Review Of: Elliott RH, Kalish BT, Cotton BA, et al. Prehospital intravenous fluid administration is associated with higher mortality in trauma patients: A national trauma data bank analysis. Annals of Surgery. 2011;253(2) [Epub ahead of print]
This study is a retrospective study of a large database, the National Trauma Data Bank. The aim was to evaluate how prehospital IVs and prehospital fluid administration in trauma patients affects mortality. Of the 1.4 million patients in the database, 776,734 were pulled for analysis. Approximately 50% of these patients had a prehospital IV documented. Analysis of several variables found the following:
• Unadjusted mortality IV fluid vs. no IV fluid: 4.8% vs. 4.5%
• Patients receiving prehospital IV fluids had an 11% increase in mortality; (odds ratio 1.11)
The authors conclude that prehospital IV fluid administration is associated with an increase in mortality and should be discouraged in all trauma patients.
Medic Marshall: I’m not sure where to start. First of all, if you read the complete study, you’ll find that the authors were unable to differentiate between the patients who received an IV only (saline lock) versus those who didn’t receive IV fluids. They just grouped them all together and recorded it as if they all received IV fluids. Does anyone else see a problem here?
Secondly, they don’t report the amount of fluid patients received in the prehospital setting; you’d think this would be an important variable to consider. I know the science of fluid administration in trauma patients is changing toward the attitude that providers ought to be more tolerant of a level of hypotension and not pour fluids into patients. Third, the authors don’t determine where the IVs were established (on-scene versus en route). I’ve always followed the philosophy, “IVs are established en route,” unless I need to spend time managing the airway and the patient needs drugs to accomplish that.
The bottom line for me is this: This study is loaded with problems, and the authors’ conclusions, with all due respect, are misguided. Imagine not starting an IV on your patient, who’s obviously bleeding out and watching their blood pressure rapidly deteriorate into the 80s, 70s, 60s and so on. So you’re just supposed to stand by and watch? That’s what the authors of this study seem to be alluding to.
Doc Wesley: I’m as confused as Marshall about the purpose of this “study.” This is data mining at its worst. The authors report the results from the runs with “complete” data and admit that more than half the records were incomplete. Additionally, they provide information on other care rendered, such as spinal immobilization, which they state occurred in less than 10% of patients and medical anti-shock trousers (MAST) application to almost half the patients. This sounds backwards to me. How many of your blunt trauma patients get spinal immobilization? I’ll bet it’s the majority. MAST? Is anyone still using it?
Finally, this paper is an example of the difference between statistical and clinical significance. Statistically, there appears to be an 11% increase in mortality with IV administration. But is this significant? No. That’s an odds ratio of 1.11. I’m not clinically impressed unless the odds ratio exceeds two or more. For example, bystander CPR and return of spontaneous circulation (ROSC) from cardiac arrest has an odds ratio of seven. That means that bystander CPR patients are seven times more likely to regain a pulse than those without bystander CPR.
Amusingly, deep in the data of this paper, the authors also report that Hispanic patients had an odds ratio of 1.3 when it came to IV administration and death. Being non-white was associated more often with death than whether a patient received an IV. The lesson to be learned here is that when the database is large, you can very easily find “associations” without cause and effect.
Objective: Prehospital intravenous (IV) fluid administration is common in trauma patients, although little evidence supports this practice. We hypothesized that trauma patients who received prehospital IV fluids have higher mortality than trauma patients who did not receive IV fluids in the prehospital setting.
Methods: We performed a retrospective cohort study of patients from the National Trauma Data Bank. Multiple logistic regressions were used with mortality as the primary outcome measure. We compared patients with versus without prehospital IV fluid administration, using patient demographics, mechanism, physiologic and anatomic injury severity, and other prehospital procedures as covariates. Subset analysis was performed based on mechanism (blunt/penetrating), hypotension, immediate surgery, severe head injury, and injury severity score.
Results: A total of 776,734 patients were studied. Approximately half (49.3%) received prehospital IV. Overall mortality was 4.6%. Unadjusted mortality was significantly higher in patients receiving prehospital IV fluids (4.8% vs. 4.5%, P < 0.001). Multivariable analysis demonstrated that patients receiving IV fluids were significantly more likely to die (odds ratio [OR] 1.11, 95% confidence interval [CI] 1.05–1.17). The association was identified in nearly all subsets of trauma patients. It is especially marked in patients with penetrating mechanism (OR 1.25, 95% CI 1.08–1.45), hypotension (OR 1.44, 95% CI 1.29–1.59), severe head injury (OR 1.34, 95% CI 1.17–1.54), and patients undergoing immediate surgery (OR 1.35, 95% CI 1.22–1.50).
Conclusions: The harm associated with prehospital IV fluid administration is significant for victims of trauma. The routine use of prehospital IV fluid administration for all trauma patients should be discouraged.