVCs. The recently published Society for Healthcare Epidemiology of America/Infectious Diseases Society of America practice recommendations for prevention of CLABSIs in acute care hospitals list PICCs as a non– evidence-based strategy to reduce the CLABSI rate.21 This recommendation was based on a study comparing PICC infections and historical controls for CVC infections, however.11 We believe that hospital characteristics play an important role in infection outcome. Comparing PICC and CVC infections in the same setting will provide a better assessment of the risk. We also support the recommendations highlighting the importance of leadership support and accountability to ensure that health care personnel are competent
www.ajicjournal.org Vol. 38 No. 2
to perform their work.21 Through partnering by infection control, case management, quality management, and nursing supported by the hospital leadership, we were able to implement our process. References 1. O’Grady NP, Alexander M, Dellinger P, Gerberding JL, Heard SO Maki DG, et al. Guidelines for the prevention of intravascular catheter–related infections. Infect Control Hosp Epidemiol 2002;23:759-69. 2. Braun BI, Kritchevsky SB, Wong ES, Solomon SL, Steele L, Richards CL, et al. Preventing central venous catheter–associated primary bloodstream infections: characteristics of practices among hospitals participating in the Evaluation of Processes and Indicators in Infection Control (EPIC) Study. Infect Control Hosp Epidemiol 2003;24:926-35. 3. Climo M, Diekema D, Warren DK, Herwaldt LA, Perl TM, Peterson L, et al. Prevalence of the use of central venous access devices within and outside of the intensive care unit: results of a survey among hospitals in the Prevention Epicenter Program of the Centers for Disease Control and Prevention. Infect Control Hosp Epidemiol 2003;24:942-5. 4. Maki DG, Kluger DM, Crnich CJ. The risk of bloodstream infection in adults with different intravascular devices: a systemic review of 200 published prospective studies. Mayo Clin Proc 2006;81:1159-71. 5. Merrer J, De Jonghe B, Golliot F, Lefrant JY, Raffy B, Barre E, et al. Complications of femoral and subclavian venous catheterization in critically ill patients: a randomized controlled trial. JAMA 2001;286:700-7. 6. Vonberg RP, Behnke M, Geffers C, Sohr D, Ruden H, Dettenkofer M, et al. Device-associated infection rates for non–intensive care unit patients. Infect Control Hosp Epidemiol 2006;27:357-61. 7. Warren DK, Cosgrove SE, Diekema DJ, Zuccotti G, Climo MW Bolon MK, et al. A multicenter intervention to prevent catheter-associated bloodstream infections. Infect Control Hosp Epidemiol 2006; 27:662-9. 8. Gnass SA, Barbosa L, Bilicich D, Angeloro P, Treiyer W, Grenovero S, et al. Prevention of central venous catheter–related bloodstream infections using non-technologic strategies. Infect Control Hosp Epidemiol 2004;25:675-7. 9. Berenholtz SM, Pronovost PJ, Lipsett PA, Hobson D, Earsing K, Farley JE, et al. Eliminating catheter-related bloodstream infections in the intensive care unit. Crit Care Med 2004;32:2014-20.
Al Raiy et al.
153
10. Lee DH, Jung KY, Choi YH. Use of maximal sterile barrier precautions and/or antimicrobial-coated catheters to reduce the risk of central venous catheter–related bloodstream infection. Infect Control Hosp Epidemiol 2008;29:947-50. 11. Safdar N, Maki DG. Risk of catheter-related bloodstream infection with peripherally inserted central venous catheters used in hospitalized patients. Chest 2005;128:489-95. 12. Horan TC, Gaynes RP. Surveillance of nosocomial infections. In: Mayhall CG, editor. Hospital epidemiology and infection control. 3rd ed Philadelphia: Lippincott Williams & Wilkins; 2004. p. 1659-702. 13. Weber DJ, Sickbert-Bennett EE, Brown V, Rutala WA. Comparison of hospital-wide surveillance and targeted intensive care unit surveillance of healthcare-associated infections. Infect Control Hosp Epidemiol 2007;28:1361-6. 14. Aloul B, Adabag AS, Goussous TY, Weinert CR. Medical resident’s self-reported adherence to guidelines during placement of central venous catheters. Infect Control Hosp Epidemiol 2008;9:86-8. 15. Taylor RW, Palagiri AV. Central venous catheterization. Crit Care Med 2007;35:1390-6. 16. Trick WE, Vernon MO, Welbel SF, Wisniewski MF, Jernigan JA, Weinstein RA. Unnecessary use of central venous catheters: the need to look outside the intensive care unit. Infect Control Hosp Epidemiol 2004;25:266-8. 17. Marschall J, Leone C, Jones M, Nihill D, Fraser VJ, Warren DK. Catheter-associated bloodstream infections in general medical patients outside the intensive care unit: a surveillance study. Infect Control Hosp Epidemiol 2007;28:905-9. 18. Crnich CJ, Maki DG. The promise of novel technology for the prevention of intravascular device–related bloodstream infection, I: pathogenesis and short-term devices. Clin Infect Dis 2002;34:1232-42. 19. Tokars JI, Klevens M, Edwards JR, Horan TC. Measurement of the impact of risk adjustment for central line-days on interpretation of central line–associated bloodstream infection rates. Infect Control Hosp Epidemiol 2007;28:1025-9. 20. Mclaws ML, Berry G. Nonuniform risk of bloodstream infection with increasing central venous catheter-days. Infect Control Hosp Epidemiol 2005;26:715-9. 21. Marschall J, Mermel LA, Classen D, Arias KM, Podgorny K, Anderson DJ, et al. Strategies to prevent central line–associated bloodstream infections in acute care hospitals. Infect Control Hosp Epidemiol 2008;29: S22-30.