Series Editor, Jonathan E. Sevransky, MD, MHS

The Highs and Lows of Blood Pressure: Toward Meaningful Clinical Targets in Patients With Shock Sheldon A. Magder, MD

Objective: Measurement of blood pressure is fundamental for the management of patients in shock, yet the physiological basis and meaning of blood pressure measurements are complex and often not well understood. This article is in two parts: part 1 deals with the mechanical and physiological aspects of blood pressure and its measurement and part 2 deals with the role of changes in regional resistances in the determination of tissue perfusion and bedside approaches to management of shock. Data Source and Selection: This review is based on physiological principles from texts and experimental studies which elucidate some of the key principles. The views expressed are the author's synthesis of the views of others and his own opinions. Data Synthesis: Arterial blood pressure is a major determinant of regional flow and is often used as a surrogate indicator of tissue perfusion, but in reality, it is a poor indicator of blood flow. Blood pressure is determined by cardiac output (total flow) and total vascular resistance. Distribution of flow for a given blood pressure is dependent on the relative values of resistances in difterent vascular beds. If this distribution of resistances were known, this would be the ideal guide to therapy. Unfortunately, regional resistances cannot be assessed in the clinical setting, and so we are left with blood pressure as a guide to therapy. Conclusions: This article discusses the implications of these points and explores factors that need to be taken into account when designing empiric trials to determine appropriate blood pressure targets for patients in shock. Even if well-studied empirically developed guidelines become available, it likely still will be important to individualize patient management and approaches for this are discussed too. {Crit Care Med 2014; 42:1241-1251)

Department of Critioal Care, MoGill University Health Centre, Montreal, QC, Canada. Dr. Magder oonsulted for end-point adjudioation oommittees for Bristol Myer Squib and Boeringer Engleheim, provided expert testimony for various law suits (plaintiff and defendant), received grant support from the Canadian Institute of Health research, and lectured for and received support for development of educational presentations from Hospira. For information regarding this artiole. E-mail: sheldon.magder@muhc. mcgill.ca Copyright ® 2014 by the Society of Critical Care Medicine and Lippincott Williams & Wilkins DOI:10.1097/CCM.0000000000000324

Critical Care Medicine

Key Words: arterial resistance; cardiac output; critical closing pressures; pressure-flow relationships; shock; vasopressor therapy

PART I: PRINCIPLES OF MEASUREMENT AND BASIC PHYSIOLOGY OF ARTERIAL PRESSURE Arterial blood pressure measurement is one of the most basic clinical acts performed by the healthcare workers, and minimal values are routinely prescribed for critically ill patients (1). Yet, the subtleties, determinants, and physiological role of blood pressure measurement often are not appreciated. Guidelines (2) and review articles (3, 4) give recommendations for treatment when arterial pressure is below a minimum value, but these recommendations are based on little data. I will state from the outset that I am not able to provide an appropriate target pressure that should trigger a clinical intervention and I suspect that a specific pressure may not be the best choice. Rather decisions likely need to be based on constellations of physiological variables. However, clinicians still need some guiding principles for situations in which monitoring is limited to just a blood pressure measurement and there is a need to respond rapidly in a severely ill patient. Development of proper targets for blood pressure in patients with shock will require well-controlled and adequately powered empiric studies as has occurred for the management of high blood pressure, although it is also worth noting that current guidelines for hypertension did not come quickly. Initially targets were higher than current recommendations because there were concerns that lowering arterial pressure too aggressively could cause harm. Only careful long-term empiric studies eventually showed the efficacy and safety of current lower targets. The current targets for management of hypertension could not have been predicted from first principles (5). There is almost a complete lack of these kinds of controlled empiric studies for the low end targets of blood pressure, but the same process needs to occur. Furthermore, it is likely that minimal blood pressure targets will not be the same in sepsis, hypovolemia, or cardiogenic shock because of differences in their pathophysiology. In this article, I will attempt to lay the groundwork for the physiological principles that should be considered when planning empiric studies on blood pressure targets in shock. I will www.ccmjournal.org

1241

Magder

try to give clinicians approaches for the management of individual patients before well-developed empiric data become available. A number of reviews have addressed some of these issues (3, 4, 6), but this article will emphasize the importance of factors that determine arterial pressure, especially vascular resistance and its distribution among organs. What Is Blood Pressure? Blood pressure is the force that distends the elastic walls of vessels. Physical measurements, including vascular pressure, are referenced relative to some value. Since our bodies are surrounded by atmospheric pressure, measurement of blood pressure (blood pressure is used as a simplification of arterial blood pressure) is presented as a deviation from atmospheric pressure, which is defined as the "zero" value. This means that a blood pressure of 120/80 mm Hg at sea level with an atmospheric pressure of approximately 760 mm Hg is actually 880/840 mm Hg. This actual value is obviously cumbersome and furthermore fluctuates with weather conditions. Blood pressure is thus normalized to atmospheric pressure which is the outside pressure for all regions of the body except structures in the chest. However, this simplification leads to important errors in the assessment of wall stress by the Laplace relationship, which says that wall stress is equal to the product of the pressure inside a vessel and the radius of the lumen of the vessel (7). The Laplace relationship only is valid for very thin-walled structures. When wall thickness is greater than a

very small fraction of the radius, calculated wall stress actually has a negative value instead of the positive value obtained with the simplified Laplace approach (7, 8). Three types of energies determine measured vascular pressure: elastic, kinetic, and gravitational (9). In a supine subject, the gravitational component is small, but gravitational energy becomes a major factor when in the upright position. Not only are hydrostatic pressures measured relative to atmospheric pressure, but pressures also are relative to the "level" of the measuring device when pressure is measured with afiuidfilled system. This is because the fluid column in the tubing has a mass and adds a nontrivial gravitational energy component to the measurement. The consensus position for the reference level for placement of the measuring device is the midpoint of the right atrium for that is where the blood comes back to the heart and is pumped out again. This can be estimated at the bedside at a point which is at a vertical distance of 5 cm below the sternal angle, which is where the second rib meets the sternum. This is valid at least up to a body angle of 60% from the horizontal (10). The actual pressure distending vessels, however, is very different in the head and the foot in an upright person. In a 180-cm person (~ 71 inches), the feet are about 120 cm below the heart and the top of the head is about 60 cm above the heart (Fig. 1). With a blood pressure at the level of the heart of 100 mm Hg, and density of blood of approximately 1 g/cm^ the actual pressure distending the arterial vessels in the foot is 183 mm Hg and at the top of the head it is 51 mm Hg.

Arterial

Arterial Pressure

95 100 Venous Pressure

.1: 5

:i 2

183

Figure 1. Gravitational effect on arterial and venous pressures. The numbers on the right in mm Hg refer to the gravitational potential energy due to the difference in height of the measuring device relative to the midpoint of the right atrium {dashed line) and assuming a 182-cm man. The dotted lines Indicate the loss of pressure due to resistance (5 mm Hg for the arterial circuit and 2 mm Hg for the venous). The other numbers refer to approximate pressure measured relative to the different positions of the leveling device and taking into account the resistance loss in pressure. Adapted from Burton (9).

1242

w\AA^'.ccmjournal.org

These values would be obtained if the transducer was leveled relative to the foot or head, respectively. It is worth noting that the added pressure produced by gravity in the foot of a standing person is close to the arterial pressure generated by the heart in a supine individual. Measured intravascular pressure is also affected by the position and type of cannula used. Vessel walls are distended by elastic pressure, and this pressure is most accurately measured by a hole on the side of the cannula. If an end-hole catheter is used, and the hole faces the flow as in a standard arterial catheter, kinetic energy in the moving blood is lost as it hits the cannula. The kinetic energy then is converted into elastic energy so that the measured pressure is higher than the pressure measured with a side-hole catheter. Normally, kinetic energy only contributes about 4 mm Hg (3%) to peak May 2014 • Volume 42 • Number 5

Concise Definitive Review

arterial pressure and 0.35 mm Hg to mean arterial pressure ( 11 ), but the contribution from kinetic energy increases when pressure is low as in the vena cava and pulmonary systems, for although they have low elastic pressures, their mean velocity of blood flow is the same as the aorta. In hyperdynamic sepsis, the combination of increased blood velocity and low arterial pressure likely results in a greater contribution of kinetic energy to total energy whieh means that blood fiow can be the same with a lower elastie pressure as measured by a euff Under normal flow eonditions, pulse pressure is amplified when measured further away for the aorta although there is a slight deerease in mean pressure. On the other hand when there is marked vasoeonstrietion, a more peripheral pressure sueh as radial arterial pressure, may me be lower than femoral arterial pressure beeause there is a greater resistanee drop due the eonstriction of the vessel. The end-whole femoral arterial pressure also has a greater kinetie eomponent for the veloeity is likely larger due to the smaller overall eross-seetional area at that level of the eirculation. The message is that measured pressures will vary with site and hemodynamic state of the patient. Recommendations for pressure targets need to consider the site and method of measurement. Trends in pressure and the relationship to metabolic measures will likely be the most important indicators. What Determines the Blood Pressure? A worthwhile intellectual exercise is to consider whether arterial pressure determines cardiac output or whether cardiac output determines the arterial pressure. Intuitively people most often have a model in their minds in which the heart produces a force, that is pressure, which drives the blood around the circulation, and this view is still argued by some (12). However, it is actually the volume per time that the heart forces through the systemic vascular resistance that creates the arterial pressure (13). The pulsatile component of arterial pressure is obviously related to the eyelie eontraetion of the heart, but the determinants of the systolie and diastolie pressure are more eomplex. When the stroke volume is pumped into the elastie aorta, the aortie wall is stretched and some of the volume is transiently taken up by the aorta and then released during the rest of the eyele in what is known as the "Windkessel effect." Peak systolic pressure is thus determined by the amount of volume ejected by the heart per beat, that is, stroke volume, the elastic properties of the aortic wall, the initial volume in the aorta at the start of cardiac ejection, and the rate of outflow from the aorta. Because of phase shifts between flow and pressure, the frequency of cardiac contractions also affects the pressure. Peak systolic pressure is further complicated by reflected waves which occur when the forward wave hits bifurcations such as branching at the ihae arteries. The diastolie pressure is more dependent on the runoff from the aorta and large vessels which in turn are dependent on peripheral resistance, but it is also affected by the duration of the cardiac cycle and the initial volume in the aorta because aortic eomplianee is eurvilinear. Paradoxically, although blood pressure does not determine cardiac output, arterial pressure Critical Care Medicine

is an important determinant of regional flows. However, this should not be confused with blood pressure being an indieator of regional flows. Why Is Arterial Pressure So High? An important prineiple of mammalian eardiovaseular physiology is that the vascular system is pressure regulated, which means that blood pressure normally is kept in a narrow range. Even during heavy aerobic exercise, systolic arterial pressure rises by less than 50% in healthy subjects. The pressure in most mammals tends to be regulated around 120/80 mm Hg. This relatively high value (at least compared with pulmonary arterial pressure) certainly cannot be because this pressure is required to move the blood through the body, for the right heart pumps the same amount of blood as the left heart through the lungs with a mean pressure of less than 20 mm Hg. Another thought might be that this high pressure is needed to overcome the gravitational loss of energy to the head and to maintain cerebral perfusion. However, blood pressures in rats and mice are similar to that of humans, but they do not have the same gravitational demands. The likely reasons for high arterial pressure in mammals are related to selective advantages for cardiac performance and the distribution of now. Cardiac muscle handles volume work (i.e., muscle shortening) much more efficiently than pressure work. Thus, by keeping the arterial pressure relatively constant, the heart works against a relatively constant load. This design component can be appreeiated when eonsidering what is required to inerease flow. Eor example, during maximal exereise in a standard size male, muselé bloodflowcan inerease from around 2 to 3 L/min at rest to greater than 20 L/ min. If systemie vaseular resistanee started very low, and resting arterial pressure was around 20 mm Hg,fiowto the working muselé eould not seleetively increase much more by a local deerease in resistance for blood pressure would be even lower. Blood flow would have to increase to all regions and total flow blood flow would have to be much higher. A solution for this would be to eonstrict all nonworking regions, but surely from a design point of view, it is much easier to dilate the area that needs flow rather than eonstrict all the regions of the body that do not need as much flow. Eurthermore, dilatation can occur through local metabolic mechanisms and does not require innervations of vessels, although the presence of neural inputs increases the effieieney of the system. Thus, the arterial system works mueh like a reservoir that supplies water to a community. The hydrostatic pressure in the tank is kept relatively constant by the height of the tank. Opening taps, which effectively decreases local resistances, allows water to flow to individual homes. However, although the arterial pressure is produced by the product of total cardiac output and total systemic vascular resistances of all regions of the body, regional flows are determined by the arterial pressure and their individual vaseular resistances relative to other regions of the body. Thus, the distribution of local arterial resistanees is the major determinant of where blood goes as will be diseussed in the next seetion.

www.ccmjournal.org

1243

Magder

Cardiosenic shock

60 Pressure

B

90

(inmHg)

60 Pressure

90 (nrniHg)

Figure 2. Hypothetical regional pressure-flow relationships during eardiogenic shock and the response to a vasopressor. The regional flows are based on values from ret, (12), A, The x-sas shows arterial pressure and they-axis flow. The slope of the lines is conductance, the inverseof resistance (1/R), The baseline condition isat (1), The line at (2) shows the effect of a drop in cardiac output and arterial pressure to 60 mm Hg without any reflex adjustment B, This shows what would happen if systemic vascular resistance (SVR) increased by a similar amount in all regions. At position (3), the pressure is restored to 90 mm Hg, but since there is no change in cardiac output, flow in each region remains at the level it was at 60 mm Hg, The dotted line indicates the muscle flow before and after the vasoconstrictor.

radius of the tube. Thus, vessel radius is the primary determinant of the resistance, and small changes in vessel radius produce large changes is vascular resistance. If the resistance to all vascular beds were the same, the flow would be the same in every region, but each vascular bed has its own resistance characteristics, which allowsflowto vary among organs (Fig. 2), and this resistance can vary according to need for flow by changes in the radii of vessels (14). However, elastin and collagen of vessel walls set limits to how much vessels can dilate (15). Fully dilated vessels function as rigid pipes and flow becomes linearly related to the pressure difference between the inflow and outflow pressure. Maximal flow in each region

What Is Resistance? In an idealflowingfluid,which is called a "Newtonianfluid,"layers form in theflowingcolumn because of the frictional loss of energy due to contact of the fluid with vessel walls and between the fluid layers. The potential of fluid layers to slide over each other is called "viscosity." This frictional loss of energy produces the pressure drop along the course of blood vessels which we refer to as the "resistance" to flow of the fluid. As determined by Poiseuille in the 19th century, flow is proportional to the length of the tube, proportional to the viscosity of the fluid, and most importantly, inversely proportional to the fourth power of the

is determined by the maximal inherent cross-sectional area of the vasculature in that bed. The vascular density and thus crosssectional area of the vessels in the heart is much larger than that of skeletal muscle, and consequently, maximal coronary blood flow per gram of tissue at a given pressure is two to three times that in skeletal muscle (Fig. 3). The kidney, too, has a very large maximal flow per mass compared with other organs (Fig. 3). There are some basic starting "rules" that help predict responses to challenges to the system under normal conditions. Regional flows are proportional to metabolic need so that for the whole body, and in individual organs, there is a tight linear relationship between cardiac output and oxygen consumpB A tion (16). The relationship is 500 especially strong in muscle 1500 Heart tissues but not very strong in - 180 b/min the kidney. It is important to I Muscle appreciate that it is total flow, Splanchnic that is, cardiac output, which is en g kidney o regulated in the body and not o E Kidney stroke volume (16). During Exercising exercise, heart rate increases Muscle Brain in proportion to the relao -.0 tive demands of the workload UHeart compared to the capacity of Splanchnic; the system (16). Since venous Muscle ' 90 (mmHg) return and thus cardiac out90 (mmHg) put during exercise are conPressure Pressure trolled by the total metabolic need of the body, and heart Figure 3. A, Hypothetical baseline regional pressure-flow relationships as shown in Figure 1, B, The same relationships normalized to the weight of each organ. When normalized to weight, the slope of the muscle pressurerate is controlled by the relaflow line is smail; it increases markedly with exercise. The figure shows a theoretical pressure-flow of the heart at tive workload, stroke volume a heart rate of 70 and 180 beats/min. The heart has the highest flow capacity per weight of tissue of the major becomes a dependent variable. vascular beds and has a marked capacity to increase vascular conductance (decrease in resistance). 1244

vv\AAv.ccmjournal. org

May 2014 • Volume 42 • Number 5

Concise Definitive Review

Neural inputs that increase heart rate without major metabolic changes, such as occur with anxiety or chronotropic drugs, decrease stroke volume without much change in cardiac output. Baseline distribution offlowis set by structural and neurohumeral mechanisms which maintain the arterial pressure at a relatively constant level with changes in blood flow. Arterial baroreceptors are the primary sensor for this process. The baroreceptor response to changes in arterial pressure gives a good example of the hierarchy of the system. In severe hypotension, there is a disproportionately greater increase in the arterial resistance in muscle vasculature compared with that of the splanchnic circulation (17). This can be seen as an attempt to protect the metabolically active abdominal organs because the relatively greater increase in vascular resistance in muscle shifts a greater fraction of the limited blood flow to the splanchnic bed. This likely occurs due to differences in adrenergic receptor density. The implication is that actions of vasoactive drugs also will not be the same in all regions. Distribution offlowto different vascular regions can overcome this hierarchy through three mechanisms: local metabolic activity, myogenic responses, and now-mediated dilatation. Metabolic activity releases substances that vasodilate vascular smooth muscle. These include adenosine, potassium ions, increased osmolarity, low Po^, Pco^, lactate, and prostaglandins, but which one dominates is not known and they likely work together with varying sensitivities in different tissues (18-21). Exercise provides a good example of the interaction of local metabolic and neurogenic factors. During exercise, the drop in systemic vascular resistance and activation of peripheral afferent nerves (22, 23) lead to increased sympathetic tone and generalized constriction of vascular beds, but local metabolic factors in the working muscle and the heart counteract this constricting effect, and there is net vasodilation in these regions. This redirects flow to these working regions. Certain regions of the body such as the brain and kidney do not have large changes in metabolic need but still depend on a relatively constant blood flow for normal function. A rise in pressure above normal would produce excessive cerebral blood flow, increase intracranial blood volume, and raise intracranial pressure. Too low pressure would jeopardize brain function. Furthermore, a rise in pressure could distend vascular walls which would decrease vascular resistance and result in even greaterflow.This feed-forward process is prevented by changes in vascular tone in response to changes in arterial pressure in what is called the myogenic mechanism (24-27). This process is especially strong in the brain and kidney and keeps flow relatively constant over a wide range of pressures. The third local regulator of regional flow is called "flow-mediated dilatation." An increase in flow increases a force on the endothelium that is called "shear stress." This increases basal release of nitric oxide from endothelial cells which dilates vascular smooth muscle (28, 29). The process has the advantage of allowing upstream flow to better match downstream needs, but it also produces a feed-forward process which is kept in check through counter regulation by myogenic and metabolic mechanisms. Critical Care Medicine

Local regulatory mechanisms are lost in sepsis, ischemia-reperfusion, and may be abnormal in patients with metabolic syndrome and vascular disease associated with endothelial dysfunction. This makes it very difñcult to predict how vessels wül respond to pharmacological agents and creates a major limitation for the use of a simple blood pressure measurement to guide therapy that is intended to increase regional flow. Calculation of resistance with Poiseuille law is based on the difference between the inflow and outflow pressures. It would seem obvious that the outflow pressure in the vascular tree should be the lowest pressure in the system, which is the central venous pressure. However, this is not the case. There is flow limitation at the arteriolar level, which produces an effective critical closing pressure or "vascular waterfall" (30,31) (Fig. 4). When there is flow limitation, flow is determined by the difference between the arterial and critical closing pressures and not the final downstream pressure. The implications of this are very significant for the interpretation of systemic vascular resistance. The mean critical closing pressure for the whole body is estimated to be in the range of 25-30mm Hg (32) but varies among tissues, and values as high as 60-70 mm Hg have been found in the resting hindlimb of dogs (31, 33) (Fig. 4). The standard calculation of systemic vascular resistance is based on the difference between arterial and central venous pressure instead of the critical closing pressure because the critical closing pressure cannot be clinically measured. Use of this incorrect downstream pressure produces an error that gets progressively larger the lower the inflow pressure because the error makes up a larger fraction of total pressure. As a result, there is an "apparent" decrease in resistance with increases in pressure or flow. It is physiologically plausible that the arterial resistance falls with an increase in pressure for this would be expected from baroreceptor-mediated regulation of arterial pressure, but the standard calculated resistance will fall whether or not there is a change in the true resistance (Fig. 5). This artifact only can be ruled out by obtaining 2 points on the pressure-flow relationship, but this cannot readily be obtained in intact persons. The consequence of this is that a drug that effectively increases cardiac output, such as the phosphodiesterase inhibitor milrinone, will appear to do so by reducing vascular resistance and afterload, but these cannot be separated from a change in contractility without 2 points on a pressure-flow line. The presence of critical closing pressures at the arteriolar level means that arterial pressure can be regulated in two ways: one is by a change in the actual resistance, which is the inverse of the slope of the pressure-flow relationship, and the other is by changes in critical closing pressures, which is the x-intercept. If pressure rises due to an increase in resistance, a given change in flow will produce a greater change in pressure, but if the rise in pressure is due to a change in critical closing pressure, a subsequent change in flow will produce the same change in pressure as occurred before the increase in the critical closing pressure. Of historical interest, critical closing pressures are also called "Starling resistors" because Ernest Starling used floppy collapsible tubes to produce a critical closing pressure in the circuit of his heart-lung preparation and was thereby www.ccmjournal.org

1245

Magder

I a » n m œ m m m m m

250-i

2005 Hz

150-

Control

100-

PART 2: THERAPEUTIC IMPLICATIONS

5025

75 100 125 150 175 200

Figure 4. Pressure-flow relationships obtained in the hindlimb of dogs at rest, with 5-Hz stimulation to simulate exercise and after a transient occlusion to produce reactive hyperemia. This summary analysis (n = 35) takes into account the compliance effects during the decrease in pressure. At rest, the flov^i became zero at a pressure indicating a critical closing pressure of almost 70 mm Hg. Data points from one animal are shown In the insert. With muscle stimulation, there was an increase in slope (increase in conductance) and a decrease in the critical closing pressure. Following arterial occlusion, there was a marked increase in the slope and further fall in critical closing pressure, and these effects were even more marked during muscle stimulation. The vertical line indicates the marked increase in flow that occurred at the same arterial pressure with these adaptations. The flow at the starting pressure would have gone from 25 to 250mL/min. Reproduced with permission from García-Cardeña et al (29).

Resi Stance

-low

Part

Pv

4

s

Jé

Pressure

Pcrit

Pressure (or Flow)

Figure 5. Error in calculated resistance produced by ignoring arterial critical closing pressure (Pcrit). True resistance is the slope of the pressure-flow line from Part to Pcrit. When Pv is used as the downstream value for calculating resistance (1/slope), the resistance decreases with decreases in pressure (or flow) as shown on the right side. Part = arterial pressure, Pv = venous pressure.

1246

able to maintain a constant pressure against which the heart contracted. The critical closing pressure in our vasculatures likely serves a similar function. Disease states can alter critical closing pressure, and this adds a further level of complexity to the system (33-37). The clinical implication of the intrinsic error in the standard calculation of vascular resistance is that changes in calculated resistance have little meaning and one should instead examine the directional changes in arterial pressure and cardiac output.

wvvw.ccmjournal.org

Further Exploration of Vascular Resistances As discussed in ?an 1: Principles of Measurement and Basi Physiology of Arterial Pressure regional blood flows are determined by the arterial pressure and resistance in the region, whereas arterial pressure itself is determined by the total vascular resistance and cardiac output. A decrease in arterial pressure without changes in local resistances or critical closing pressures only can occur when there is a decrease in cardiac output (Fig. 2). If the distribution of vascular resistances does change, the flow in each region will change in proportion to the baseline resistance in each region. Similarly, if all resistances are increased by a vasoconstricting drug in such a way that the distribution of vascular resistances does not change, arterial pressure will increase, but blood flow will remain the same in all regions (Fig. 1). Thus, vasopressor use is based on the assumption that resistance wül increase less in vital organs, such as the brain, kidney, and heart, because local mechanisms will reduce the increase in their regional May 2 0 1 4 • Volume 42 • Number 5

Concise Definitive Review

resistance. This assumption is less likely to be true when high doses of vasopressors are used and regulatory mechanism no longer function normally. Another key implication is that it is unlikely that vasopressor therapy will be helpful if there is not an accompanying increase in cardiac output to provide an increase in overall fiow to be redistributed. An exception likely is loss of normal a-adrenergic tone, with for example spinal shock, for the vasopressor may restore the normal distribution pattern produced by sympathetic tone (38). It is important to differentiate between blood flow in units of volume per minute versus blood flow in units of volume per minute normalized to body weight (Fig. 3). Because of its large mass and cross-sectional area, total conductance (inverse of resistance) in skeletal muscle is greater than conductance of the kidney and heart so that a drop in cardiac output produces a much greater loss of blood fiow in muscles than in the heart and kidney. However, when fiow is normalized to tissue mass, the loss of fiow to each gram of tissue is much greater in the kidney and the heart for these tissues have a greater density of vessel per mass. When plotted this way, the kidney has the steepest pressure-flow line and thus the greatest loss of flow per mass in the body, which explains why it is such a sensitive marker of a decrease overall perfusion. Furthermore, the capacity of renal blood flow to increase above the resting level is smaller than that in the heart and splanchnic region which means that the kidney has limited potential to adapt to changes in arterial pressure (Fig. 6). The heart can increase its vascular conductance more than five-fold from the value at resting heart rate. As a dramatic example of this, we were able to make measurements of coronary blood flow in dogs in which we maximally vasodilated the vasculature even at mean arterial pressures of 20-30 mm Hg (Fig. 6).

Another key point is that if blood pressure falls because of a decrease in vascular resistance, and all vascular resistances decrease in proportion, blood flow to all regions can be sustained at very low pressures (Fig. 7). This is why some people, more commonly women, can have systolic pressures of 90 mm Hg or less and still have normally functioning organs. Their regional vascular resistances are all set at a lower level but with the same proportions among vascular beds as persons with higher baseline arterial pressures. Thus, a key piece of information for setting targets for blood pressure management is knowledge of the proportion of resistances among vascular beds and how they change with diseases and therapy. Unfortunately, it is not possible to obtain this information in humans, especially in the clinical setting. Furthermore, measurement of microvascular flow in a single region is unlikely to be helpful in and of itself and multiple regions likely will need to be assessed and their characteristics compared with the vascular characteristics of vital organs, or at least, it must be established that the region being studied is representative of what is happening in vital organs.

Goals of Therapy The primary role of the circulation is to provide adequate bulk flow to deliver energy substrates and to remove waste. As such, it is tissue perfusion that counts and even though it is well appreciated that blood pressure is notflow,in practice blood pressure ends up being used as a surrogate forflow.However, pressure is a poor guide to actual blood flow for at a given arterial pressure, regionalflowis highly dependent on the local vascular resistance and outflow pressures which are regulated both locally and centrally. Consider an extreme example, if someone made a device that spontaneously clamps the thoracic aorta when pressure is too low, it would produce a significant rise in upper body pressure, but the flow to the lower extremities would be zero. Pressure thus produced would be what is termed a "tangible benefit" in that clinicians see something that is comforting, but the rise in pressure is not necessarily producing the desired outcome, which is increased overall tissue perfusion (39). Pressure is easy to obtain, but emphasis on pressure can be misleading without some assessment of overall flow. From the physiological point of view, the distribution of regional 10 20 30 40 50 60 70 80 120 180 10 20 30 40 50 60 7080 120 180 resistances is likely the most Arterial Pressure (mmHg) Arterial Pressure (mmHg) important variable for determining outcome of therapy, but Figure 6. Pressure-flovii lines for kidney and heart obtained in dogs under baseline condition and after maxithis value cannot be obtained mal dilatation of the vasculature with nitroprusside. Pressure-flow relations were obtained by hemorrhaging the animals, and blood flow was measured with radio-labeled microspheres. There is much less reserve in the renal practically and we are left with vasculature and in the baseline condition; kidney blood flow falls with first decrease in pressure. The reserves pressure being a limited but are much larger in the heart because the basal coronary flow is lower than that in the kidney. Note that meaessential tool. surements were still obtainable at very low arterial pressures. Reproduced with permission from Magder (15). Critical Care Medicine

vi/vvw.ccmjournal.org

1247

Magder

deereases with long diastoles, but the longer diastole also means that there is more time for eoronary perfusion and the lower I Splanchnic heart rate means that the metabolie needs of the heart are lower. Kidney Similarly, a low diastolie pressure with a high cardiac output indicates that systemic vascular resistance is low and at least some Brain regions must be getting high Muscle flows. It is hard to know which regions will loose when attempts are made to raise this pressure. Diastolie pressure is also set by Heart the arteriolar critieal elosing pressure. If this value deereases, the same flow ean oeeur with a 90 (mmHg) 50 90 (mmHg) lower diastolie pressure. I personPressure ally find the diastolie pressure to Pressure be the least useful value. Figure 7. Theoretical pressure-flow lines af rest and with vasodilation. The left side is the same as Figure 1. On It would seem that the mean the right side, the resistances have all been decreased to maintain the same proportion of flow in all regions. In this pressure is the best compromise. example, the pressure could be decreased by half, and flow would nof be compromised in any organ. The mean is the value that has The question then becomes what aspects of the pressure, most often been used in assessing autoregulation of renal blood that is, systolie, diastolie or mean, best relate to tissue fiow? flow (42). However, it should be appreciated that this is partly Sinee pressure-flow relations, elosing pressures, and regulatory driven by practical experimental reasons. In animal studies, processes vary markedly among the different organs, there can flow is often controlled by a pump without pulsation; thus, be no simple answer. A particularly complex situation arises there is only mean fiow. There also is mueh less variability in when cardiac output is high and the pressure is low as is typical the mean than in systolie pressure, and the mean is not afteeted in sepsis (40). In this situation, the pressure clearly is not an by the frequeney response of the measuring deviee (41). This indicator of cardiac output for the output is high. However, makes it easier to deteet differenees among groups in a study. cardiac output, too, is a poor indicator of microvascular perfuHowever, mean pressure is not obtained in early resuseitation sion for the high flow likely is a result of peripheral shunt-like studies when the patient does have an arterial line and a sphygactivity. To make things more complicated, when vascular tone momanometer is being used. One should also appreeiate that normalizes with the resolution of sepsis, eardiae output falls newer automated systems aetually obtain a mean pressure and as regional perfusion improves. However, the integrated piederive the arterial and diastolie pressures and are subjeet to ture of improving vaseular responsiveness, deereasing eardiae artifaets. Because of this, I recommend that the values on these output, and resolving metabolic dysfunction likely can provide systems always be checked with a sphygmomanometer, which guidance that vasopressors can be decreased. bring us back to systolic pressure! Systolic pressure generally provides a good overall assessAn interesting example of the arbitrariness of clinical evalment of cardiac performance and is the most readily available uation oeeurs in patients who have an intra-aortie balloon number because of the ease of obtaining it with a sphygmopump. I titrate vasopressor drugs in these patients to the augmanometer. However, in cardiogenic shock, it is not uncom- mented diastolie pressure although the mean arterial pressure mon to observe patients with normal or even elevated arterial is used by many. My rational is that the balloon pump is being pressures and very low cardiac outputs. Systolic pressure can used to lower two values, the systolie arterial pressure and the be followed easily with indwelling catheters, but there are some pressure at end diastole to unload the ventrieles. On the other important limitations. It is essential to ensure that the signal hand, the balloon is being used to inerease the rest of diastolie is not over or underdamped. This is determined by observing pressure to improve coronary perfusion pressure. If the mean the quality of the signal on the monitor, and it can be checked pressure is targeted, this increases the load on the left venby auscultation or palpation and a sphygmomanometer (41). tricle, which defeats the whole purpose of the balloon pump. Furthermore, the target for the augmented diastolie pressure DiastoKc pressure is particularly important for coronary perdoes not need to be as high as usual systolic pressures for the fusion because perfusion of the lefi ventricle is primarily during augmented diastolie pressures are still mueh higher than usual diastole. However, the value of the diastolie pressure whieh is reeorded is the last and lowest value and does not give a true indi- diastolie pressures. As interesting example of fallaeious reasoning oeeurs when the balloon pump triggers on every seeond eation of the overall diastolie perfusion foree. Diastolie pressure B

1248

www.ccmjournal.org

May 2014 • Volume 42 • Number 5

Concise Definitive Review

rather than every beat (1:2). This usually results in a higher augmented diastolic pressure than occurs with one to one augmentation because of the simple reason that when augmentation is 1:2, balloon inflation occurs after a beat in which the systolic pressure was not lowered by a prior balloon deflation. Since it starts higher, it will end higher but only on every second beat. Accordingly nurses and physicians often feel more comfortable lowering the dose of vasopressors because they see this higher pressure value on the console even though the patient is clearly getting half the support! Vasoconstrictors Pharmacological actions of vasoconstrictors in intact organisms are very complex (43 ). Consider the actions of commonly used norepinephrine. In sepsis, norepinephrine can decrease vascular capacitance and thereby raise mean systemic fllling pressure and either decrease or not change venous resistance (44). Together these changes increase venous return. Norepinephrine also increases cardiac function, which coupled with the increase in venous return increases cardiac output. It increases vascular resistance which coupled with the increase in cardiac output can lead to better tissue perfusion and potentially better distributions of blood flow by restoring vascular tone to regions that were inappropriately vasodüated (44,45). However, these positive effects depend on the baseline characteristics of the system. If the heart is functioning on the flat part of the cardiac function curve and the norepinephrine does not increase cardiac function, cardiac output will not increase. Furthermore, if systemic vascular resistance rises substantially, the increase in afterload could decrease cardiac output unless matched by an increase in inotropic activity or heart rate. If there are no reserves in unstressed volume, there is no volume to recruit by a decrease in vascular capacitance, and this too will limit the increase in venous return and potential for an increase in cardiac output. If resistance in small veins increases more than the arteriolar resistance, there will be an increase in capillary nitration, which will lower mean systemic ñlling pressure and decrease venous return and cardiac output. This discussion does not even address the regional differences in the drug's pharmacological effects, the variability with doses, and the role of the intrinsic disease. With regard to the latter point, the pathophysiology of eardiogenic and hypovolemic shock is fundamentally different from that of septic shock. In eardiogenic shock, systemic vascular resistance is high and likely so is venous resistance; if the cardiac dysfunction is structural due to a valve disorder or ischémie myocardium, there is less likelihood of an improvement in cardiac output. It is thus not surprising that cardiologists have had great fear of using norepinephrine (46). In septic patients, the problem is more one of a loss of vascular tone, and it is not uncommon to see improvements in organ function such as renal function with restoration of tone with norepinephrine (47,48). Patients with ischemia-reperfusion may lie in the middle with an initial "cardiogenic-like" picture and then a subsequent vasodilatory state secondary to endothelial and vascular smooth muscle dysfunction following restoration of flow. The proper clinical Critical Care Medicine

blood pressure targets for the use of vasoactive drugs must thus take into consideration the nature and stage of the pathophysiological process and the responses of the components of the circulatory system (43). It is for this reason that I argue that only the accumulation of empiric data will make it possible to come up with reasonable recommendations for the blood pressure targets and the use of vasoactive drugs, and even then, there will be no substitute for careful bedside observation and monitoring, which should likely include some measure of blood flow. Vasopressor therapy is very often ordered to maintain a sufficient pressure to maintain renal function. This is not difficult in a patient with urine output and functioning kidneys. If the urine output goes up with the vasopressor therapy and creatinine does not, then clearly the therapeutic goal has been achieved and the therapy should be continued and vasopressors titrated to a level necessary to maintain urine output. In this simple condition, the pressure itself should not be the determinant for continuous management but may be of help in testing responsiveness. If the drug raises renal perfusion pressure but also constricts renal arteries in an equal proportion to vessels in other vascular beds, the rise in pressure will be of no help to the kidney for renal blood flow will not increase. Furthermore, when the kidney faus, it would seem that it no longer makes sense to target therapy to the needs of the kidney, for clearly other factors besides perfusion pressure are determining renal function, and maintaining the arterial pressure at a higher level may be detrimental to other organs. Back to the Bedside and Individual Patients Given the lack of adequate data for general guidelines, how can physicians at the bedside determine that a treatment is helpful for the patient or harmful? One must start by remembering that the objective of increasing arterial pressure is to increase blood flow to at least some tissues, and if that does not happen, the rise in pressure is of no value. In a patient who is warm, awake, communicating, and urinating, the task is easy for the flow must be adequate for these functions to be normal no matter how low the blood pressure. I recently had a most extreme example in my outpatient clinic. A patient that I have followed for more than 5 years walked into (and out of.) my clinic with a systolic pressure of 68 mm Hg in both arms on repeated measurements. This man had clearly managed to regulate the distribution of flow so that normal organ function was maintained. However, the situation is usually more complex in most critically ill patients. One of the first clinical guides to be lost is the patient's state of wakefulness due to direct neurotoxicity from inflammatory cytokines or sedation for intubation or other procedures. If raising the arterial pressure improves wakefulness, then CNS responsiveness in that person can be an excellent indicator of the adequacy of tissue perfusion. However, more often, the decrease in wakefulness does not readily respond initially to pressure changes. If the kidneys are still functioning and there is good urine output, it is very likely that the pressure is adequate since kidneys are so sensitive to decreases in perfusion. It also has been well documented that vwvw,ccmjournal,org

1249

Magder

vasopressors can increase urine flow in patients with sepsis, and thus, urine output is a simple measure to follow. Unfortunately, a large proportion of critically ill patients have impaired renal function so that urine output is not a useful guide. The next tool that likely is helpful is blood lactate in a central vessel. The value itself may not be as useful as much as the trend in values (49, 50). If lactate rises as the pressure is increased by vasopressors, it seems unlikely that increasing vasopressors further will be helpful. At this point, some measure of flow is needed. This could be a surrogate of flow such as central venous oxygen saturation (51) or a direct measure of flow by a device that measures flow perhaps even at the microcirculatory level (52). The question to be asked is does the vasopressor-induced increase in blood pressure also increase cardiac output and reduce lactate? If not, another therapy should be tried for it is unlikely that tissue perfusion has been improved. An exception is a rise in lactate from "washout" of lactate due to improved flow. This can be identified by improving overall function. The reverse also can be used. The dose of vasopressor could be decreased while observing serum lactate and indicators of flow. If flow and metabolic indicators do not decrease, the lower blood pressure value is likely tolerable.

CONCLUSION Although blood pressure measurement has a primary place in the management of tissue perfusion, this discussion should indicate that there are many limitations to the use of blood pressure by itself. Blood pressure measurements always need to be put into the context of a patient's overall status. This can be as simple as observing that the patient's overall function is normal or in complex situations by obtaining measurements of blood flow and metabolic indicators. Guidelines for targets for initiating therapy need to be developed from properly designed prospective empiric trials. However, given the complexity of the physiology, the astute clinician also needs to consider changes in measured variables in response to therapy in individual patients and use these responses to titrate the doses of potentially helpful but also potentially harmful vasopressor therapy for no single value of blood pressure will fit all patients' needs.

REFERENCES 1. Lamonfagne F, Cook DJ, Adhikari NK, et al: Vasopressor adminisfrafion and sepsis: A survey of Canadian intensivists. 7 Crit Care 2011 ; 26:532.e1-532.e7 2. Dellinger RP, Levy MM, Rhodes A, et al; Surviving Sepsis Campaign Guidelines Committee including the Pédiatrie Subgroup: Surviving sepsis campaign: International guidelines for management of severe sepsis and sepfic shock: 2012. Crit Care Med 2013; 41:580-637 3. Levy B, Collin S, Sennoun N, ef al: Vascular hyporesponsiveness to vasopressors in septio shook: From bench fo bedside. Intensive Care /Wed 2010:36:2019-2029 4. Augusto JF, Teboul JL, Radermacher P, ef al: Interprefafion of blood pressure signal: Physiological bases, clinical relevance, and objectives during shock sfafes. Intensive Care /Wed 2011; 37:411-419 5. Rosendorff C, Black HR, Cannon CP, ef al; American Heart Associafion Council for High Blood Pressure Research; American Heart Assooiafion Council on Clinical Cardiology; American Hearf Association Council on Epidemiology and Prevention: Treatment of

1250

/.ccmjournai.org

hypertension in the prevention and management of ischémie hearf disease: A scientific stafemenf from the American Heart Associafion Council for High Blood Pressure Research and the Councils on Clinical Cardiology and Epidemiology and Prevention. Circulation 2007; 115:2761-2788 6. McGhee BH, Bridges EJ: Monitoring arterial blood pressure: What you may not know. Crit Care Nurse 2002; 22:60-64, 66 7. Shrier I: Critical Closing Pressures, Vascular Waterfalls, and the Control of Blood Flow fo the Hindlimb. Montreal, Canada, Thesis for Faculty of Graduate Studies and Research McGill University, 1993 8. Azuma T, Oka S: Mechanical equilibrium of blood vessel walls. Am J Physion97^; 221:1310-1318 9. Burton AC: Total fluid energy, gravitational potential energy, effects of posture. In: Physiology and Biophysics of the Circulation: An Introductory Text. Chicago, IL, Year Book Medical Publishers Incorporated, 1965, pp 95-111 10. Bickley LS, Hoekelman RA (Eds): Bates Guide to Physical Examination and History Taking. Seventh Edition. Philadelphia, PA, Lippincotf, 1999 11. Burton AC: Kinetic energy in the circulation. In: Physiology and Biophysics of the Circulation: An Introductory Text. Chicago, IL, Year Book Medical Publishers Incorporated, 1965, pp 102-112 12. Brengelmann GL: Counterpoint: The classical Guyton view that mean systemic pressure, righf atrial pressure, and venous resistance govern venous return is not correct. J AppI Physioi (1985) 2006; 101:1525-1526 13. Magder S: Point: The classical Guyfon view that mean systemic pressure, righf atrial pressure, and venous resistance govern venous return is/is not correct. J AppI Physiol (1985) 2006; 101:1523-1525 14. Ross J Jr: Dynamics of the peripheral circulation. In: Best and Taylor's Physiological Basis of Medical Practice. West. J (Ed). Eleventh Edition. London/Baltimore, MD, Williams and Wilkins, 1985, pp 119-131 15. Magder S: Pressure-flow relations of diaphragm and vital organs wifh nitroprusside-induced vasodilafafion. J AppI Physiol (1985) 1986; 61:409-416 1 6. Magder SA: The ups and downs of heart rate. Crit Care Med 201 2; 40:239-245 1 7 Deschamps A, Magder S: Baroreflex control of regional capacifance and blood flow disfribution with or wifhout alpha-adrenergic blockade. Am J Physiol 1992; 263:H1755-H1763 18. Duling BR: Microvascular responses to alterations in oxygen tension. Circ Res 1 972; 31:481 -489 19. Berne RM: Regulation of coronary blood flow. Physioi Rev 1964; 44:1-29 20. Berne RM: Metabolic regulation of blood flow. Circ Res 1964; 15(Suppl):261-268 21. Guyton AC, Carrier O Jr, Walker JR: Evidence for fissue oxygen demand as the major factor causing autoregulation. Circ Res 1 964; 15(Suppl):60-69 22. Mitchell JH, Shephard JT: Control of the circulation during exercise. In: Exercise-The Physiological Challenge. McNeil Hill P (Ed). Auckland, NZ, Conference Pub, 1 993, pp 55-85 23. McCloskey Dl, Mitchell JH: Reflex cardiovascular and respirafory responses originating in exercising muscle. J Physiol 1972; 224:173-186 24. Johnson PC: Autoregulation of blood flow. C;>c Res 1986; 59:483-495 25. Sfanden NB, Quayle JM, Davies NW, et al: Hyperpolarizing vasodilators activate ATP-sensitive K-l- channels in arterial smooth muscle. Science 1989; 245:177-180 26. Bayliss WM: On the local reactions of the arterial wall to changes of internal pressure. J Physiol 1902; 28:220-231 27 Johansson B, Mellander S: Sfafic and dynamic components in fhe vascular myogenic response fo passive changes in lengfh as revealed by elecfrical and mechanical recordings from fhe rat portal vein. Circ Res 1975; 36:76-83 28. Dimmeler S, Assmus B, Hermann C, et al: Fluid shear stress stimulates phosphorylation of Akt in human endothelial cells: Involvement in suppression of apoptosis. Circ Res 1998; 83:334-341 May 2014 • Volume 42 • Number 5

Concise Definitive Review 29. García-Cardeña G, Fan R, Shah V, et al: Dynamic activation of endothelial nitric oxide synthase by Hsp90. Nature 1998; 392:821-824 30. Permutt S, Riley RL: Hemodynamics of collapsible vessels with tone: The vascular waterfall. J AppI Physiol 1963; 18:924-932 31. Magder S: Starling resistor versus compliance. Which explains the zero-flow pressure of a dynamic arterial pressure-flow relation? Circ Res 1990; 67:209-220 32. Sylvester JL, Traystman RJ, Permutt S: Effects of hypoxia on the closing pressure of the canine systemic arterial circulation. Circ Res 1981; 49:980-987 33. Shrier I, Hussain SNA, Magder SA: Carotid sinus stimulation influences both arterial resistance and critical closing pressure of the isolated hindlimb vascular bed. Am J P/7ys/o/1993; 33:H1560-H1566 34. Shrier I, Magder S: Response of arterial resistance and critioal pressure to changes in perfusion pressure in canine hindlimb. Am J Physiol 1993; 265:H1939-H1945 35. Shrier I, Baratz A, Magder S: Effects of adenosine on pressure-flow relationships in an in vitro model of compartment syndrome. J AppI Physion 997; 82:755-759 36. Shrier I, Magder S: NG-nitro-L-arginine and phenylephrine have similar effects on the vascular waterfall in the canine hindlimb. J AppI Physiol (1985) 1995; 78:478-482 37 Shrier I, Magder S: Effects of nifedipine on vascular waterfall and arterial resistance in canine hindlimb. Am J Physiol 1995; 268:H371-H376 38. Thiele RH, Nemergut EC, Lynch C 3rd: The clinical implications of isolated alpha(i) adrenergic stimulation. Anesth Anaig 2011; 113:297-304 39. Magder S: Phenylephrine and tangible bias. Anesth AnaIg 2011 ; 113:211-213 40. Bersten AD, Hersch M, Cheung H, et al: The effect of various sympathomimetics on the regional circulations in hyperdynamic sepsis. Surgery 1992; 11 2:549-561

Critical Care Medicine

41. Gardner RM: Direct blood pressure measurement-dynamic response requirements. Anesthesiology ^98^ ; 54:227-236 42. Bersten AD, Holt AW: Vasoactive drugs and the importance of renal perfusion pressure. New Horiz 1995; 3:650-661 43. Rudis Ml, Basha MA, Zarowitz BJ: Is it time to reposition vasopressors and inotropes in sepsis? Crit Care Med 1 996; 24:525-537 44. Datta P, Magder S: Hemodynamic response to norepinephrine with and without inhibition of nitric oxide synthase in porcine endotoxemia. Am J Respir Crit Care Med 1 999; 160:1 987-1993 45. Martin C, Viviand X, Leone M, et al: Effect of norepinephrine on the outcome of septic shock. Crit Care Med 2000; 28:2758-2765 46. Nasraway SA: Norepinephrine: No more "leave 'em dead"? Crit Care Med 2000; 28:3096-3098 47. Desjars P, Pinaud M, Bugnon D, et al: Norepinephrine therapy has no deleterious renal effects in human septic shock. Crit Care Med 1989; 17:426-429 48. Schaer GL, Fink MP, Parrillo JE: Norepinephrine alone versus norepinephrine plus low-dose dopamine: Enhanced renal blood flow with combination pressor therapy. Crit Care Med 1985; 13:492-496 49. Jansen TC, van Bommel J, Schoonderbeek FJ, et al; LACTATE study group: Early lactate-guided therapy in intensive care unit patients: A multicenter, open-label, randomized controlled trial. Am J Respir Crit Care Med 2010; 182:752-761 50. Jansen TC, van Bommel J, Woodward R, et al: Association between blood lactate levels. Sequential Crgan Failure Assessment subscores, and 28-day mortality during early and late intensive care unit stay: A retrospective observational study. Crit Care Med 2009; 37:2369-2374 51. Rivers E, Nguyen B, Havstad S, et al; Early Goal-Direoted Therapy Collaborative Group: Early goal-directed therapy in the treatment of severe sepsis and septic shock. A/Eng/J Med 2001; 345:1368-1377 52. Magder S: Bench-to-bedside review: An approach to hemodynamic monitoring-Guyton at the bedside. Crit Care 201 2; 16:236

www.ccmjournal.org

1251

Copyright of Critical Care Medicine is the property of Lippincott Williams & Wilkins and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.