Early Resuscitation: The Controversy
Hypovolemic trauma patients will reliably respond to volume infusion with an increase in blood pressure, thus in the past ATLS recommended administration of 2L IVF as rapidly as possible. This recommendation has recently been questioned, as the downsides of dilutional anemia in the trauma setting (decreased O2 delivery, hypothermia, coagulopathies, and electrolyte abnormalities) and hypertension in the setting of uncontrolled hemorrhage (increased bleeding) the have come to light. Aggressive fluid administration often leads to a transient rise in BP which leads to increased bleeding and another episode of hypotension, producing a vicious cycle. For this reason, resuscitation has been divided into two phases: early (during active hemorrhage) and late (when hemorrhage has been controlled). Deliberate hypotensive is well-accepted in the operative arena [Enderby GEH: Hypotensive Anaesthesia, Edinburgh: Churchill Livingstone; 1985], but is highly controversial in the setting of trauma – that said, consideration should be given to minimizing fluid intake until hemorrhage has been controlled, at which point fluid can be titrated to desired effect.
In 1965, an animal (dog) study comparing blood loss secondary arterial injury in the setting of multiple differing conditions (BP, vasoconstrictors/vasodilation, fluid status, transfusion) showed that blood loss was minimized in hypotensive animals, followed by controls (no volume or vasoconstrictors), those who received vasoconstrictors, and those who received volume infusions (i.e. those who received volume had the most blood loss) [Shaftan GW et al. Surgery 58: 851, 1965]. Sharftan’s study is supported by multiple other animal studies showing that large volumes of resuscitation fluid can be harmful [Owens TM et al. J Trauma 39: 200, 1995; Riddez L et al. J Trauma 44: 1, 1998; Sakles JC et al. Ann Emerg Med 29: 392, 1997; Capone A et al. Resuscitation 29: 143, 1995], the best of which show no difference in cardiac output and/or regional perfusion between moderate and large-volume resuscitation.
A 1994 consensus panel on resuscitation from hemorrhagic shock suggested that spontaneous hemostasis and long-term survival could be maximized by limiting resuscitation fluids during the period of active bleeding (early hemorrhage/resuscitation) while seeking to keep perfusion only above the threshold for ischemia [Shoemaker WC et al. Crit Care Med 24S: S12, 1996] – note that some mammalian species are capable of sustaining MAP as low as 40 mm Hg for periods as long as 2 hours without deleterious effects [Miller, Chapter 63] although the human cerebral autoregulatory curve is thought to have a lower inflection point at 50 mm Hg in non-hypertensive patients (higher in those with a history of hypertension).
Bickell et al (RCT, 100 patients)
Bickell et al randomized 100 penetrating torso trauma patients to standard of care (up to 2L of prehospital crystalloid, average was 870 cc infused) versus delayed resuscitation (no fluid until reaching the OR, averaged 92 cc prehospital). At the time of surgery, blood pressures were identical but the standard of care group had markedly lower hemoglobin (10.7 vs 11.5 g/dL). In the operating room, there were no differences in the amount of fluid administered, although the prehospital resuscitation group required more rapid fluid administration. The delayed-resuscitation group had improved survival to discharge (70% versus 62%, p = .04) as well as reduced length of stay (11 days vs. 14 days, p = 0.006) [Bickell WH et al. NEJM 331: 1105, 1994]
Dutton et al (RCT, 110 patients)
Bickell’s study was followed by Dutton, who randomized 110 trauma patients with hypotension (defined as SBP < 90 mm Hg) to a systolic goal of 100 mm Hg versus 70 mm Hg until the completion of surgery. In contrast to Bickell, Dutton found no difference in mortality, although the study may have been confounded by a difference in injury severity scores in favor of the 100 mm Hg group (19.6 vs 23.4) as well as the profundity of shock at presentation. At 24 hours, lactate and base deficit normalized in both groups, and both fluid and blood product requirements were similar. Importantly, the low-pressure group had a significantly longer hospital and ICU length of stay [Dutton RP et al. J Trauma 52: 1141, 2002]
The notion that aggressive fluid resuscitation can be detrimental prior to controlling hemorrhage is supported by two retrospective studies. In the first, 4856 EMS patients were compared with 926 non-EMS patients, and linear model analysis controlling for age, gender, mechanism of injury, cause of injury, Injury Severity Score (ISS), and severe head injury showed a crude mortality rate of 9.3% in the EMS group versus 4.0% in the non-EMS group (RR 2.32, p < .001), after adjustment for ISS, the RR fell to 1.60 (p = .002). Subgroup analysis showed that patients with ISS > 15 were twice a likely to die if transported by EMS (RR 28.8% vs 14.1%) [Demetriades D et al. Arch Surg 131: 133, 1996]. In the second, 527 patients at Maryland’s Shock Trauma Center (which has had the RIS since 1990) were compared to the STC Trauma Registry as well as historical controls that did not have access to the RIS. For patients with access to the RIS, 9724 ml were infused, and overall survival was significantly less than expected (52.9% vs. 61.8%, p < 0.001). In penetrating trauma the survival rates were similar, thus the differences are likely due to blunt trauma (48.8% vs. 63.0% survival, p < 0.001). In patients who received less than 6000 ml via the RIS, there was no difference, whereas in those that did, there was a significant mortality difference (37.2% vs. 57.2%, p < 0.0001). Compared to matched controls, those who received fluids via the RIS had a 4.8 times RR of dying (95% CI 2.4-7.1) [Hambly PR et al. Resuscitation 31: 127, 1996]
Despite the accumulating evidence supporting fluid restriction in the ED/preoperative phase of trauma, one must also keep in mind the different physiology of the patient in hemorrhagic shock (maximal vasoconstriction secondary to endogenous epinephrine) and the patient sedated or under anesthesia (vasodilation due to exogenous pharmacologic agents). At the least, the use of anesthetic agents should be minimized. Furthermore, patients at-risk for ischemic complications (TBI, known ischemic heart disease, elderly) have been excluded from most of the above studies, and intentional hypotension may be advantageous in those groups, until it is shown, intentional hypotension should not be used in at-risk populations.
After hemorrhage has been controlled, fluids can be more safely administered. The emphasis should be on normalization of organ function and perfusion, and not on some arbitrarily defined hemodynamic goals or laboratory values. Heart rate, blood pressure, and urine output are notoriously unreliable guides to resuscitation. One must be wary of “occult hypoperfusion syndrome” in which normotension is maintained in a severely hypovolemic patient through extreme vasoconstriction [Blow O et al. J Trauma 47: 964, 1999]
In fact, there are no ideal markers for resuscitation. Urine output is often confounded by intoxication, pharmacologic agents, or renal injury. Lactate requires time, and cardiac output requires a PA catheter.
Tissue-specific techniques (ex. skin, SQ, and skeletal muscle [McKinley BA and Butler BD. Crit Care Med 27: 1869, 1999] oxygenation) are investigational, but show some promise. Currently, however, oxygenation measurements are impractical. Measurement of tissue carbon dioxide has shown some promise, ex. in the GI tract [Puyana JC et al. J Trauma 46: 9, 1999], but may soon become more practical, as sublingual pCO2 has been shown to correlate with the hemodynamic signs consistent with shock [Weil MH et al. Crit Care Med 27: 1225, 1999]
Access and Delivery of Fluids
IV access is critical and should be established as soon as possible. Place at least two 16 ga. IVs in any trauma patient.
Order of preference for IV access is as follows: 1. AC 2. Other peripheral 3. Subclavian 4. Femoral 5. IJ 6. Intraosseous
The subclavian approach is favored for quick central access in the trauma patient, as it is rarely traumatized, is easily accessed, and can be left in place. Place on the same side as chest tubes (when present). If unattainable, however, the next step is femoral access – major concerns re: femoral placement are the increased incidence of DVT [McGee DC and Gould MK. NEJM 348: 1123, 2003] and infection, but in a trauma situation are reasonable if removed as soon as possible.
Thermal equilibrium (or at the least, avoidance of hypothermia) should be attained. Hypothermia can worsen coagulopathies and acidosis, as well as increase metabolic demands, thereby leading to a supply:demand mismatch. Note that many trauma patients will arrive with significant hypothermia, necessitating the use of forced air warmers in addition to heated IV fluids.
Rapid infusion devices are controversial, with some theoretical benefits (higher temperature, less acidosis [Dunham CM et al. Resuscitation 21: 207, 1991]) but potentially leading to increased rebleeding [Hambly PR and Dutton RP. Resuscitation 31: 127, 1996]. In the absence of further evidence, RID should only be used to keep blood pressure above some (arbitrary) critical value.
Choice of Resuscitation Fluids
With regards to fluid administration, there is no data to support the use of colloids over crystalloids in the trauma setting. For hemorrhagic shock, however, PRBCs are considered the standard treatment. Patients receiving 10U or more of PRBCs should probably receive replacement plasma in a 1:1 ratio. Prior to 5U of PRBCs, plasma is not indicated – from 5-9U of PRBCs the utility (or need) of replacement plasma is not known. Also, in patients who are bleeding profusely and uncontrollably, plasma can be given before labs or a pre-determined criteria is hit, as rapid blood and factor loss can lead to a vicious cycle of increased bleeding. Note also that rapid blood transfusion can lead to “citrate intoxication” in which serum calcium is bound, leading to profound myocardial depression and hypotension. Patients receiving massive blood transfusions should therefore have frequent ionized calcium assessments [Irving GA. Can J Anaesth 39: 1105, 1992].