Abstract

Background: Peripheral intravenous (IV) cannulation is a key
intervention in the prehospital setting, but inappropriate use may cause unnecessary pain, distress or risk of infection. The aim of this study was to examine the rate and appropriateness of prehospital cannulation and the relative importance of factors associated with increased likelihood of cannulation.
Design and setting: Cross-sectional survey of patients transported in Lincolnshire, East Midlands Ambulance Service.
Methods: Retrospective non-identifiable data for September 2006 were extracted. Clinical conditions were classified according to whether they warranted, did not warrant or were uncertain as to the need for cannulation. Other potential indications for cannulation including IV drug administration, reduced consciousness, systolic hypotension, respiratory depression and haemorrhage were combined to determine whether cannulation was indicated. Other variables were investigated as predictors of cannulation. The method of analysis was agreed at the outset.
Results: Paramedics cannulated 14.6% (1295/8866) of patients. IV drug administration, clinical indication, reduced conscious level, respiratory depression and hypotension were associated with greater likelihood of cannulation (p,0.001). Cannulation was more likely in older patients but was not associated with gender, haemorrhage or hypoglycaemia. Multivariate logistic regression showed IV drug administration as the strongest predictor of cannulation. Cannulation rates varied threefold by ambulance station (mean 13.4%, 5.8% to 19.0%). It was estimated that 202 (15.6%) of the cannulations performed could potentially have been avoided.
Conclusion: Rates of cannulation were higher than previous
studies with wide variations between ambulance stations. 15.6% of cannulations performed could have been avoided, thus reducing pain, distress and other potential complications such as thrombophlebitis, extravasation and infection. The generalisability of this study was limited by use of a single site, short duration and dependence on accurate retrospective data. The data demonstrating wide variations suggest that there may be scope for consideration of interventions to reduce cannulation rates.