Studies that involve multiple EMS systems and bigger numbers are often perceived as stronger research. In this month’s column, I’m going to look at two studies that used electronic trauma registries to compare different prehospital interventions. Interestingly, the smaller and more local study may actually have a more valuable take-home message.
Prehospital IVs & Trauma Patients
Haut ER, Kalish BT, Cotton BA, et al. Prehospital intravenous fluid administration is associated with higher mortality in trauma patients: A national trauma data bank analysis. Ann Surg. February 2011; 253(2):371–377.
Dr. Haut and co-authors retrospectively reviewed the electronic records of patients in a large U.S. trauma center data bank. They compared outcomes of a trauma patient cohort who received prehospital IV therapy to those who didn’t receive an IV. They report that patients receiving IV fluids are significantly more likely to die (odds ratio of 1:11). They conclude that prehospital IV fluid administration in trauma should be discouraged.
The facts: The study’s title and conclusion are about fluid administration, yet the authors could only report if an IV was initiated. How much fluid was administered, if any, wasn’t available and, therefore, wasn’t considered. They also didn’t report how many of these IVs were started on scene, possibly extending scene times.
Of 1,466,887 patients in the database, only 53% (776,734 patients) had prehospital procedures recorded, and of those, only 40% (311,071)—21% of total records—were considered complete. Although not representative, this still may have been a nice sample to review. The problem is that some of their numbers seem bizarre. For example, only 8% of these trauma patients received spinal immobilization, but 28% received pneumatic anti-shock garments (PASG). This and other mathematical irregularities may point to a skewed or very regional sample.
After reading the full report, I’d caution anyone about running to their medical director for a protocol change based on this study. The evidence that was gathered and presented doesn’t support the conclusion, and more importantly, it doesn’t imply any kind of cause-and-effect relationship.
This study is a perfect example of a headline that grabs our attention but doesn’t have enough substance to back up its claim. Weighed scientifically, the conclusion of this study should only be that researchers can fall prey to the same tactics as tabloid newspapers. I hope everyone reading this remembers to read the full article before jumping to conclusions based on research abstracts and titles.
Should Police Transport Penetrating Trauma Patients?
Band RA, Pryor JP, Gaieski DF, et al. Injury-adjusted mortality of patients transported by police following penetrating trauma. Acad Emerg Med. 2011; 18:32–37.
“No difference” is the message from Roger Band, MD, and co-authors after comparing outcomes of trauma patients transported by police vs. ambulance. The setting is Philadelphia, where police have been transporting penetrating trauma victims for more than two decades. The policy states that if EMS is on scene first they’ll transport the patient, but if police arrive first and EMS is delayed, police can make the transportation decision. Additionally, no formal policy was in place for care provided by police. This was a retrospective cohort study for a five-year time period (January 2003–December 2005). All patients were transported to a Level 1 urban trauma center.
Of the 2,127 patients, police transported 26.8%. Nearly 71% were gunshot wounds, and 29% were stab wounds. Patients transported by police had a higher injury severity score, rarely received medical care (including bleeding control) and had a higher mortality rate.
The authors used a mathematical model to account for differences in severity, trying to make the comparison fair. Using an adjusted severity score, they found no difference in outcomes. Their carefully worded discussion and recommendation is to conduct additional research and for now, continue allowing police to transport.
Unlike the IV study described above, the authors of this study used the records of all patients arriving at a single Level 1 trauma center. By keeping the project smaller and focused on one system, their data and statistical methods are cleaner and lead to a more useful conclusion. The bottom line in both studies is that EMS needs to be more proactive with ensuring good data collection and more controlled prospective research to prove the usefulness of its interventions. More importantly, mandatory hospital trauma databases that track patient outcomes for accreditation should be coordinating more closely with prehospital systems so that future research can be collaborative and more accurate. JEMS
Trauma registry: An electronic database that holds information about injuries, care and outcomes of trauma patients. Such registries are often used for research because they can be easily accessed. Unfortunately, their prehospital data may not be as accurate as their hospital data.
Cohort study: A study of a group of people who share a common characteristic within a defined period of time. In this month’s study, the cohort is a group of people exposed to trauma who were entered into a national data bank between 2001 and 2005.
Odds ratio (OR): A statistical term used to compare two values and predict whether an event is more likely to occur with one group than another. In this case, the OR is being used to predict whether trauma patients who receive a prehospital IV are more likely to die than those who don’t. An OR of “1” would mean that both groups are equally likely to die. This study found an OR of 1:11, and the authors concluded a greater likelihood of death exists if the patient receives an IV.
What we know: IVs in trauma patients should be started en route to a hospital; scene times should be minimized; and IV fluid administration should be used carefully and only in cases of serious hypotension.
What this study adds: more evidence that electronic databases have incomplete prehospital records. EMS needs to have better coordination with trauma centers to improve tracking and prehospital care research.
This article originally appeared in April 2011 JEMS as “Fluids for Trauma Patients & Police Transport.”