Original Contributions

Fluid resuscitation of trauma patients: how fast is the optimal rate?

Yasuaki Mizushima MD*, Hideo Tohira MD, Yasumitsu Mizobata MD, Tetsuya Matsuoka MD, Junichiro Yokota MD

Osaka Prefectural Senshu Critical Care Medical Center, Osaka, 598-0048, Japan

Accepted 16 March 2005

Abstract The Advanced Trauma Life Support guidelines recommend an initial rapid infusion of fluid (1-2 L) in trauma and hemorrhage victims as a Diagnostic procedure to aid Treatment decisions. Although patient response to initial fluid resuscitation is the key to determining therapeutic strategies, the appropriate rate of infusion is not clearly defined. Ninety-nine adult (age N16 years) blunt Trauma victims with hypotension were enrolled. Patients were classified into 3 groups according to Hemodynamic state after initial fluid resuscitation and requirement of surgical intervention. Total volume and rate of infusion differed significantly between the groups ( P b .05). Patients requiring fluid administration at higher rate were all hemodynamically unstable and required immediate surgical intervention. Moreover, rate of infusion was the best predictor of the patients who required immediate surgical intervention. Moderate fluid infusion rate should be considered to allow identification of the patient’s response to initial fluid resuscitation.

D 2005

Introduction

Despite major advances in the management of trauma victims, traumatic injury remains one of the leading causes of death during the first 3 decades of life [1]. Moreover, effective treatment of critically injured patients with hypo- volemic shock continues to be a formidable challenge. The main management strategies in hemorrhagic shock are the arrest of bleeding and replacement of circulating fluid volume. The Advanced Trauma Life Support guide- lines recommend an initial rapid infusion of fluid in the

T Corresponding author. Tel.: +81 724 64 9911; fax: +81 724 64 9932.

E-mail address: [email protected] (Y. Mizushima).

management of trauma and hemorrhage victims as a diagnostic procedure to aim in further treatment decisions [2]. Patients still unstable after aggressive volume resusci- tation must undergo Urgent surgery to control the source of bleeding.

Although patient response to initial fluid resuscitation is the key to determining an appropriate therapeutic strategy, the rate of fluid infusion has not been clearly defined. Moreover, recent controversies have arisen regarding end points in fluid resuscitation [3]. Several authors suggest that aggressive fluid resuscitation before control of bleeding may result in increased blood loss [4,5]. However, hypotensive patients should be stabilized rapidly with moderate fluid infusion to achieve and maintain perfusion of essential organs. Both inadequate and excessive fluid resuscitation

0735-6757/$ - see front matter D 2005 doi:10.1016/j.ajem.2005.03.015

Fig. 1 Initially, patients were classified into 3 groups according to hemodynamic parameters after initial fluid resuscitation and surgical intervention. Hemodynamically unstable was defined as sustained hypotension (SBP b90 mm Hg) or prolonged tachycardia (HR N120 beats per minute), with no or unsatisfactory response to initial fluid resuscitation.

should be avoided. Therefore, we conducted a study to determine the optimal rate of fluid infusion in the Initial resuscitation of trauma victims with hypotension.

Material and methods

Adult (age z16 years) blunt trauma victims with systolic arterial blood pressure (SBP) of 90 mm Hg or less upon admission were identified for retrospective analysis. All patients included in the study were initially evaluated and managed at our institute. Patients who were dead on arrival or not admitted directly to our hospital were excluded from the study. Standard trauma Resuscitation protocols were used for all other components of care. None of the patients had undergone prehospital administration of intravenous fluids.

If shock was present, at least two large-bore percutaneous catheters were placed, and a rapid Bolus infusion of Ringer’s lactate was started immediately. central venous access was used as necessary. If a hemodynamically normal state was achieved, the rate of fluid infusion was reduced and adjusted to maintain adequate vital signs. If the patient remained unstable, further rapid fluid and blood infusions were performed. Data on total fluid volume and time of initial resuscitation until surgical intervention or transfer to the intensive care unit (ICU) were collected for each patient. All surgical interventions to control hemorrhage were reviewed and reevaluated. The clinical outcome, Injury Severity Score (ISS), and predicted Probability of survival of each study patient were obtained from medical records and the trauma registry.

Table 1 Characteristics of the study patients
	A	B	C
No. of patients	33	27	39
Age (y)	47.6 F 19.2	45.9 F 20.5	51.5 F 21.5
Initial SBP (mm Hg)	79.8 F 11.0	71.5 F 13.5	63.7 F 20.4T,TT
Initial HR (beats per minute)	90.2 F 27.7	101.2 F 25.8	111.6 F 135.2T,TT
ISS	22.1 F 13.6	28.8 F 15.4	39.7 F 13.1T,TT
RTS	6.0 F 1.7	5.7 F 1.5	3.7 F 2.2T,TT
TRISS	0.74 F 0.29	0.71 F 0.31	0.29 F 0.29T,TT
Mortality	2 (6.1%)	3 (11.1%)	32 (82.1%)T,TT
T P b .05 compared with A. TT P b .05 compared with B.

Hemodynamic response to initial fluid re”>Table 2 Site of hemorrhage or cause of hypotension A B C Hemorrhage				prolonged tachycardia (heart rate [HR] N120 beats per minute) with absent or unsatisfactory response to initial fluid resuscitation. Surgical intervention included radiological Transarterial embolization to control bleeding in patients with
Pleural	1		5	pelvic fracture.
Peritoneal	2	8	15
Retroperitoneal	3	13	8	3.2. Clinical characteristics
External hemorrhage Multiple extremity fracture	15	6	3	The demographic and outcome comparisons are listed in
Multiple sites			8	Table 1. Initial SBP decreased significantly in group C in
Other				comparison with that in other groups. Also, initial HR was
neurogenic shock	5			significantly higher in group C than in group A or B. The
No identifiable cause	7			ISS was significantly higher in group C than that in group A or B. Revised Trauma Score and TRISS were

Data are presented as mean F SD. Differences between groups were considered to be statistically significant at P b.05, as determined by 1-way analysis of variance, Tukey test, or v² and Fisher exact tests.

Results

Ninety-nine patients were entered into the study during a period of May 1999 through October 2002. There were 66 men and 33 women. Mean age was 48.6 F 20.4 years. The causes of injury included motor vehicle crash (n = 31), fall (n = 29), pedestrian-automobile accident (n = 32), and other events (n = 7).

Patient groups

Initially, patients were classified into 3 groups according to hemodynamic parameters after initial fluid resuscitation and surgical intervention (Fig. 1): patients who were hemodynamically stable after initial fluid resuscitation and required no further surgical intervention (group A, n = 33); patients who were hemodynamically stable after initial fluid resuscitation but required surgical intervention to control hemorrhage (group B, n = 27); and patients who were hemodynamically unstable after initial fluid resuscitation and required further fluid resuscitation and immediate surgical intervention (group C, n = 39). Hemodynamic instability was defined as sustained hypotension (SBP b90 mm Hg) or

significantly lower in group C that in group A or B. However, there was no significant difference between group A and B in initial SBP, ISS, RTS, or TRISS. Actual mortality rates were 6.1 % in group A, 11.1% in group B, and a significantly high 82.1% in group C.

The sites of hemorrhage in each group are shown in Table 2. Multiple hemorrhAge SItes were present in 20.5% of patients in group C.

Hemodynamic response to initial fluid resuscitation

Patients in groups A and B responded to initial fluid resuscitation. Systolic blood pressure was restored 79.8 F

11.0 to 122.1 F 15.7 mm Hg in group A and 71.5 F 13.5 to

126.3 F 15.7 mm Hg in group B, and HR decreased 90.2 F

27.7 to 85.7 F 7.5 beats per minute in group A and 101.2 F

25.8 to 94.1 F 14.5 beats per minute in group B (Fig. 2). However, patients in group C failed to respond to fluid resuscitation; marked tachycardia (111.6 F 35.2 to 113.5 F 25.2 beats per minute) with depressed SBP (63.7 F 20.4 to 77.2 F 16.1 mm Hg) continued even after initial resusci- tation (Fig. 2).

Fluid resuscitation

Initial resuscitation time between admission and surgical intervention in groups B and C was significantly shorter than the time between admission and transfer to the ICU in group A (Table 3). However, there was no significant

Fig. 2 Hemodynamic response to initial fluid resuscitation. Patients in groups A and B responded to initial fluid resuscitation; SBP was restored and HR was decreased. However, patients in group C failed to respond to fluid resuscitation; they remained in marked tachycardia with depressed SBP. OA indicates on admission; ER, end of initial resuscitation.

A		B	C
Initial fluid volumes (mL)	1488 F 1136	2612 F 1329T	4061 F 1979T,TT
Initial resuscitation time (min)	98.6 F 39.3	77.4 F 32.1T	65.5 F 27.8TT
24 h Fluid (mL)	6284 F 2757	13400 F 7133T	20053 F 9745T,TT
24 h PRBC (U)	2.1 F 3.7	12.8 F 12.6T	38.4 F 23.3T,TT

difference in initial resuscitation time between group B and

Table 3 Fluid volumes and times of the resuscitation

Initial resuscitation time indicates time from admission to until surgical intervention (group B and C) or from admission to until transfer to ICU (group A);

24 h Fluid, intravenous fluid administration during the first 24 hours; 24 h PRBC, packed red blood cell transfusion during the first 24 hours.

T P b .05 compared with A.

TT P b .05 compared with B.

C. Total fluid administered during the initial resuscitation differed significantly between all groups. Also, a greater total amount of fluid and packed red blood cell transfusion (PRBC) were administered to group C patients than to group A or B patients during the first 24 hours after injury (Table 3). There was a significant difference between all groups in the total amount of fluid and PRBC administered during the first 24 hours.

Hemodynamically stable patients did not require fluid administration at more than 60 mL/min during the initial resuscitation (Fig. 3). Patients requiring fluid administration at higher rate were all hemodynamically unstable and required surgical intervention (Fig. 3).

Receiver operating characteristic curves

We constructed receiver operating characteristic curves (derived from rate of infusion, ISS, RTS, and initial SBP) and compared the areas under the ROC curves to predict a group C patient. Receiver operating characteristic curves indicated that rate of infusion (the area under the ROC curve, 0.95 [95% confidence interval, 0.92-0.99]) was the best predictor of group C patients characteristics (Fig. 4). On

Fig. 3 Relation between the rate of fluid administration and SBP after initial fluid resuscitation. Hemodynamically stable patients did not require a fluid administration rate of more than 60 mL/min in the initial fluid resuscitation. Patients requiring a rate of fluid administration at greater rate were all hemodynamically unstable and required surgical intervention.

the ROC curves, the cutoff value for the rate of infusion fluid (the value closest to the upper left corner of the ROC plot) was 45 mL/min (sensitivity, 79.5%; specificity, 91.7%).

Discussion

Although early, rapid fluid resuscitation remains the cornerstone of treatment for trauma and hemorrhage victims, the most important management principle in treating hemor- rhagic shock is to find the source of blood loss and stop it [6,7]. Patient response to initial fluid resuscitation should be observed to identify those patients with ongoing bleeding requiring surgical control [2,8]. It is generally accepted that 3 response patterns can be identified to initial fluid admin- istration in hemorrhagic shock: rapid responders, transient responders, or minimal or nonresponders [2,8]. Nonrespond- ers fail to show any Hemodynamic improvement after fluid administration because of ongoing hemorrhage at a greater rate. These patients need immediate surgical intervention

Fig. 4 Receiver operating characteristic curves of fluid admin- istration rate (Rate), ISS, RTS, and initial SBP for predicting group

C. The areas under the ROC curves were 0.95, 0.77, 0.23, and 0.31, respectively.

rather than volume replacement to control the hemorrhage. Otherwise, the prognosis is poor.

Concern has been expressed about aggressive fluid resuscitation before control of bleeding. Some investigators argue that it may disrupt thrombus formation, increase bleeding, and decrease survival [3-5,9]. Restricted fluid resuscitation may have a positive effect on uncontrolled hemorrhage but a negative effect on tissue perfusion in shocked patients who respond to fluid with stabilization of their vital signs. Therefore, it appears that a rational, moderate fluid infusion rate should be considered to allow identifica- tion of the patient’s response to initial fluid resuscitation.

The result indicated that patients requiring fluid infusion at more that 60 mL/min were all hemodynamically unstable and required immediate surgical intervention. These patients were considered nonresponders who needed immediate surgical intervention to control hemorrhage rather than aggressive fluid resuscitation. Interestingly, increasing the fluid administration at higher rate did not produce hemo- dynamic stability. Aggressive fluid resuscitation at higher rate may result only in excessive fluid administration and hemodilution in cases of uncontrolled hemorrhage. Receiver operating characteristic curves showed that the rate of infusion was the best predictor of whether uncontrolled hemorrhage requiring urgent surgical intervention would occur in our hypotensive patients rather than initial SBP, ISS, or RTS. On the ROC curves, the cutoff value for the rate of infusion fluid (the value closest to the upper left corner of the ROC plot) was 45 mL/min (sensitivity, 79.5%; specificity, 91.7%). Therefore, 45 mL/min might be the fluid administration rate at which the type of response to initial resuscitation can be identified in trauma victims with hypotension. It would take 20 to 45 minutes for an adult patient to receive the standard 1-2 L crystalloid infusion recommended by the Advanced Trauma Life Support guidelines at this rate.

Lewis [10] developed a computer Simulation model to evaluate prehospital fluid administration and found that fluids would be beneficial to increase blood pressure only if the bleeding rate was moderate and the infusion rate was approximately equal to it. He suggested that bleeding rates greater than 100 mL/min or less than 15 mL/min would not be affected by intravenous therapy [10]. Our findings agree with previous findings indicating that a rapid infusion rate does not restore blood pressure if the concomitant bleeding is occurring at a greater rate.

The fact that the rate of fluid administration was not always consistent throughout the resuscitation could be considered a limitation to this study. The rate may have been inconsistent for patients in groups A and B, who achieved a

hemodynamically normal state after initial resuscitation, because the rate of fluid administration was reduced after the initial bolus infusion. However, the rate of fluid administration in the group C patients was consistent during the initial resuscitation because the rate of infusion could not be reduced because of hemodynamic instability.

It should be noted that group C patients had a high mortality rate in this study. The predicted survival rate (TRISS) in group C was low enough to explain the high morality rate; however, the mortality rate can be reduced in such patients if the time between initial resuscitation and surgical intervention was decreased. It is undoubtedly that an initial imaging assessment (chest x-ray, pelvic x-ray, and Abdominal ultrasonography) provides useful information to detected sources of hemorrhage. Nonetheless, identifying the rate of infusion required also could be a good strategy for determining whether immediate surgical intervention is indicated. Further studies are required to evaluate this possibility and to determine the role of initial fluid infusion rate in the management of trauma and hemorrhage victims. In summary, our findings show that increasing the fluid administration rate did not produce hemodynamic stability. Aggressive fluid resuscitation at a higher rate may result in excessive fluid resuscitation and may cause a poor outcome in patients. The rate of infusion was the best predictor of whether uncontrolled hemorrhage would occur in our

hypotensive patients.

References

1. Shires G. Principles of trauma care. New York7 McGraw-Hill Co; 1985.
2. American College of Surgeons G. Advanced Trauma Life Support Program for physicians. 6th ed. Chicago (Ill)7 American College of Surgeons; 1997.
3. Dries DJ. Hypotensive resuscitation. Shock 1996;6:311 - 6.
4. Bickell WH, Wall Jr MJ, Pepe PE, et al. Immediate versus delayed fluid resuscitation for hypotensive patients with penetrating torso injuries. N Engl J Med 1994;331:1105 - 9.
5. Dutton RP, Mackenzie CF, Scalea TM. Hypotensive resuscitation during Active hemorrhage: impact on in-hospital mortality. J Trauma 2002;52:1141 - 6.
6. Mattox KL, Feliciano DV, Moore EE. Trauma. 4th ed. New York7 McGraw-Hill; 2000.
7. Hoyt DB. Fluid resuscitation: the target from an analysis of Trauma systems and patient survival. J Trauma 2003;54:531 - 5.
8. Greaves I, Porter K, Ryan J. Trauma care manual. London7 Arnold; 2001.
9. Martin RR, Bickell WH, Pepe PE, et al. Prospective evaluation of preoperative fluid resuscitation in hypotensive patients with penetrat- ing truncal injury: a preliminary report. J Trauma 1992;33:354 - 61.
10. Lewis Jr FR. Prehospital intravenous fluid therapy: physiologic computer modelling. J Trauma 1986;26:804 - 11.