The US media seem to be making a big thing of a recent article published ahead of print which demonstrates an association between increased mortality from trauma and the insertion of an intravenous line with or without the administration of fluid.
This was a retrospective cohort study of over 770 000 patients from the National Trauma Data Bank. Approximately half (49.3%) received ‘prehospital IV’, which could mean fluids, or could just mean insertion of an intravenous cannula: ‘we could not definitively differentiate IV fluid administration versus IV catheter placement alone‘.
Unadjusted mortality was significantly higher in patients in the prehospital IV group, although the abstract inaccurately reports this to be ‘in patients receiving prehospital IV fluids‘ (4.8% vs. 4.5%, P < 0.001).
Multivariable logistic regression was used to examine the relationship between prehospital IV and mortality in the 311,071 patients with complete data. After adjustment, prehospital IV patients had significantly higher mortality than those without a prehospital IV. The odds ratio of death associated with prehospital IV placement was 1.11 (95% CI 1.06–1.17). When Dead-On-Arival patients were excluded from the group as a whole, the association persisted (OR 1.17, 95% CI 1.11–1.23).

On subgroup analyses, the association between IV placement and excess mortality was maintained in nearly all patient subsets; the effect was more exaggerated in penetrating trauma victims.
Media speculation as to the reason for this association abounds, like USA Today‘s ‘those who are given pre-hospital IV fluids are actually 11% more likely to die than those who aren’t, not only because of transport delays but also in part because of the increased risk for bleeding that can accompany a fluid-induced increase in blood pressure‘. However the study did not record any pre-hospital times and could not tell which patients received fluid, let alone what the effect of fluid on blood pressure was.
The authors are open about this and other limitations: ‘The NTDB did not report prehospital transport times or differentiate urban versus rural care. Thus, we could not examine whether excess mortality in patients treated with IVs was directly associated with delays in transport to definitive care. We were also not able to control for transport time within the multiple regression model or perform a stratified analysis by urban versus rural patients. Perhaps this analysis would have identified a subset of patients who may benefit from IV placement‘.
No doubt this will be added to the pile of mainly hypothesis-generating literature quoted by the scoop-and-run brigade whose black-and-white worldview includes paramedics who want to delay proper treatment and a homogeneous trauma population whose lives can only be saved by a trauma surgeon in a hospital. Those who have evolved colour vision find this an interesting, but hardly practice-changing study; caution regarding injudicious fluid administration has been the game plan for many civilian and military pre-hospital providers since early last decade, and it is clear that different patients with different injury patterns, different degrees of physiological derangement, and different distances from the right hospital will continue to have different clinical needs specific to their presentation, some of which are likely to be of benefit if provided in the field, through an intravenous line.
Prehospital Intravenous Fluid Administration is Associated With Higher Mortality in Trauma Patients: A National Trauma Data Bank Analysis
Ann Surg. 2010 Dec 20. [Epub ahead of print]