| Methods |
Multi‐centre RCT (Greece) |
| Participants |
Adult ICU patients who required a CVC for 3 days or more. Neutropenic patients, pregnant women and patients with allergy to silver or chlorhexidine were excluded |
| Interventions |
Three‐arm comparison: Oligon SPC impregnated CVCs versus CVCs with silver‐gluconate impregnated patch (placed over the skin underneath the CVC insertion site) versus non‐impregnated CVCs |
| Outcomes |
Catheter colonization, CRBSI, ICU death, local adverse effects, sepsis and the number of catheters removed due to suspected sepsis |
| Notes |
Only 2 of the 3 assigned groups (Oligon group and control group that received non‐impregnated CVCs) are included in this review, as the silver‐gluconate impregnated patch does not fit within the interventions prespecified. Sources of funding: not stated |
| Risk of bias |
| Bias |
Authors' judgement |
Support for judgement |
| Random sequence generation (selection bias) |
Low risk |
Quote: “Patients were randomly allocated to one of the three groups, separately for each participating ICU, and according to computer‐generated randomization sequences.” |
| Allocation concealment (selection bias) |
Low risk |
Quote: “This procedure (randomization) was performed by the supervising ICU. The randomization number and the corresponding study group were sealed in envelopes in numeric order. Envelopes were posted to the ICUs with accompanying instructions to be opened by respecting their numerical order” |
| Blinding of participants and personnel (performance bias)
All outcomes |
High risk |
Quote: “Double‐blind design was not possible as a result of apparent differences between the compared products" |
| Blinding of outcome assessment (detection bias)
Microbiological outcomes like catheter colonization |
Unclear risk |
It was not stated clearly whether the assessors of the microbiological outcomes were blinded to the status of the participants |
| Blinding of outcome assessment (detection bias)
Clinical outcomes like CRBSI |
Low risk |
Although it was not stated clearly whether assessors of clinical outcomes were blinded to the status of the participants, the infectious disease physicians who verified the data were blinded
Quote: : “Two ICU infectious diseases experts scrutinized all data blindly to the randomization group to identify concomitant infections and avoid erroneous attribution of the recorded events to the study catheters.” |
| Incomplete outcome data (attrition bias)
All outcomes |
Low risk |
Catheter cultures were available in 376/465 participants randomized (80.9%). The proportions of uncultured catheters (not cultured for a variety of reasons) appeared balanced across the 3 groups. For person‐level outcomes such as CRBSI, all randomized participants were included in the analysis |
| Selective reporting (reporting bias) |
Low risk |
All major outcomes specified in the Methods were reported in sufficient detail in the Results |
| Other bias |
Low risk |
None identified |
