Just a quick review of a thought provoking paper in the Annals of Emergency Medicine from a group of docs in Melbourne Australia. It’s thought provoking as I suspect that this reflects practice in many EDs across the developed world.
In a larger tertiary teaching hospital the investigators tracked patients with peripheral IVs placed in the ED. They wanted to know whether they were used after placement. This is important as patients don’t particularly like peripheral IVs and there is a small but clinically important risk of subsequent infection and patient harm.
[learn_more caption=”So, who was included in this study?”] 3829 patients were screened of whom 618 had IVs placed, 48 were excluded as for data collection or clinical (high severity cases) leaving 570 for analysis.
It’s a reasonable sample size and probably big enough to give us a picture of typical practice in this department.[/learn_more] [learn_more caption=”…and the main findings are?”] Perhaps unsurprisingly, but certainly dissapointingly half the IVs were not used in the ED. Some patients in this group were admitted to the wards and by my estimation an additional 35 were used on the wards. Personally I think these should be considered in the ‘used’ group as I am sure the patient does not give a monkey’s where they are used, just that they ‘are’ used. That changes the stats slightly….
- Number that are used becomes 321/570 = 56%.
- Number that remain unused = 249/570 = 44%
That’s not quite as impressive but the message remains the same. Lots of IVs placed in the ED are not subsequently used (apart from taking bloods I presume – although 43 patients did not have blood drawn through the cannula). The authors do this analysis themselves, but the title of the paper suggests that the half figure is more important – I don’t[/learn_more][learn_more caption=”Unanswered questions”] To give the authors credit the limitations section of this paper is good and raises many important points. It is single centre and retrospective which may bias some of the results, though the findings chime with my experiences in the UK so I am inclined (my own personal bias) to believe it. The key limitation is whether it is possible to differentiate at the time of insertion whether a cannula is needed? It’s fine to look back and say that it was not used, but what is more important is for us to try and identify the groups of patients who need a cannula right away for drugs/fluids OR who are high risk for need later. That is not answered in this paper, though the authors do recognise this as the differentiation between unused and unnecessary.
The topic is not especially new, there have been previous papers looking at prehospital placement, in US EDs, in acute medical units and other settings (see the authors references for more links)
[/learn_more] [learn_more caption=”What shall I do then?”] Three things…
- May I suggest a quick pause button before you place the next cannula. Does the patient really need it now?
- Have a chat to your colleagues who place IVs. What guidance have they had? What makes them choose a cannula over a simple blood draw or delay? It might be quite enlightening to have that conversation…….
- If one of my colleagues is looking for an audit project that would be quick, easy and clinically important then I think this might be great.[/learn_more]
2 thoughts on “JC: Is that IV really needed?”
Pingback: JC: Should we glue cannulas to patients? St.Emlyn's - St.Emlyn's
Pingback: SGEM#148: Stuck on You – Skin Glue for Peripheral IVs | The Skeptics Guide to Emergency Medicine