Criteria abstracted from The Users' Guide to Medical Literature, from the Health Information Research Unit and Clinical Epidemiology and Biostatistics, McMaster University

Highlighted lines and questions below provide links to the pertinent description of criteria in The EBM User's Guide, now available at the Canadian Centres for Health Evidence

Article Reviewed:

Please visit the new Evidence Based Journal Club Reviews

A comparison of two antimicrobial-impregnated central venous catheters.

Darouiche RO, Raad II, Heard SO, et al.

N Engl J Med 1999;340:1-8. [abstract]

Reviewed by Stephen Schexnayder MD, University of Arkansas for Medical Sciences/Arkansas Children's Hospital

Review posted November 12, 1999

---

I. What is being studied?:

The study objective:

To compare infection risks with two different antimicrobial coated central venous catheters. Because previous data had shown both rifampin/minocycline coated catheters (RF/MN) and chlorhexidine/silver sulfadiazine (CH/SSD) coated catheters reduced infection rates, these two catheter types were compared (1, 2).

The study design:

Prospective randomized clinical trial in twelve university-affiliated hospitals.

The patients included:

Adult patients who were considered high risk (defined as being in an intensive care unit or immunocompromised) who were expected to need central venous access for at least three days were enrolled after informed consent and no exclusion criteria were met. 865 catheters were placed in 817 patients

The patients excluded:

Pregnant women and patients with a history of allergy to any of the antimicrobial agents.

The interventions compared:

Catheters coated with RF/MN were coated on both the internal and external surfaces; CH/SSD catheters were coated on the external surface only. All catheters were inserted with sterile technique including full barrier precautions. No catheters were changed over guidewires.

The outcomes evaluated:

Catheter colonization (CC) and catheter-related blood stream infection (CR-BSI). CC was detected by culture of four-centimeter segments of the tip and subcutaneous segments using roll-plate method, followed by sonification. CC was defined by either 15 or more colony-forming units in culture of the catheter segments or 1000 or more colony forming units in cultures from sonification of either the tip or subcutaneous segment of the catheter. CR-BSI was defined as a positive peripheral blood culture with the same organism isolated (same species with identical antimicrobial susceptibility) from the CC cultures. Patients with CR-BSI also had to have clinical manifestations of sepsis and no other apparent source of bloodstream infection.

II. Are the results of the study valid?

Primary questions:

1. Was the assignment of patients to treatments randomized?

Yes. A computer-generated randomization scheme was used, where blocks of six catheters were sent to each hospital.

2. Were all patients who entered the trial properly accounted for and attributed at its conclusion?

Was followup complete?

Complete data was available for 85% of catheters. The remaining 127 catheters (58 MN/RF and 69 CH/SSD) were not cultured because of a variety of reasons (84 removed without study coordinator notification, 19 grossly contaminated during removal, and 24 for other reasons). All these were excluded from study analysis. This is a rather high rate of excluded patients. In a worst case analysis, had a significantly skewed number of MN/RF catheters been colonized in these excluded patients, the study's results and conclusions may have been altered.

Although it is understandable why many of these catheters were excluded from analysis, it would have been valuable to continue to include the patients in an intention-to-treat analysis. Although the strict diagnosis of CR-BSI could not be confirmed without an appropriate culture from the catheter, it would have been of value to report on the number of bloodstream infections overall as well.

Were patients analyzed in the groups to which they were randomized?

Yes. No patients crossed over between groups.

Secondary questions:

3. Were patients, health workers, and study personnel "blind" to treatment?

Yes, according to the paper. However, both catheters are commercially available and are distinctly different colors, and this issue is not addressed in this paper. Additionally, each catheter has a unique junction between the catheter and the tubing, so operators and investigators could learn which type of catheter was being used.

4. Were the groups similar at the start of the trial?

Yes. Patients' underlying disease as well as age and sexes were similar. Approximately two-thirds of patients in both groups were in intensive care units. Risk factors for infections that were considered (total parenteral nutrition, immunosuppressive therapy, bone marrow transplantation, neutropenia, mechanical ventilation, other intravascular catheters, and urinary catheters) were similar between groups. Approximated 90% of both groups received systemic antibiotics. Insertion sites were also similar (approximately one-half were subclavian catheters, followed by approximately 40% being placed in the internal jugular vein, and 10% in the femoral vein.

5. Aside from the experimental intervention, were the groups treated equally?

Yes. Skin preparation was done using 10% providone-iodine and operators used maximal barrier precautions. Dressings were changed and the insertion sites inspected three times a week. The percentage of patients who received systemic antibiotics was not different between groups, although the specific antibiotics were not mentioned. It is possible one group received antibiotics more active against typical CVL pathogens (such as vancomycin) than the other group.

III. What were the results?

1. How large was the treatment effect?

The relative risk for catheter colonization as detected by at least one method of assessment was 2.9 times higher for the CH/SSD than the RF/MN catheters (22.8% vs. 7.9%, 95% confidence interval for relative risk, 1.94-4.33). The absolute risk reduction was 14.9%. This was independent of whether catheters were in place for less than or more than seven days. This effect was seen in all hospitals that contributed over 32 catheters to the study. This effect was noted with coagulase negative staphylococci (relative risk 4.16, 95% confidence interval 2.42-7.14), gram-positive bacilli (relative risk 7.46, 95% confidence interval 0.94-58.8), or gram-negative bacilli (relative risk, 3.96, 95% confidence interval 1.35-11.63). Rates of colonization of catheters with coagulase positive staphylococci, enterococci, and yeast were not statistically different between the two groups.

For CR-BSI, 14 infections were found; 13 in the CH/SD catheter group (3.4%) and 1 in the RF/MN catheter group (0.3%), giving the CH/SSD catheter group a relative risk of 12.05, (95% confidence interval 1.59 - 90.9). Two of the patients in the CH/SSD group died as a result of bloodstream infections. Rates of CR-BSI per 1000 catheter days (a standard measure for examining CVL-related infections) were 0.3 in the RF/MN group (95% confidence interval 0.01 1.85) and 4.1 in the CH/SSD group (95% confidence interval 2.22-6.99).

By use of a multivariate logistic-regression analysis, the following factors were found to be significantly associated with catheter colonization:

Factor	Odds Ratio	95% confidence interval
Femoral or jugular vein placement (vs. subclavian)	3.05	1.86 - 5.01
CH/SSD catheter (vs. RF/MN)	2.80	1.68 - 4.66
ICU setting	2.60	1.47 - 4.62
Male sex	2.45	1.43 — 4.20
Mechanical ventilation	1.97	1.14 — 3.41

 

When catheter-related blood stream infection was considered, multivariate analysis revealed three factors associated with increased infections as shown below:

Factor	Odds Ratio	95% confidence interval
Duration of catheterization > 7 days	8.66	1.90 — 39.40
CH/SSD catheter (vs. RF/MN)	11.04	1.42 — 71.2
Male sex	9.15	1.18 — 71.2

2. How precise was the estimate of the treatment effect?

The 95% confidence intervals reported for catheter colonization show that infection is 1.94 to 4.33 times more likely with the CH/SSD catheters. While the effects was maintained across several types of bacterial (coagulase negative staphylococci, gram positive bacilli, and gram negative bacilli), the confidence intervals for these individual organisms are very wide, particularly for gram positive bacilli.

CH/SSD catheters were 12 times more likely to be associated with catheterrelated blood stream infection than the RF/MN catheters, but the 95% confidence intervals are extremely wide giving a risk from slightly under two times more likely to over 90 times more likely.

IV. Will the results help me in caring for my patients?

1. Can the results be applied to my patient care?

Possibly. While this study was performed in adults, the most frequent cause of CR-BSI (coagulase negative staphylococci) is the same in children as found in this study.

2. Were all clinically important outcomes considered?

Yes. While length of stay (LOS) is not addressed, increased LOS has been previously shown to be strongly correlated with CR-BSI.

3. Are the likely treatment benefits worth the potential harms and costs?

It would appear that the benefits outweigh the cost concerns. These catheters cost approximately $15 per catheter more. Cost data for pediatric CR-BSI infection is unknown, but it is well accepted that it increases length of stay and hospital costs (3). The incremental cost for the catheter is less than a single blood culture. Based on adult data, the incremental cost is minimal, and might result in savings to the hospital. The catheters might actually save lives in adults, where the incidence of death from catheter-related infection is substantial. An accompanying editorial shows, based on a crude cost analysis (with higher incremental costs for the adult catheters than for pediatric catheters), the use of this technology would be more cost effective than air bags in vehicles in terms of quality adjusted years of life saved (4).

Based on the number needed to treat (NNT), 7 patients would have to be treated with a RF/MN catheter to prevent one case of colonization, or 32 patients would have to be treated to prevent one CR-BSI.

References

1. Raad D, Darouiche ER, Dupuis J. Central venous catheters coated with minocycline and rifampin for the prevention of catheter-related colonization and bloodstream infections. Ann Intern Med 1997;127:267-274. [abstract] [PedsCCM EB Journal Club Review]
2. Maki DG, Stolz SM, Wheeler S, et al. Prevention of Central Venous Catheter-Related Bloodstream Infection by Use of an Antiseptic-Impregnated Catheter. Ann Intern Med 1997;127:257-266. [abstract] [PedsCCM EB Journal Club Review]
3. Pittet D, Tarara D, Wenzel RP. Nosocomial bloodstream infection in critically ill patients. JAMA 1994:271:1598-1601. [abstract]
4. Wenzel RP, Edmond MB. The evolving technology of venous access. N Engl J Med 1999:340:48-50. [abstract]

Comments

Back to the EB Journal Club Index