There is actually very little quality evidence comparing insertion sites for central venous access. There was a Cochrane review in 2007, but that included only 1 RCT (which always makes me wonder whether there was any point in going to all that effort). This was updated in 2012, with the addition of two later studies (here and here), and also a non-blinded study with high bias risk published (in Chinese) in the Chinese Journal of Practical Nursing, (which is not listed on Pubmed). Whilst published in time for inclusion, the latter article is not mentioned in the 2007 Cochrane review!
A Cochrane (or any other) review is only as good as the evidence included, and the authors write “….there appeared to be a considerable risk of bias in the included studies and we have to say that there was still a lack of systematic and strong evidence on optimal insertion sites for clinical practice”.
There are at least two other reviews on the topic. The first was by Reusch, published pre-Cochrane in 2002. This did not include any randomised trials. The second was published in 2012 by Marik looking only at infection; this included 2 RCTs and 8 cohort studies.
Combining these reviews, what do we ‘know’? (RRs given as 95% CIs)
Regarding thrombosis risk:
- Femoral is greater risk than subclavian (based on 1xRCT, RR 2.8-47.52)
- There’s no difference between femoral and jugular
- There’s no difference between femoral and subclavian and jugular combined (which doesn’t make sense if the previous statements are correct)
- Firstly it depends whether we’re talking colonisation or bloodstream infection.
- Femoral is colonised more than subclavian (based on 1xRCT, RR 1.95-21.21)
- There’s no difference between femoral and jugular colonisation.
- Jugular is no more likely to lead to bloodstream infection than subclavian (based on 3 cohort, RR 0.62-8.09)
- Considering only RCTs, or all studies in the Marik review minus the 2 outliers, there’s no difference in infection between femoral, subclavian or jugular.
- Femoral related infections were more likely in earlier studies (Marik)
So that’s clear then?!
I can’t get too excited about the site of a central line, but maybe I should (for those that do there’s meeting for you here). The ‘Matching Michigan’ campaign recommended avoidance of the femoral route, but the evidence base for that seems less than watertight. We know that the impact of nosocomial infection can be devastating; perhaps rather than endlessly continuing to try and find a place for steroids in sepsis or the optimal filtration dose to the nearest ml. it might be worth trying to work out if insertion site does make a difference? Or maybe that’s ambitious but hopeless?!
As always, please leave a comment below and we’ll see if we can get some discussion going….