Brad Gander is an Associate Practitioner and Student Paramedic working in Brighton. Brad has an interest in critical care and wrote the following blog piece as part of our #CritHaemWeek. Tweet Brad @BG_999 and let him know what you think!
Trauma is a worldwide leading cause of death in people aged 15-44 years (1). 30-40% of total mortality attributed to trauma is as a result of haemorrhage with up to 56% of these deaths occurring in the pre-hospital setting (2). A number of important decisions must be made in this period such as initial therapies and requests for additional pre- or in-hospital resources. The cornerstone of these decisions is the prompt recognition of the patient’s clinical condition, and the presence of life threatening blood loss. Traditional vital sign abnormalities associated with haemorrhagic shock include tachycardia and hypotension, such as those proposed by the Advanced Trauma Life Support (ATLS; 3) shock classifications, shown in the table below. These classifications suggest a relationship between the percentages of blood volume (BV) lost and increase in heart rate (HR), however several studies have contested the accuracy of using such vital signs as an indicator of haemorrhage.
ATLS Shock Classification (modified to include only heart rate in relation to percentage of blood volume lost)
This article will briefly discuss the physiological changes that occur during ‘simple’ haemorrhage, defined as a loss of circulating volume without significant damage to tissues (4), and evaluate several studies conducted to investigate the utility of HR as a measure of haemodynamic compromise.
Ultimately, the goal of the compensatory mechanisms that occur during BL is to deliver oxygen to the vital organs in order to maintain their function. In cases of BL, decreased circulating volume is detected by baroreceptors within the aortic arch and carotid bodies, which then reduce their inhibition of the sympathetic nervous system. In turn, the cardiovascular system, usually controlled by the parasympathetic nervous system, is subject to greater influence from the sympathetic nervous system and HR is increased. The result of this increase in HR allows for oxygen to be delivered to the organs at the same rate as it was prior to the decrease in circulating volume and subsequent decline in oxygen carrying capacity of the blood. Alongside these mechanical changes a number of cellular-level changes ensue as a response to reduced oxygen delivery, resulting in an increased blood glucose supply and an increase in the ability of the tissues to extract oxygen from the blood (5).
An article by Schultz and McConachie (6) describes the HR to simple haemorrhage as biphasic or, in extremis, triphasic:
The first phase of this response occurs when up to 30% of total blood volume is lost. As a result of the previously described compensatory mechanisms, adequate organ perfusion is maintained and the patient will have a mild tachycardia and normal arterial blood pressure.
In cases of continued blood loss, exceeding 30% of blood volume, a second phase of changes to the cardiovascular system occurs. C-fibres within the left ventricle send signals to the parasympathetic nervous system (7), creating an increased parasympathetic system output and inhibition of the sympathetic nervous system. This reduces sympathetic influence on the cardiovascular system, possibly intended as a mechanism to slow the heart rate and allow for increased diastolic filling time (4), and as a result the heart rate and arterial pressure decreases. As these changes occur oxygen delivery to the organs cannot be sustained and becomes insufficient.
The third phase often represents a pre-terminal phase of haemorrhagic shock. The patient will display a significant tachycardia, low arterial blood pressure and signs of poor vital organ perfusion.
The theory of a triphasic response is supported by a study conducted in 1986 by Sander-Jensen et al (8) on 20 patients with haemorrhagic shock. The mean BL of these patients was 36% of total volume and all were found to have a relatively ‘normal’ HR of 73 beats per min (bpm) accompanied by a mean blood pressure (BP) of 81/55 mmHg. Upon volume replacement via administration of blood and crystalloids these values returned to those expected in the first phase (112/72 mmHg and 102 bpm). After continued treatment and reversal of haemorrhagic shock the observations of these patients returned to normal clinical parameters (131/79 mmHg and 82 bpm). These findings demonstrated the second phase of haemorrhagic shock can be considered a reversible period in the patient's condition.
A similar study was conducted by Jacobsen and Secher and published in Clinical Physiology and Functional Imaging in 1992 (9). This prospective study explored the relationship between HR and mean arterial pressure (MAP) in 34 patients with haemorrhagic shock. 18 patients had a BL of less than 1.9 litres [0.9-3.0], corresponding to 34% [16-46%] of estimated blood volume. These patients had an average HR of 83 bpm and a MAP of 62. In the remaining 16 patients, whom had an average BL of 4 litres [3.3-5.0] corresponding to 89% [80-100%] of estimated total blood volume, the average HR was 120 beats/min with an average MAP of 52 mmHg. Six patients died as a result of severe bleeding with an average HR of 129 beats/min and a MAP of 40mmHg.
In a far larger trial than those discussed previously (n = 10,825), a retrospective cohort study conducted by Brasel et al (10) studied the validity of HR as a predictor of haemodynamically significant injury. The study included those with blunt or penetrating trauma and found HR lacked sensitivity and specificity in determining the need for emergent intervention (laparotomy, thoracotomy, or angiography), packed red blood cell transfusion in the first 24 hours, or severe injury (ISS >25). Tachycardia was defined as a HR of >100 bpm and had a sensitivity of 33% in blunt trauma and 37% in penetrating trauma with a specificity of 76% to 79%. In the subgroup of patients with a HR of >120 bpm, specificity was increased to 95% and 96% for blunt and penetrating trauma patients, but sensitivity remained poor at 12% and 11% respectively.
In slight contrast to the findings of the previously mentioned papers, a study by Guly et al in 2011 (11) aimed at testing the validity of ATLS shock classifications and found HR progressively increased in correlation with estimated percentage of blood loss. Patients with ‘Class I’ shock (<15% blood loss) had a median HR of 82 bpm, which increased to 95 bpm in ‘Class IV’ shock (>40% blood loss). Mean BP also decreased from 135 mmHg to 120 mmHg in the group with the most severe BL. The results of this study indicate, whilst the theory of a progressive increase in heart rate may be correct, it is not to the degree suggested by the ATLS classifications of shock.
Several limitations exist when considering the validity of these studies in pre-hospital practice. Firstly they were not conducted in the pre-hospital environment and therefore do not take into account interventions or treatment performed prior to hospital admission. All studies included patients involved in blunt and penetrating trauma; the former is often a precipitant to extensive tissue damage and thus creates different physiological changes (6). Medication details were not included within the data, therefore it was not possible to examine the effects of beta-blockade or other medications on HR. Several other factors of HR variability exist, such as age, anxiety, physical fitness, temperature and the presence of additional injuries involving the central nervous system.
The changes that occur during the second phase of haemorrhagic shock highlight an important learning point and consideration when assessing patients with suspected haemorrhage. The physiological changes that occur during this phase can produce clinical findings that do not match what is often considered the predicted response of haemodynamically compromised patients, such as progressive tachycardia.
As noted by Brasel et al (10), it is important to recognise "HR alone is not sufficient to determine the need for emergent interventions for haemorrhage". Whilst there may be a slight correlation between HR and total blood volume lost, the presence of major haemodynamic compromise in patients with a ‘normal’ HR following simple haemorrhage should not be discounted.