Prove it: Trauma patients need two IV lines

Many researchers question the value of prehospital fluid administration for trauma patients, which is an important concern given the short transport intervals normally associated with urban environments

 The Review
 What It Means for You

Medic 16 responds to a motor-vehicle collision on a county road about thirty miles from the city center. A vehicle ran off the road, rolled twice, and came to rest on its roof. The rollover ejected the driver who sustained head injuries that were incompatible with life.

The restrained and non-ejected front seat passenger is conscious and alert and complaining of pain in the right lower leg. A quick visual exam reveals an obvious midshaft closed lower leg fracture that did not produce a disruption in distal blood flow. The remainder of the patient's physical exam is unremarkable for trauma. The patient's blood pressure is 128/86 mmHg, the heart rate is 96 bpm, the respiratory rate is 16 bpm, the room-air pulse oximetry reading is 96%, and the Glasgow Coma Scale (GCS) score is 13.

After taking spinal stabilization procedures, Medic 16 begins transport to the hospital. Local protocol requires the medics to transport all survivors from a passenger compartment involving a death to a Level 1 trauma center. On the way, the medic places the patient on an electrocardiogram (ECG) monitor and establishes one large-bore intravenous (IV) line with normal saline at a to keep open (TKO) rate. The transport interval is uneventful.

The medics roll the cot into the emergency department passed the waiting trauma team. A nurse asks why there is only one IV and the medic responds that because of the patient's stable condition, the second line (also required by local protocol) was unnecessary. The nurse accepts the explanation but seems perturbed.

After completing the paperwork, the medic asks the nurse to sign the chart accepting care for the patient. As she signs her name, she adds that if patients are bad enough for trauma team activation, they should probably have two large-bore IV lines to guard against the possibility of deterioration before arrival at the hospital.

On the way back to the station, the medic thinks about what the nurse said. Do trauma patients, even critical ones, really need a second IV line?

Researchers at the Robert Wood Johnson Medical School in New Jersey evaluated whether a second IV line provided any benefit for trauma patients (Merlin et al., 2011). Specifically, the researchers wanted to know if the second IV line influenced the patient's heart rate, blood pressure, total amount of fluid infused in the prehospital environment, rehospitalization rate, or 30-day mortality.

The study population originated from the local EMS system, which consisted of a basic life support (BLS) and an advanced life support (ALS) component serving 800,000 people in an urban and suburban response area. The BLS component included a mixture of paid and volunteer personnel staffing eight ambulances. The ALS component included six hospital-based ambulances staffed by two paid paramedics per ambulance.

To obtain the data, investigators searched the electronic medical records of the local EMS system using the terms “trauma” and “IV” for a 12-month period beginning 01 May 2008 through 30 April 2009. The retrospective database search produced 320 patients who met the inclusion criteria of a traumatic mechanism of injury with an IV established in the field by paramedics. Medics transported all of these patients to a level-one trauma center. The State EMS office defined the physiological, anatomical, and mechanism of injury criteria that paramedics used to make destination decisions.

The New Jersey State Trauma Triage criteria (State of New Jersey Department of Health and Senior Services, Office of Emergency Medical Services, 2007) are very similar to the Trauma Triage Criteria established by a national expert panel on field triage assembled by the Centers for Disease Control and Prevention [CDC] (Sasser et al. 2009). For example, the physiological section in the New Jersey triage guidelines adds abnormal heart rate to the physiological criteria established by the CDC. New Jersey adds high-voltage electrical injuries to the CDC list of anatomical injuries that medics should transport to a trauma center. All of the mechanism of injury criteria requiring trauma center evaluation in the New Jersey guidelines appears in the CDC guidelines with the exception of vehicle extrication greater than 20 minutes with an injury and being an unrestrained passenger in a vehicle roll over. On the other hand, the CDC's list of injury mechanism requiring trauma center transport is more comprehensive (including death in the same passenger compartment) and the CDC includes a list of special circumstances such as age extremes that the New Jersey guidelines do not address.

The New Jersey EMS treatment protocols recommended that medics establish two patent IV lines when transporting any patient who met the criteria for trauma center destination. State protocols allow medics to use an intraosseous line as one of the two lines, either by starting with an IO and following with a peripheral IV or starting with the peripheral IV and following with an IO. Of course, a variety of factors determines whether medics actually establish a second IV line. As a result, some patients arrived with one IV line and others with two.

Because the study was retrospective in nature, the researchers presumed that medics obtained the recorded vital signs before establishing the IV and administering fluid therapy. The researchers used the first set of vital signs taken in the emergency department as the post-therapy vital signs. Researchers compared patient injury patterns using the Injury Severity Score (ISS) to verify that both study groups were similarly injured. Research assistants contacted patients by telephone after hospitalization to determine rehospitalization and mortality rates.

The results of the investigation appear in Table 1. The unadjusted (no regression analysis) column represents raw data for all patients. By the time the patients arrived in the hospital, the average difference in the volume of fluid infused between the two groups was 348 ml, with a 95% confidence interval of 235.6 ml and 461.1 ml. This term means that 95% of all the patients in the study had fluid volumes between those two values. The p value for the fluid volume variable was less than .05, which suggests that the difference in infused fluid volume between the two groups was probably the result of the second IV line and not random chance. There are no surprises there. All of this statistical mumbo-jumbo simply confirms what one might expect: that a second patent IV line allows a medic to infuse greater volumes of fluid during the prehospital interval.

The real question is whether this extra fluid provides any benefit for the patient. The results in Table 1 demonstrate that for the entire patient sample, the additional fluid did not significantly alter the patients' pulse oximetry values, Glasgow Coma Scale scores, heart rate, or systolic blood pressure. There was also no significant change in diastolic blood pressure for a subgroup of patients who presented to paramedics with an initial diastolic blood pressure less than 60 mmHg.

There was however, an average change in diastolic blood pressure of about five mmHg in the entire patient group. The associated p value (0.015) suggests there is about a 1.5 in 100 probability that random chance produced the change. Looking at that statistic from another perspective, there is a 98.5 percent chance that the change in diastolic blood pressure resulted from the extra fluid infused. In a subgroup of patients who presented on the scene with a diastolic blood pressure less than 80 mmHg, the average change in diastolic blood pressure caused by the extra fluid was even more significant; about 8 mm Hg with a less than 1 in 100 chance of being caused by random chance.

The statistically significant average change in diastolic blood pressure that resulted from the additional fluid remained when the researchers held multiple variables constant. The greatest effect of the additional fluid on diastolic blood pressure occurred in patients who presented on scene with a diastolic blood pressure less than 60 mmHg. In this subgroup of patients, the additional fluid resulted in an average change of about 31 mmHg in diastolic blood pressure; with less than 4% chance that random chance caused the change.

Table 1: Difference in variable value between 1-IV group to 2-IV group



Adjusted for age, gender, whether the variable was below median on scene

Adjusted for age, gender, volume of fluid, and whether the variable was below median on scene

Pulse oximetry




Fluids volume difference between the two groups

348.4 ml
95% CI: 235.6, 461.1
p value = < 0.0001

357.2 ml
95% CI: 224.1, 470.4
p value < 0.0001

not applicable

Change on Glasgow Coma Scale score
All subjects
Initial GCS < 15




Change in heart rate
All subjects
HR > 100 bpm on scene




Change in systolic blood pressure
All subjects
SBP < 132 mmHg on scene
SPB < 100 mmHg on scene









Change in diastolic blood pressure for all subjects

5.32 mmHg
95% CI: 1.02, 9.62
p value = 0.015

5.32 mmHg
95% CI: 1.02, 9.62
p value = 0.015

5.41 mmHg
95% CI: 0.76, 10.07
p value = 0.023

Change in diastolic blood pressure when DBP < 80 mmHg on scene

8.18 mmHg
95% CI: 2.16, 14.19
p value = 0.0077

8.58 mmHg
95% CI: 2.79, 14.37
p value = 0.0037

9.98 mmHg
95% CI: 3.38, 16.40
p value = 0.0029

Change in diastolic blood pressure when DBP < 60 mmHg on scene



31.63 mmHg
95% CI: 2.20, 61.05
p value = 0.035

The last two columns in the table represent mathematical estimations of the effects of the infused fluid on the physiologic criteria after modifying the data to account for other variables. These estimations result from an analysis technique known as regression analysis. Regression analysis allows researchers to isolate the effects of a single variable when many variables occur simultaneously. As an example, it is reasonable to assume that a 500 ml fluid dose will result in more dramatic heart rate changes if the patient was initially hypotensive than initially normotensive. If one determines how fluid affects the heart rates of hypotensive patients AND how fluid affects the heart rates of normotensive patients, one could start with the average blood pressure of a population and predict the heart rate change caused by a fluid bolus. Notice that the extra fluid infused through the second IV (the single variable) did not significantly alter the patients' pulse oximetry values, Glasgow Coma Scale scores, heart rates, or systolic blood pressures regardless of the effects of the many other variables.

Table 2: Injury Severity and Outcome


One IV Line

Two IV Line

Two IV Line

Injury Severity Scores




Rehospitalization rate




Rehospitalization rate




There was no statistical difference in the injury severity scores between the two groups suggesting that one group was not more severely injured (Table 2). Following discharge, research assistants were able to contact less than one-fourth (23%) of the surviving patients or their families. Of the contacts, there was no statistical difference in rehospitalization or 30-day mortality rates.

The authors concluded that establishing a second IV line in trauma patients did not significantly affect the patient's heart rate, blood pressure, rehospitalization rate, or 30-day mortality.

What it means for you
Many researchers question the value of prehospital fluid administration for trauma patients (Krausz, 2003; Madigan, Kemp, Johnson, & Cotton, 2008; Shoemaker et al., 1996). Animal models demonstrate that large-volume fluid resuscitation following uncontrolled hemorrhage actually increases total blood loss and decreases survival time compared to no fluid resuscitation (Solomonov, Hirsh, Yahiya, & Krausz, 2000). Researchers in Houston found a mortality reduction when EMS personnel did not administer prehospital fluids to victims of penetrating torso injury (Bickell et al., 1994).

Because of the short transport intervals normally associated with urban environments, even critical trauma patients probably receive no benefit from prehospital fluid administration (Smith, Bodai, Hill, & Frey, 1985). Researchers in Washington, D. C. could not demonstrate any benefit of prehospital fluid administration when the out-of-hospital interval was less than 30 minutes (Dalton, 1995). In a study involving almost 7,000 trauma patients with an average out-of-hospital interval of 36 minutes, researchers could not demonstrate a survival benefit to prehospital fluid administration even when the patient was hypotensive on EMS arrival (Kaweski, Sise, & Virgilio, 1990). Researchers in Pennsylvania found that although prehospital fluid administration did improve systolic blood pressure in patients suffering blunt trauma who were hypotensive on scene, it did not improve mortality rates or length of hospital stay (Dula, Wood, Rejmer, Starr, & Leicht, 2002).

A Practice Management Guideline Committee of the Eastern Association for the Surgery of Trauma reviewed all available literature that focused on prehospital vascular access and fluid resuscitation of trauma patients written in the English language and published between 1982 and 2007 (Cotton et al., 2009). The Committee concluded that there was insufficient evidence supporting a rate or volume of fluid administration to trauma patients in the prehospital setting.

One limitation to this study as with many studies involving the EMS environment is the accuracy of the data. The authors report that there were many errors in the charts, such as misspelled names and words and inaccurate dates. Vital sign measurements normally occur under more stressful conditions than found when completing the demographic section of a patient care record. It is reasonable to assume that under those conditions, medics may place a greater emphasis on performing skills they deem as life saving than on the accuracy of recording vital signs. In addition, there is no guarantee that the medics obtained the patient's vital signs using standardized methods. In fact, field medics recorded the initial set of vital signs while ED staff recorded the second set. It is possible that different healthcare providers assessing the same patient could obtain significantly different vital signs even when taking the measurements moments apart.

Another factor that limits the applicability of these findings is the location of the study. This study occurred in an urban setting at a level-one trauma center. Although the authors did not report out-of-hospital intervals for these patients, it is reasonable to assume that patients in more rural or frontier environments will likely spend more time out of hospital. Patients who must travel farther distances to receive surgical care may actually benefit from the extra fluid provided by a second IV line. The researchers did not structure this study to answer that question.

Finally, the researchers could only gather long-term outcome data for a small number of patients. Because of inaccurate contact information, the research assistants could not contact half of the patients following hospital discharge. Another 23% never returned messages left by the research assistants. The authors attempted to review all records at their trauma center to see if any of the missing patients returned, however they did not report if they were successful. One can only speculate as to the possibility of different results had the researchers been able to make contact with those missing patients.

This is the first published study to evaluate the prehospital placement of one versus two IV lines for trauma patients. Although the authors were able to demonstrate changes in diastolic blood pressure as the result of the additional fluid provided by the second IV line, they could not demonstrate significant improvements in patient condition or outcome. The available literature challenges the practice of prehospital fluid resuscitation for trauma patients. At a minimum, it seems reasonable for EMS agencies to reexamine the time-honored practice of requiring medics to establish two large-bore IV lines in all trauma patients.


Bickell, W. H., Wall, W. J., Pepe, P. E., Martin, R. R., Ginger, V. F., Allen, M. K. & Mattox, K. L. (1994). Immediate versus delayed fluid resuscitation for hypotensive patients with penetrating torso injuries. New England Journal of Medicine, 331(17), 1105-1109.

Cotton, B. A., Jerome, R., Collier, B. R., Khetarpal, S., Holevar, M., Tucker, B., Kurek, S., Mowery, N. T., Shah, K., Bromberg, W., Gunter, O. L., Riordan, W. P. Jr., & the Eastern Association for the Surgery of Trauma Practice Parameter Workgroup for Prehospital Fluid Resuscitation. (2009). Guidelines for prehospital fluid resuscitation in the injured patient. Journal of Trauma, 67(2), 389-402. doi:10.1097/TA.0b013e3181a8b26f

Dalton, A. M. (1995). Prehospital intravenous fluid replacement in trauma: an outmoded concept? Journal of the Royal Society of Medicine, 88(4), 213P–216P.

Dula, D. J., Wood, G. C., Rejmer, A. R., Starr, M., & Leicht, M. (2002). Use of prehospital fluids in hypotensive blunt trauma patients. Prehospital Emergency Care, 6(4), 417– 420.

Kaweski, S. M., Sise, M. J., & Virgilio, R. W. (1990). The effect of prehospital fluids on survival in trauma patients [abstract]. Journal of Trauma, 30(10), 1215–1218.

Krausz, M. M. (2003). Fluid resuscitation strategies in the Israeli army. Journal of Trauma, 54(5 Suppl), S39-S42. doi:10.1097/01.TA.0000064506.47688.51

Madigan, M. C., Kemp, C. D., Johnson, J. C., & Cotton, B. A. (2008). Secondary abdominal compartment syndrome after severe extremity injury: Are early, aggressive fluid resuscitation strategies to blame? Journal of Trauma, 64(2), 280-285. doi:10.1097/TA.0b013e3181622bb6

Merlin, M. A., Kaplan, E., Schlogl, J., Suss, H., DosSantos, F. D., Ohman-Strickland, P., & Shiroff, A. (2011). Study of placing a second intravenous line in trauma. Prehospital Emergency Care, 15(2), 208–213. doi:10.3109/10903127.2010.545612

Sasser, S. M., Hunt, R. C., Sullivent, E. E., Wald, M. M., Mitchko, J., Jurkovich, G. J., Henry, M. C., Salomone, J. P., Wang, S. C., Galli, R. L., Cooper, A., Brown, L. H., Sattin, R. W. and the National Expert Panel on Field Triage, Centers for Disease Control and Prevention (CDC). (2009). Guidelines for field triage of injured patients. Recommendations of the national expert panel on field triage. Morbidity and Mortality Weekly Review: Recommendations and Reports, 58(RR-1), 1–35.

Shoemaker, W. C., Peitzman, A. B., Bellamy, R., Bellomo, R., Bruttig, S. P., Capone, A., Dubick, M., Kramer, G. C., McKenzie, J. E., Pepe, P. E., Safar, P., Schlichtig, R., Severinghaus, J. W., Tisherman, S. A., & Wiklund, L. (1996). Resuscitation from severe hemorrhage [abstract]. Critical Care Medicine, 24(2 Suppl), S12-S23.

Smith, J. P., Bodai, B. I. , Hill, A. S., & Frey, C. F. (1985). Prehospital stabilization of critically injured patients: A failed concept [abstract]. Journal of Trauma, 25(1), 65–70.

Solomonov, E., Hirsh, M., Yahiya, A., & Krausz, M. M. (2000). The effect of vigorous fluid resuscitation in uncontrolled hemorrhagic shock after massive splenic injury. Critical Care Medicine, 28(3), 749–754.

State of New Jersey Department of Health and Senior Services, Office of Emergency Medical Services. (2007). Trauma triage guidelines, Retrieved from

About the author

Kenny Navarro is an Assistant Professor in the Department of Emergency Medicine at the University of Texas Southwestern Medical School at Dallas. He also serves as the AHA Training Center Coordinator for Tarrant County College. Mr. Navarro serves as an Emergency Cardiovascular Care Content Consultant for the American Heart Association, served on two education subcommittees for NIH-funded research projects, as the Coordinator for the National EMS Education Standards Project, and as an expert writer for the National EMS Education Standards Implementation Team. Send correspondence concerning any articles in this section to Kenneth W. Navarro, The University of Texas Southwestern Medical School at Dallas, 5323 Harry Hines Blvd MC 8890, Dallas, Texas 75390-8890, or e-mail

  1. Tags
  2. Research Reviews
  3. Trauma

Join the discussion


logo for print