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

Heparin-bonded central venous lines reduce thrombotic and infective complications in critically ill children.

Pierce CM, Wade A, Mok Q.

Intensive Care Medicine 2000; 26:967-972. [abstract]

Reviewed by Joel Barron, MD and Andrianik Madikians, MD, UCLA, Los Angeles, CA

Review posted December 6, 2000

---

I. What is being studied?:

The study objective:

To determine whether heparin-bonded central venous catheters reduce the incidence of catheter-related thrombosis and infection in critically ill children.

The study design:

Prospective double-blind randomized controlled study.

The patients included:

Two hundred and nine patients, 123 males and 86 females, aged 0-16 years, admitted to the intensive care unit, and needing a central venous line.

The patients excluded:

Patients were excluded who had had a central venous catheter inserted within the preceding month, those on thrombolytic therapy, those with a history of a previous thrombosis and post-operative cardiac patients.

The interventions compared:

Insertion of heparin-bonded (HB) versus standard non-heparin-bonded (NHB) polyurethane central venous catheter in intensive care patients. Catheters were appropriately sized for the child and selected from size 4 or 5 F diameter, double or triple lumen, and 5, 8 and 15 cm in length.

The outcomes evaluated:

• Incidence of vascular thrombosis of catheterized vessel: thrombus was defined as echogenic mass within vessel lumen disrupting flow, as determined by color Doppler ultrasound, performed within 3 days of insertion and then every three days thereafter.
• Blood stream infection: blood cultures were considered significant for bacteremia only if the same organism was grown from two sets of cultures, from two different sites, and if the organism differed from that cultured on admission.
• Catheter-related blood stream infection: defined as catheter colonization (at least 15 cfu cultured from the catheter tip) with the same organism that was cultured from two sets of blood cultures from two separate sites.

II. Are the results of the study valid?

Primary questions:

1. Was the assignment of patients to treatments randomized?

Consecutive admissions to the intensive care unit were randomized, by selection of consecutively numbered opaque envelopes, to receive either a HB or a NHB line as the need for central venous access arose.

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

Was followup complete?

All 209 patients were accounted for at the end of the study. At the study's completion data were available from 97 patients with HB and 103 patients with NHB lines. Nine patients were excluded, four from NHB group and five from HB group. In one patient, the line was changed within the first 24 hours for a catheter suitable for hemofiltration. In eight patients, the line was removed before sufficient microbiological data could be collected. Complete data was collected for all patients who entered the study except those who were excluded.

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

It seems that patients were analyzed within their assigned groups with regard to the outcomes measured by the study.

Secondary questions:

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

Radiologists, microbiologists, data collection personnel were blinded to treatments groups. It is unclear whether the operators inserting the catheters were also blinded, although they mention that the catheters were identical in appearance; presumably this means the operators were also unaware of group assignment.

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

The groups were similar, but not identical at the start of the trial. Table 1 compares the demographic attributes between the two groups; there is no clear difference in male/female ratio, age, site of line placement, number of days the catheters were in place, use of TPN, etc.

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

With regard to use of the catheters, treatment was similar (outlined in Table 1). There is no mention in the text whether therapeutic measures were consistent among patients with the same disease entity. However, given what appears to be complete blinding, and the similarities of the groups, it seems unlikely that there were any systematic differences in other aspects of patient care between groups.

III. What were the results?

1. How large was the treatment effect?

See Table 2. The incidence of positive blood cultures alone was 3/97 (3%) in the HB lines and 15/103 (15%) in the NHB lines. The incidence of positive CVL-associated bacteremia was 1/97 (1%) in the HB lines and 19/103 (18%) in the NHB lines. The total incidence of infection (both bacteremia alone and CVL associated bacteremia) which was treated was 4/97 (4%) in HB lines and 34/103 (33%) in the NHB lines.

The relative risk reduction in catheter related infection in HB vs. NHB lines was 94%, with a 95% CI of 60% to 99%.

There were 5.17 infections in the HB compared to 46.96 in the NHB lines per 1000 catheter days (p < 0.0005). The number needed to treat to avoid one episode of infection was estimated to be 4.

There was no thrombosis in the 97 HB lines and 8 (8%) episodes of thrombosis in the 103 NHB lines.

Survival Curve (Figure 2): Plots proportion of lines without infection as a function of time in HB vs. NHB catheters. There is a dramatic difference between the curves, with an approximate 40% difference by 25 days of infection-free catheters between the groups.

Univariate analysis of hazard ratios for time to infection revealed a hazard ratio of 0.11 for heparin bonding.

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

Comparison of the total infection rate between the two groups yielded p < 0.00005, and a relative risk reduction of 88% (95% CI 67%, 95%). The number need to treat to avoid one episode of infection was 4 (95% CI 3, 5). A very small NNT with a narrow confidence interval illustrates the significance of the treatment effect. The 95% CI for the hazard ratio of heparin bonding was 0.04 to 0.31. This suggests that at worst, the risk of infection with heparin bonding is one third that of NHB lines.

The reduction in absolute risk of thrombosis for HB was 8% (95% CI 2.6%, 13%), with p < 0.007 and NNT = 13 (95% CI 7, 38). Statistical analysis of the study results affirm their precision and support the studyÕs conclusions of positive therapeutic effect.

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

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

The results of this study are directly applicable to my patient care. Heparin-bonded CVL's are demonstrated with statistical significance to reduce the risk of thrombosis and CVL-associated infection in the population studied.

2. Were all clinically important outcomes considered?

The three outcomes measured were the most clinically relevant. Another that was documented was the incidence of thrombocytopenia in patients with HB catheters. This was not seen in the study, although the incidence of heparin-induced thrombocytopenia is low enough that this study might not have been large enough to identify cases. They did not comment on mortality.

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

The authors address this question. The additional cost of each HB catheter is 2.50 - 3.50 pounds ($3.50 - $5). Based on estimates of NNT to prevent one episode of infection would be 7 - 17 pounds ($10 - $24), and to prevent one episode of thrombosis 15 - 115 pounds ($21.50 - $165), "both of which figures are significantly less than the cost of a course of antibiotics or thrombolytic therapy." (Pierce et al, 971) The findings of this study clearly demonstrate dramatic improvement in patient care at a reasonable cost.

Additional comments

There is no mention of who put in the lines, nor was mention made regarding the technique used to place, secure and care for the lines. In addition, the degree of difficulty of line insertion was not addressed, e.g. was access difficult because of a junior operator inserting the line, or history of multiple previous lines, anatomic variation, etc.?

References

1. Appelgren P, Ransjo U, Bindslev L, Espersen F, Larm O. Surface heparinization of central venous catheters reduces microbial colonization in vitro and in vivo: results from a prospective, randomized trial. Crit Care Med. 1996;24(9):1482-9. [abstract]
2. Randolph AG, Cook DJ, Gonzales CA, Andrew M. Benefit of heparin in central venous and pulmonary artery catheters: a meta-analysis of randomized controlled trials. Chest. 1998;113(1):165-71. [abstract]
3. Harris E, Albert A. Survivorship Analysis for Clinical Studies. New York: Marcel Dekker, 1991.

Comments

Back to the EB Journal Club Index