Correspondence
Replacement of Central Vascular Catheters
N Engl J Med 1993; 328:445-446February 11, 1993
- Article
To the Editor:
Cobb et al. (Oct. 8 issue)1 reported on a trial of scheduled replacement of central venous catheters that compared catheter replacement after three days at a new puncture site with replacement using guide-wire-assisted exchanges. The authors found an increased incidence of catheter colonization and catheter-related bloodstream infection among the patients who underwent guide-wire-assisted catheter exchanges. In addition, they found more catheter-related mechanical complications, such as pneumothorax, among the patients who had catheter replacement at a new puncture site.
Unfortunately, the patients of Cobb et al. underwent catheter insertion principally through the subclavian and internal jugular veins. Central venous cannulation can be accomplished safely and effectively through the external jugular, percutaneous brachial, and brachial-cutdown routes in a substantial number of patients. These routes do not have the same mechanical complications as puncture of the subclavian and internal jugular veins. Their use as part of the planned strategy in a study of this type might produce a different outcome.
In addition, 90 of 250 patients (36 percent) were excluded from the study. Among the reasons for exclusion were anticoagulation therapy and danger from repeated punctures due to the risk of pneumothorax. Patients with these contraindications are particularly well suited to central-line insertion through the external jugular or brachial routes. They may also represent some of the patients at highest risk, in whom long-term central venous access is necessary. Their exclusion from the study was a serious flaw.
. . . The authors' conclusions that changing catheters with the use of a guide wire increases the risk of bloodstream infection and that replacement involving the insertion of a catheter by the subclavian and internal jugular routes is associated with mechanical complications are well justified. A broader strategy for the selection of new catheter-insertion sites and the consideration of other factors, such as the use of long-term antibiotic therapy, may help in developing a useful approach to the long-term maintenance of a central venous catheter.
1 ReferencesTed Feldman, M.D.
University of Chicago, Chicago, IL 606371
Cobb DK ,High KP ,Sawyer RD , et al. A controlled trial of scheduled replacement of central venous and pulmonary-artery catheters. N Engl J Med 1992;327:1062-1068
Full Text | Web of Science | Medline
To the Editor:
The prospective, randomized trial reported by Cobb et al. had important findings: rates of mechanical complications from repetitive venipuncture are prohibitively high; changing catheters over a guide wire is associated with a high incidence of positive cultures; and routine catheter changes appear to involve a higher incidence of positive cultures than changes only when clinically indicated.
Our small, randomized trial1 of a homogeneous group of patients receiving total parenteral nutrition with triple-lumen catheters showed no difference between groups in the incidence of positive cultures, but catheters replaced when indicated were more likely to be associated with evidence of clinically important infection. Why do these two studies have such different conclusions?
In Cobb's study, since the routine-change groups had cultures performed every three days, the risk of having a positive culture was obviously higher. Were these positive cultures associated with clinical syndromes, or were they unexpected results? Of the patients who were excluded after randomization, how many had been randomly assigned to one of the routine-change groups? Most important, how do the authors explain the very high incidence of positive blood cultures in the patients who had a fresh venipuncture every three days? The high rate of mechanical complications is understandable, but surgical wisdom would suggest that this group should have the lowest rate of catheter sepsis. Why was there such a high incidence of infection during the first three days of catheter use?
In terms of outcome, how many patients had a serious syndrome of clinical sepsis? Was there a difference in the duration of stay in the intensive care unit? Was there a difference in the total hospital stay? Gertrude Stein said that for there to be a difference, something had to make a difference. Was there a real difference? This is an extremely important question, and hundreds of intensive care units will modify their behavior on the basis of these data. Will the authors help us understand how to apply their results?
1 ReferencesPeter J. Fabri, M.D.
University of South Florida, Tampa, FL 336121
Powell C ,Kudsk KA ,Kulich PA ,Mandelbaum JA ,Fabri PJ . Effect of frequent guidewire changes on triple-lumen catheter sepsis. JPEN J Parenter Enteral Nutr 1988;12:462-464
CrossRef | Web of Science | Medline
To the Editor:
The study by Cobb et al. would have been more statistically correct and clinically relevant had it addressed the number of infections per patient rather than per catheter used, because the patients in the routine-exchange group had multiple catheters inserted. Another flaw in the article was the elimination of a large number of patients after randomization (32 patients), and in unequal proportions among the study groups. . . .
. . . With regard to the infection rate in the guide-wire-exchange group, the difference from the other groups approaches statistical significance with a P value of 0.06, although the numbers are small. This suggests a higher infection rate with guide-wire use. The most useful data from the study indicate a significantly higher (P = 0.005) rate of mechanical complications associated with the use of new puncture sites; for this group, per-catheter rates are more appropriate.
Given its small size, this study does not have adequate power to yield conclusions about the efficacy of routine catheter exchange. Although they are of borderline statistical significance, the data suggest that infection is increased in exchanges using a guide wire, and they are more conclusive in indicating that the risk of mechanical complications increases with new insertions.
J. Douglas Rizzo, M.D.
Brian L. Strom, M.D., M.P.H.
Jesse Berlin, Sc.D.
University of Pennsylvania School of Medicine, Philadelphia, PA 19104Author/Editor Response
The authors reply:
To the Editor: We agree with Dr. Feldman that other strategies, such as the use of peripherally inserted central venous catheters, should be studied in patients in intensive care.
The trial cited by Dr. Fabri1 showed no significant effect of scheduled replacement on catheter sepsis (8 episodes in 63 patients [13 percent], vs. 10 in 63 patients whose catheters were replaced only when necessary [16 percent]; P = 0.80). Our definition required positive blood and catheter cultures as evidence of catheter sepsis2. If Fabri and his coworkers had used this definition, their rates would still not have been significantly different (5 of 63 [8 percent] vs. 4 of 63 [6 percent], P = 1.00), and there would have been no trend favoring scheduled replacement. We diagnosed bloodstream infection only in patients with clinical sepsis (our indication for blood cultures); we did not consider the rate in patients with scheduled puncture of a new site to be high (0.9 percent per catheter, or 2.8 percent per patient). The higher infection rates in Fabri's trial could relate to other differences (e.g., the use of povidone-iodine ointment and cotton-gauze dressings in our trial vs. transparent dressings and no ointment). With regard to the question about a “real difference,” mechanical complications occurred more frequently with punctures at new sites; infectious complications were more frequent with exchanges using a guide wire. Scheduled replacement did not make a difference.
As for the comment by Dr. Rizzo and colleagues about the need to address the number of infections per patient, our 3 does show rates per patient, and the discussion addresses this risk. Patients were excluded after randomization as follows: 8 (4 percent) because there were no outcome data, 9 (5 percent) because of failures of aseptic technique (which could have distorted results), and 15 (8 percent) because the criteria for randomization were not met. We believe that excluding such small proportions for these reasons does not compromise the quality of our results.
As for the question of statistical power to draw conclusions about scheduled replacement, pooling our data with those cited by Fabri and those from the trial by Eyer et al.3 yields rates of catheter colonization and bloodstream infection that show no trend favoring scheduled replacement: 27 percent of 211 patients so treated had colonized catheters (as compared with 14 percent of 187 patients with unscheduled catheter changes), and 11 percent had catheter-related bloodstream infections (as compared with 10 percent with unscheduled changes). The power to detect an important reduction (e.g., a relative risk of 0.20) in bloodstream infection (i.e., one comparable to that seen with a silver-impregnated collagen cuff [relative risk, 0.14]4 or an antiseptic-impregnated catheter [relative risk, 0.20]5) from pooling these three trials was 91 percent (alpha = 0.05, two-sided). The power to detect an important reduction in colonization, comparable to that observed with the silver-impregnated cuff (relative risk, 0.27),4,6 was 93 percent (alpha = 0.05, two-sided).
6 ReferencesDavid K. Cobb, M.D.
Tulsa Infectious Diseases Clinic, Tulsa, OK 74104Kevin P. High, M.D.
Yale University, New Haven, CT 06514Barry M. Farr, M.D., M.Sc.
University of Virginia, Charlottesville, VA 229081
Powell C ,Kudsk KA ,Kulich PA ,Mandelbaum JA ,Fabri PJ . Effect of frequent guidewire changes on triple-lumen catheter sepsis. JPEN J Parenter Enteral Nutr 1988;12:462-464
CrossRef | Web of Science | Medline2
Maki DG ,Weise CE ,Sarafin HW . A semiquantitative culture method for identifying intravenous-catheter-related infection. N Engl J Med 1977;296:1305-1309
Full Text | Web of Science | Medline3
Eyer S ,Brummitt C ,Crossley K ,Siegel R ,Cerra F . Catheter-related sepsis: prospective, randomized study of three methods of long-term catheter maintenance. Crit Care Med 1990;18:1073-1079
CrossRef | Web of Science | Medline4
Flowers RH III ,Schwenzer KJ ,Kopel RF ,Fisch MJ ,Tucker SI ,Farr BM . Efficacy of an attachable subcutaneous cuff for the prevention of intravascular catheter-related infection: a randomized, controlled trial. JAMA 1989;261:878-883
CrossRef | Web of Science | Medline5
Maki DG, Wheeler SJ, Stolz SM, Mermel LA. Clinical trial of a novel antiseptic central venous catheter. In: Program and abstracts of the 31st Interscience Conference on Antimicrobial Agents and Chemotherapy, Chicago, September 29-October 2, 1991. Washington, D.C.: American Society for Microbiology, 1991:176. abstract.
6
Maki DG ,Cobb L ,Garman JK ,Shapiro JM ,Ringer M ,Helgerson RB . An attachable silver-impregnated cuff for prevention of infection with central venous catheters: a prospective randomized multicenter trial. Am J Med 1988;85:307-314
CrossRef | Web of Science | Medline
- Citing Articles (1)
Citing Articles
1
Guy Bruninx, Jean-Christophe Matte, Françoise VanWilder, Christian Delcour. (1996) Catheter migration of a port-a-cath system. Cardiovascular and Interventional Radiology 19:6, 435-437
CrossRef
