Peripheral cannula: using that old line…

1 December 2012

CLINICALLY APPRAISED TOPIC (CAT): How do you decide when you need to replace a peripheral IV line?


Replacing peripheral cannula when clinically indicated is as safe as a routine replacement strategy. Introducing clinically indicated replacement may release healthcare resources to be used for other purposes.


When updating your policy on changing peripheral intravenous (IV) cannula, you checked the Cochrane review on routine versus clinically indicated replacement. Although the review found no significant benefit in routine changes, there were slightly greater rates of phlebitis in the clinically indicated groups. Using the precautionary principle you are inclined towards maintaining the policy of routine changes every 72–96 hours. However, you note the review only included evidence to 2009 and thus you decide to check for any new trials.


Does routine versus clinically indicated replacement reduce complications in adults with peripheral intravenous cannulae?


Pubmed clinical queries– (therapy, broad): Peripheral AND cannula AND replacement.


Rickard CM, Webster J, Wallis MC, et al. Routine versus clinically indicated replacement of peripheral intravenous catheters: a randomised controlled equivalence trial. Lancet 2012; 380:1066-74.


Two-arm, parallel group, randomised controlled equivalence trial conducted in three hospitals in Queensland, Australia between May 2008 and September 2009. 3379 patients were assessed for eligibility and 3283 were randomised. Participants were adults aged 18 years or more, with a peripheral IV cannula in situ that had been inserted in any clinical area by any clinical team member, and expected treatment of four or more days. Patients were ineligible if they had a bloodstream infection, planned removal of the cannula within 24 hours, an intravenous catheter that had been in situ for longer than 72 hours, or a cannula that had been inserted in an emergency (hospital policies stipulated such cannula should be replaced within 24 hours). Skin disinfection was with 2 per cent chlorhexidine in 70 per cent ethanol, cannula were standardised (Insyte Autogard 30 mm) and covered with Tegaderm 1624W dressings. Dressings were replaced weekly, or when soiled or loose and there was no restriction of type of infusion.

Clinically indicated cannula changes (n=1593)

Participants in the clinically indicated group had cannula removed only with completion of therapy, phlebitis, infiltration, occlusion, accidental removal or suspected infection.

Routine cannula changes (n=1690)

Participants in routine change group had cannula replaced every third day unless clinical reasons required replacement e.g. cannula failed before the third day. The day three change was to occur at about 72 hours from insertion.


The primary outcome was phlebitis based on two or more of the following being reported: patient-reported pain or tenderness with severity of two or more out of 10, erythema extending at least 1 cm from insertion site, swelling extending at least 1 cm from insertion site, purulent discharge, or palpable venous chord. Phlebitis measures were done daily and 48 hours after removal. The equivalence was an absolute difference (AD) of +/- 3 per cent in phlebitis rates. Secondary outcomes included catheter-related bloodstream infection, all-cause bloodstream infections, local venous infection, colonisation of intravenous catheter tip, infusion failure, number of cannula needed per patient, overall duration of intravenous therapy, costs per patient for course of intravenous therapy, mortality with intravenous catheter in situ or within 48 hours of removal.


Simple randomisation on 1:1 ratio stratified by hospital using computer-generated numbers with randomisation occuring at point of study entry using hand-held computer. There appears to be 100 per cent follow-up and primary analysis was on an intention-to-treat basis, with all participants being included in the analysis. A per protocol analysis was also conducted. Patients, clinical and research staff could not be blinded, but laboratory staff were blinded, as was an independent rater for catheter-related bloodstream infections. The two groups did not differ on where first cannula inserted, cannula gauge, who inserted cannula, insertion site, or prescribed treatments. No difference in treatment between the two groups were likely, apart from the intervention, and the mean duration of therapy was similar in both groups (98 versus 99 hours). Overall the study quality was high.


The patients did not differ between groups by age, sex, type of admission, comorbidities, presence of wound infection, wound drain or stoma. Protocol adherence was 85 per cent in the clinically indicated group and 70 per cent in the routine changes group. Median dwell time was 84 hours (interquartile range 64–118) in the clinically-indicated group and 70 hours (interquartile range 57–77) in the routine group. There was no significant difference between the groups (see table) for the primary outcome in either the intention-to-treat analysis or the per-protocol analysis. The range of dwell times were 48–561 hours in the clinically indicated group and 48–96 hours in the routine group for the per protocol analysis. There were no significant differences in any of the secondary outcomes, except on cost where the significant difference in number of cannulae (1.7 vs 1.9 per patient) drove a significant difference in costs ($61.66 versus $69.24).


Large equivalence trial with 95 per cent power to observe a +/-3 per cent equivalence margin at a phlebitis rate of 4 per cent.

Pragmatic trial sufficiently similar to New Zealand clinical practice environment.

Reviewer: Dr Andrew Jull, Associate Professor, University of Auckland and Nurse Advisor, Quality, Auckland District Health Board.