JC: Should we glue cannulas to patients? St.Emlyn’s

Use of glue cyanoacrylate to secure peripheral intravenous cannula. St.Emlyn’s. randomised controlled trial critical appraisal.

A sticky solution for peripheral IVs

Inserting a peripheral intravenous catheter (cannulas) into a patient seems part and parcel of their journey into hospital (Ed though perhaps it should not be). This intervention is most commonly started in our emergency departments. Yet just last year a Cochrane review was published showing that there is still no clear evidence to guide us in how best to keep them in place1. Up to 69% of cannulas fail because of ineffective methods of securing them2.

Hang on, are you telling me a cannula fails because of the way we stick it down?

Well, yes.

If a cannula is not secured well, it might make tiny movements. These movements could lead to the cannula becoming dislodged, irritation of the walls of the vein causing phlebitis, or the entry of micro-organisms, causing infection.

Is there a better way?

A team in Caboolture, Queensland, Australia looked at just this, having read a study which seemed to show promising results for the use of skin glue. To be precise we are talking about the use of Cyanoacrylate, medical grade tissue adhesive (NOT superglue). They went on to conduct their own randomised control trial. The abstract is inserted below (though as always you should read the whole paper):

Screenshot 2016-02-09 09.29.58

What did they do?

A non-blinded, randomised controlled trial was conducted over 4 months. Patients who required admission into hospital had a new cannula inserted into their upper limb and clinical trial software was used to randomise for the fixation method. In the standard care group the cannula was secured using Tegaderm IV Transparent Film Dressing.

Those patients randomised to tissue adhesive fixation had a drop of cyanoacrylate at the insertion site as well under the hub of the cannula. The glue was allowed to dry for 30 seconds after which a Tegaderm IV Transparent Film Dressing was placed on top. Both groups had cannula patency checked with a 10ml flush of NaCl 0.9%.

Did you say non-blinded?

Unfortunately tissue adhesive has a subtle colour and appearances. This means at intervention and follow up blinding was not possible.

So was anyone excluded?

Yes, there were exclusion criteria. If you had an allergy to tissue adhesive or were under 18 years you were not included in the trial. Anyone with upper limb phlebitis, venous thrombus, or infection near cannula insertion site was excluded. Lastly, the trial did not include patients who were very likely to have their cannula removed in cases such as agitation.

What were the outcomes and how were they assessed?

The primary outcome was cannula failure at 48 hours post insertion, with “failure” defined as any or a combination of: phlebitis, infection, occlusion or dislodgement.

If the patient was still in hospital they were reviewed, however if they had been discharged they had a telephone consultation.

The secondary outcome was the individual mode of cannula failure. This was ascertained by a review in person, documentation in the patient’s notes and a patient questionnaire. They did not find a statistically significant difference in infection, phlebitis or occlusion although the event rates for these complications was low and therefore this could be a type 2 error.

Did they get much data?

In total 369 cases were analysed; 179 in the tissue adhesive group and 190 in the control group. A 20 gauge cannula was the most commonly inserted in each group (128 in standard care, 120 in skin glue), with the favoured insertion site being the antecubital fossa. That’s a reasonable number of patients and the study was powered to detect a difference of 7% which we would agree is a clinically important difference.

Does it work?

This study found that 17% of cannulas failed in the skin glue group compared to 27% in standard care. The most significant reduction came in cannulas failing due to dislodgement (13 vs 26). For the stats nerds amongst us that’s a number needed to treat of 10 – a really powerful intervention (if true).

This looks great: are there any flaws?

In this trial the research nurses are performing a non-blinded intervention, as well as following up the outcome of that intervention. Some of the outcomes of failure included ‘clinical impression of cellulitis’, which is a very subjective outcome. This has potential to introduce observer bias into the study; if you are looking for cannulas without tissue adhesive to fail, will the researcher need only ‘soft’ signs of cellulitis to declare it failed, in comparison to the ones secured with tissue adhes

Only 179 of these cases were assessed by face-to-face follow up, leaving 209 followed up by telephone consultation. How good is visual assessment – do you need to see it to believe it, or does a standardised questionnaire give the same results?   This leaves the study open to recall bias; if the patient is being asked about their symptoms but they haven’t been blinded to the intervention, are they more likely to report symptoms in the control group?

Costs are interesting. If considering this you should find out how much glue costs in your institution and consider whether the decreased failure rate may offset the additional cost.

Lastly, there is the issue of removal. It appears that they used glue removal wipes, but there was still some pulling experienced by patients. This needs further enquiry as patient experience was not really looked at in this study.

Bottom line

Use of tissue adhesive in this trial does appear to reduce the incidence of cannula failure rate at 48 hours. This might be worth considering for patients likely to require admission of more than 48 hours’ duration, in order to reduce the anxiety of repeated cannula insertions for the patient.

The key question is whether it will change your practice? We’re not sure on the basis of this single study, but we’d love to see some more research in this area. On the basis of this and other studies looking at glue placement of CVC lines it is certainly worthy of consideration.






Don’t forget to listen to the SGEM podcast on this paper too. Simon and Ken talk through the paper in depth, and with modicum of nerdiness.

The SGEM – Stuck on you.


FRCEM style questions.

  1. Why is it important to have primary and secondary outcomes?
  2. In terms of number needed to treat (NNT) what is the NNT for failure of IV cannula when comparing glue with standard in this study?

Further reading

  1. http://www.cochrane.org/CD011070/WOUNDS_effectiveness-of-dressings-and-other-devices-that-are-used-to-keep-a-peripheral-venous-catheter-in-place
  2. Skin Glue Reduces the Failure Rate of Emergency Department-Inserted Peripheral Intravenous Catheters: A Randomized Controlled Trial. http://www.ncbi.nlm.nih.gov/pubmed/26747220
  3.  Superglue for CVCs http://resus.me/superglue-for-cvcs/
  4. Tissue adhesive as an alternative to sutures for securing central venous catheters
    Anaesthesia. 2007 Sep;62(9):969-70
  5. Cyanoacrylate tissue adhesives – effective securement technique for intravascular catheters: in vitro testing of safety and feasibility
    Anaesth Intensive Care. 2012 May;40(3):460-6
  6. Is that IV really needed? http://www.stemlynsblog.org/is-that-iv-really-needed/

Before you go please don’t forget to…

Posted by laura howard

Laura Howard. MBChB, MRCEM is an Editorial Board Member on the St Emlyn’s blog and podcast. She is an emergency physician trainee and clinical and doctoral (PhD) fellow in Emergency Medicine Manchester Metropolitan University. She co-founded the ED Spa Project (@edspa_mcr). She was the RCEM Young Investigator of the Year 2016. Her reseach interests include emergency medicine, wellbeing, compassionate care, triage. She is regularly invited to speak on the topic of well-being. Her PhD is on triage (funded by the Manchester Triage Group) and she is a member of the Wellbeing Committee at the Royal College of Emergency Medicine. You can find her on twitter as @laurahoward10

  1. That failure rate is really surprising, certainly in A&E it’s rare for a well secured cannula to fail


    1. It’s failure up to 48 hours


  2. Great post Laura. Interesting stuff may well be worth a pop(!) especially in those where high risk of failure. Could see it being very useful in the sweaty agitated although excluded here – dry skin blob of glue before tegaderm.


  3. Great post Laura, think it’s particularly worth considering in those with difficult IV access. Always especially frustrating when their carefully placed cannula becomes dislodged!


  4. […] peripheral IVs be secured with cyanoacrylate? St. Emlyn’s dissects a recent article on the topic. […]


  5. Great post! I was working in Caboolture Hospital as this trial was ongoing as a ward resident, and if anecdotal evidence is anything to go by – this study proved a game changer for me. I now routinely put in the first IV with “Cavilon” (the name of the adhesive glue used in the trial), and all of my precious ultrasound guided IVC’s, PICCS etc get anchored with Cavilon as well now! As a lowly resident, I wasn’t worried about infection rates (now am!), long term failure rates (now am!) or any of the major outcomes. I was worried about how the glue “worked” on a day to day level, and I found the best part of this study was that patients reported the cannula ‘moved less’, ‘hurt less’ and (in the agitated sweaty patients who always needed a resite) tended to get the cannula “through the night” without dislodgement compared to not using Cavilon. It’s now my routine practice – anyone out there wanting to give it a try, steal some adhesive from your local Wound Care Nurse’s trolley, and give it a go!


    1. Like the idea but can’t access the whole paper. Anyone know what they used to get the glue off? Were they special wipes and if so are they available in the uk?


  6. […] Simon Carley at St. Emlyn’s: JC – Should we Glue Cannulas to Patients? […]


Thanks so much for following. Viva la #FOAMed

Translate »