Preferences of critical care registrars in fluid resuscitation of major trauma patients: concordance with current guidelines.
We conducted a survey at the Royal Melbourne Hospital, a large tertiary hospital and major trauma centre in metropolitan Melbourne, Victoria where 3072 trauma patients were admitted in 2009, of which 787 were classified as major trauma. We surveyed 102 registrars working in critical care (anaesthesia, intensive care and emergency). Surveys were collected between June 2009 and June 2010. Participation was voluntary and anonymous. The survey asked several demographic questions before giving an example of a typical major trauma patient who was haemodynamically unstable (systolic blood pressure 85 mmHg, heart rate 110 per minute) with evidence of lower limb injuries. Participants were asked a variety of questions about how they would manage this patient's fluid resuscitation, as well as that of a similar patient who had a head injury, and what their blood transfusion limits would be for patients with these scenarios. Results were compiled with subgroup analysis of seniority of registrars when there was a variation from currently accepted practice. The study was approved by the hospital's Human Research Ethics Committee as a registered quality assurance project (QA2009033).
There were 66 responses (representing a 65% response rate), with 37 participants from emergency, 19 from intensive care and 10 from anaesthesia. Thirty-four participants were senior registrars with three or more years of experience in their specialty, and 32 were junior registrars with less than three years experience. Seniority was represented evenly among the departments.
Participants were asked how confident they would feel managing the sample trauma patient. Results are seen in Table 1. Emergency registrars tended to be the most comfortable treating this patient, but this difference was not statistically significant between the three departments (P=0.16). Seniority correlated closely with confidence, with 41% of senior registrars reporting a high level of confidence treating similar patients compared with only 9% of junior registrars (P <0.01).
The fluids of choice when resuscitating the major trauma patient with no evidence of head injury, assuming that the patient had been given less than 1 litre of fluid en route to hospital, are shown in Table 2. Crystalloids (mostly saline) were chosen by 85% of registrars. Colloids were chosen only by intensive care registrars, a difference which was statistically significant (P <0.01).
The fluids of choice for a similar patient but with an additional head injury are shown in Table 3. Three junior emergency registrars changed their choice from a crystalloid to a colloid, although this change was not statistically significant (P=0.24). The reasons given for this change were to increase the plasma oncotic pressure in an attempt to decrease cerebral oedema.
Of surveyed doctors, 85% said they would infuse their resuscitation fluid of choice at a rate of 1 litre stat, while 15% chose to give 500 ml stat and then reassess.
Participants were then asked which physiological parameter they would find most useful when assessing whether this same trauma patient was fluid replete or not. The most popular parameter was heart rate (29%) followed by mean arterial pressure (24%), systolic blood pressure (21%), central venous pressure if available (14%) and urine output (9%).
A targeted question to determine whether participants were familiar with the current guidelines (2) for fluid resuscitation in trauma then asked at what point they would give a unit of universal donor blood if the haemoglobin was unknown and the patient was still unstable. The majority (62%) were familiar with this guideline and correctly responded after 2 litres, although 21% would choose to wait until after 3 litres, 7.5% would choose to transfuse after only 1 litre, and 7.5% would choose to transfuse only if the patient was visibly bleeding. There was a significant difference between registrars from intensive care and from the other two departments (although there was no difference with seniority), with the former preferring variable points to transfuse and the latter transfusing almost universally after 2 litres of fluid (P <0.01).
Participants were then asked at what level of haemoglobin (Hb in g/l) they would transfuse a major trauma patient. The question was divided into four categories and the responses are seen in Table 4. Transfusion was selected more frequently in the older age group (P=0.02), as most participants would choose to transfuse at a higher Hb level in this group, irrespective of whether the patient was bleeding or not. Similarly, transfusion was more popular at higher levels if the patient was actively bleeding, irrespective of age group (P <0.01). Thus, 64% of participants would transfuse a bleeding patient across all age groups once the Hb fell below 90 g/l, while only 7.5% of participants would transfuse the non-bleeding patient at this same level (with the other participants transfusing these patients at lower Hb levels).
Participants were asked to comment, based on their previous experiences with similar major trauma patients, on how often they would see acidosis on the first arterial blood gas analysis. Responses are shown in Table 5. Anaesthesia and intensive care registrars reported seeing a higher percentage of trauma patients with an initial acidosis compared with the observations made by emergency registrars (P=0.02). Only 56% of registrars overall would alter their fluid resuscitation based on the finding of a pH of 7.25 alone.
Finally, participants were asked when they felt that optimal fluid resuscitation had first usually been achieved in the major trauma patient. Very few respondents (<5%) thought that these patients were adequately resuscitated pre hospital (i.e. by paramedics), 48% nominated the emergency department, 26% the intensive care unit and 21% the operating theatre. Anaesthesia and emergency registrars generally selected their own department as the place where optimal resuscitation took place, whereas the responses of intensive care registrars were more variable.
Our survey of 102 critical care registrars achieved a 65% response rate, which was pleasing as it provided a considerable body of information concerning the preferences of front-line medical staff for fluid resuscitation in major trauma patients. Detailed comparison of the responses with currently recommended guidelines permitted us to reflect on the effectiveness of existing teaching programs and to point to areas where specifically targeted education might be more appropriate. It was reassuring to note that most responses from the three surveyed departments (anaesthesia, intensive care and emergency) reflected familiarity with currently accepted guidelines (1-3). Senior registrars in particular would be most comfortable treating the sample major trauma patient compared with junior registrars (41% vs 9%), although the majority of registrars would be comfortable to some degree (78% total), a result which may reflect the large number of trauma patients admitted to our centre.
There has been continued controversy over the years regarding the use of crystalloids versus colloids as a primary resuscitation fluid. Most studies have shown no significant differences in overall outcome (1,4), except in trauma patients where the use of crystalloids, which replenish both intravascular and interstitial deficits, has been associated with improved survival, particularly in head-injured patients (5,6). Most doctors in our survey would appropriately choose a crystalloid, usually normal saline, as their preferred resuscitation fluid for the trauma patient. Some guidelines suggest Hartmann's solution in preference to normal saline in the setting of major trauma, as it may avoid hyperchloraemic acidosis and thus cause less exacerbation of any preexisting acidosis from compromised tissue perfusion and may cause less physiological derangement (1,7). However, this suggestion was not reflected in clinical preferences in our hospital.
Most registrars would give one litre of fluid immediately to a major trauma patient before reassessment. Aggressive fluid resuscitation, however, should be avoided, particularly in the setting of uncontrolled internal haemorrhage, in favour of immediate transfer to the operating theatre (8,9). In addition to potentially worsening haemorrhage and haemodilution, large volume crystalloid resuscitation can have other adverse outcomes, such as cerebral oedema, pulmonary oedema, abdominal compartment syndrome and death (1,10,11). Theories that delayed resuscitation (permissive hypotension) is associated with improved survival may be due either to excessive fluid resuscitation attempts delaying time to the operating theatre and control of haemorrhage, or to previous use of rapid infusion systems which were found to be associated with increased mortality (11).
Colloids are claimed to achieve a better restoration of intravascular volume, although they are associated with poorer outcomes in patients with head injury (4-6). In this subset of patients in our survey, most registrars appropriately would choose a crystalloid. Some changed their responses from crystalloids to colloids in this scenario, in spite of evidence in the Australian SAFE study which determined that these fluids are associated with a poorer outcome (5,6). However, only a few junior registrars were responsible for this practice blemish. Several registrars from anaesthesia and intensive care changed their preference from Hartmann's solution to normal saline in the head-injured patient in an attempt to achieve higher plasma sodium levels, although only one opted for hypertonic saline which is likely to be more useful for this purpose.
Current clinical guidelines suggest that a systolic blood pressure <90 mmHg and/or heart rate >130 beats per minute are indicators of shock (1,3). However, both indices can be misleading, because young patients tend to maintain their blood pressure even when there is ongoing haemorrhage, and some other patients paradoxically become bradycardic following massive blood loss (1). Systolic blood pressure targets should be higher (>100 mmHg) if there is a suspected head injury in order to maintain cerebral perfusion pressure (3). Haemoglobin, base deficit, lactate, urine output and central venous pressure can be used as goal-directed measures of effective resuscitation. Central venous pressure can be useful when differentiating between hypovolaemic and cardiogenic shock, although this index is usually not readily available in the early resuscitation period. Participants were asked which of these indices they would find helpful when deciding whether or not their patient had been adequately fluid resuscitated. The wide spectrum of responses is likely to reflect the fact that many of these indices are examined simultaneously and that no single factor is indicative of adequate filling. Nonetheless, heart rate and mean arterial pressure were the most popular indices, probably because they are the most readily available. No participant chose jugular venous pressure, as this is likely to be more subjective and difficult to assess (often in part due to the likely presence of a cervical collar). On the other hand, many chose central venous pressure if available, as this may help guide filling by discriminating between hypovolaemic and cardiogenic causes of haemodynamic instability, as discussed in more detail below.
The Advanced Trauma Life Support guidelines suggest an initial two litres of crystalloid infusion, followed by blood transfusion and blood products if there is ongoing haemodynamic instability (2). Most registrars seemed familiar with this guideline, although intensive care registrars in particular demonstrated a greater variety of responses. This may reflect the tendency for them to work in a more monitored environment, where transfusion decisions can be made on a more individual patient basis.
Massive blood loss in severe trauma results in a hypocoagulable state, primarily due to the dilution and consumption of platelets and clotting factors. Current guidelines (3) suggest using a 1:1:1 ratio of transfused packed red blood cells, fresh-frozen plasma and platelets. There are differing opinions on optimal limits for blood transfusion, with some studies suggesting more restrictive targets (Hb <70 g/l) in young, euvolaemic patients, but a clinically-based decision is considered more appropriate in the shocked trauma patient (1). As may be anticipated, the majority of participants would choose to transfuse the older trauma patient at a considerably higher haemoglobin level than the younger patient. Lower levels would be tolerated in the younger patient, although evidence of active bleeding was a stronger predictor of early transfusion than was age. Most felt that a Hb <90 g/l would be appropriate to transfuse the bleeding trauma patient, but a Hb <80 g/l would be considered a more appropriate transfusion threshold if the patient was not bleeding or even a Hb <70 g/l if the patient was also less than 60 years old. It is recognised that these guidelines are approximate only, and in practice, patients are treated on an individual basis. Furthermore, it can be difficult to establish in some cases whether active bleeding is present, and transfusion may be ordered at higher haemoglobin levels to treat possible ongoing losses. In this regard, transfusions are also routinely ordered as part of current guidelines (2) for ongoing instability, irrespective of documented continuing blood loss or haemoglobin level.
A recent US guideline (1) recommends performing arterial blood gas analysis as part of initial assessment of the severe trauma patient and, if the base deficit is >6, then inserting a central venous line in the emergency department. The central venous pressure can then be used to guide further interventions, as described above. A raised central venous pressure (>15 mmHg) is suggestive of cardiogenic shock (e.g. pericardial tamponade, tension pneumothorax, myocardial injury or dysfunction) and a low central venous pressure (<10 mmHg) is suggestive of hypovolaemic shock, usually from acute ongoing blood loss. This guideline suggests that acidosis itself may be used as an indirect marker of inadequate oxygenation at a tissue level in these patients (usually due to hypovolaemia) necessitating more invasive monitoring, although this did not translate to a clinically useful predictor for many registrars in our centre. Indeed, many registrars did not feel that acidosis would be a helpful predictor of fluid resuscitation requirements in the major trauma patient. The higher proportion of anaesthesia and intensive care registrars who reported that they frequently encountered an initial acidosis amongst these patients probably reflects the fact that the more critically unwell subset of major trauma patients is more likely to be in operating theatres and intensive care units, whereas the wider spectrum, including those who are more stable, is seen in the emergency department.
Our study was conducted in a single tertiary institution, was of a relatively small sample size and was a snapshot at the time of the survey. Participation was voluntary, and thus the responses necessarily did not include all critical care registrars. Importantly too, it was not an observational audit of actual clinical practice but instead a survey of practice preferences under a set of structured scenarios.
We surveyed 66 critical care registrars in a tertiary institution to determine their views on fluid resuscitation of the major trauma patient. We found general consistency across all three departments of anaesthesia, intensive care and emergency, with appropriate responses concerning the type and volume of fluids to use and the end-points of fluid administration. However, there was a minority of responses that reflected lack of familiarity with current clinical guidelines or in some cases lack of experience. We believe that this type of survey can provide a useful indication of likely clinical practice in an important area, and in particular that it can suggest specific areas for targeted education. This then could be an efficient tool for optimising clinical guideline implementation, especially given the limitations of other methods of introducing and promoting current best-practice guidelines.
We thank Professors Jack Cade (Intensive Care), Kate Leslie (Anaesthesia) and Russell Gruen (Trauma Surgery), all of the Royal Melbourne Hospital for their advice and encouragement, as well as the Royal Melbourne Hospital Trauma Data Registry. We particularly thank the many registrars who gave their time to participate in this survey.
(1.) Moore FA, McKinley BA, Moore EE. Guidelines for shock resuscitation. J Trauma 2006; 61:82-89.
(2.) American College of Surgeons, American College of Surgeons Committee on Trauma. Advanced Trauma Life Support for Doctors. ATLS Student Course Manual Eighth Year 2008, Chicago, Illinois, USA.
(3.) Chiara O, Bucci L, Sara A, Bassi G, Vesconi S. Quality and quantity of volume replacement in trauma patients. Minerva Anestesiol 2008; 74:303-306.
(4.) Perel P, Roberts I, Pearson M. Colloids versus crystalloids resuscitation in critically ill patients (intervention review). The Cochrane Library, July 2009.
(5.) SAFE Study Investigators. A comparison of albumin and saline for fluid resuscitation in the intensive care unit. N Engl J Med 2004; 350:2247-2256.
(6.) SAFE Study Investigators. Saline or albumin for fluid resuscitation in patients with traumatic brain injury. N Engl J Med 2007; 357:874-884.
(7.) Healey MA, Davis RE, Liu FC, Loomis WH, Hoyt DB. Lactated Ringer's is superior to normal saline in a model of massive haemorrhage and resuscitation. J Trauma 1998; 45:894-899.
(8.) Pepe PE, Dutton RP, Fowler RL. Pre-operative resuscitation of the trauma patient. Curr Opin Anaesthesiol 2008; 21:216-221.
(9.) Pepe PE, Mosesso VN, Falk JL. Prehospital fluid resuscitation of the patient with major trauma. Prehosp Emerg Care 2002; 6:81-91.
(10.) Revell M, Greaves I, Porter K. Endpoints for fluid resuscitation in haemorrhagic shock. J Trauma 2003; 54:S63-67.
(11.) Hambly PR, Dutton RP. Excess mortality associated with the use of a rapid infusion system at a level 1 trauma centre. Resuscitation 1996; 31:127-133.
J. A. CADE *, M. TRUESDALE [(dagger)]
Departments of Anaesthesia and Emergency, The Royal Melbourne Hospital, Melbourne, Victoria, Australia
* M.B., B.S., B.Med.Sc., Registrar, Anaesthesia.
([dagger]) M.B., B.S., F.A.C.E.M., Grad. Dip. Health Services Management, Director Emergency Services.
Address for correspondence: Dr J. Cade, Department of Anaesthesia, Royal Melbourne Hospital, Parkville, Vic. 3050. Email: email@example.com
Accepted for publication on October 24, 2010.
TABLE 1 Confidence looking after major trauma patients (%) Very confident Somewhat Not confident confident Anaesthesia 10 60 30 Senior 10 30 10 Junior 0 30 20 Intensive care 21 58 21 Senior 21 26 5 Junior 0 32 16 Emergency 32 46 22 Senior 24 24 3 Junior 8 22 19 Total 26 52 23 TABLE 2 Fluid preferences for a non-head injured trauma patient (%) Normal saline Hartmann's Gelatin Blood Anaesthesia ([dagger]) 60 40 -- -- Senior 40 10 -- -- Junior 20 30 -- -- Intensive care 53 21 10 5 Senior 26 5 5 5 Junior 26 16 5 -- Emergency 78 8 -- 14 * Senior 38 3 -- 11 Junior 40 5 -- 3 Total 68 17 3 9 Albumin Anaesthesia ([dagger]) -- Senior -- Junior -- Intensive care 10A Senior 10 Junior -- Emergency -- Senior -- Junior -- Total 3 ([dagger]) Anaesthesia registrars chose only crystalloids. [DELTA] Intensive care registrars were more willing to use a colloid early (20% selecting gelatin or albumin, specifically senior registrars), and this was statistically significant (P <0.01). * Emergency registrars showed a preference for the early use of blood as a resuscitation fluid, particularly senior registrars, but this was not statistically significant (P=0.16). TABLE 3 Fluid preferences for a head-injured trauma patient (%) Normal saline Hartmann's Gelatin Blood Anaesthesia 70 20 -- -- Senior 50 -- -- -- Junior 20 20 -- -- Intensive care 58 16 10 5 Senior 32 -- 5 5 Junior 26 16 5 -- Emergency 73 5 8 14 Senior 38 3 -- 11 Junior 35 3 8 3 Total 68 11 7.5 9 Albumin Other Anaesthesia -- 10 * Senior -- -- Junior -- 10 Intensive care 10 -- Senior 10 -- Junior -- -- Emergency -- -- Senior -- -- Junior -- -- Total 3 1.5 * 3% saline. TABLE 4 Transfusion preferences by number of participants Hb <70 g/l Hb <80 g/l Hb <90 g/l Patient >60 y, bleeding 1 15 50 Patient >60 y, 23 37 6 not bleeding Patient <60 y, bleeding 4 28 34 Patient <60 y, 45 17 4 not bleeding TABLE 5 Likelihood of acidosis on first arterial blood gas analysis (%) 10-50% of <10% of patients patients >50% of patients Anaesthesia -- 60 40 Intensive care 5 53 42 Emergency 19 65 16 Total 12 61 27
|Printer friendly Cite/link Email Feedback|
|Author:||Cade, J.A.; Truesdale, M.|
|Publication:||Anaesthesia and Intensive Care|
|Date:||Mar 1, 2011|
|Previous Article:||Analysis of human cultured myotubes responses mediated by ryanodine receptor 1.|
|Next Article:||Scientific publications in international anaesthesiology journals: a 10-year survey.|