Pathophysiology, monitoring, and therapy of shock with organ failure 5

Applied Cardiopulmonary Pathophysiology 14: 5-15, 2010

Pathophysiology, monitoring, and therapy of shock with

organ failure
W. C. Shoemaker, M. Beez
Department of Surgery, Division of Trauma/Critical Care, Los Angeles County and Univer-
sity of Southern California, Los Angeles, CA, USA

Introduction: example, „If the B.P. goes down, give fluids;

The conventional approach to if the fluids don’t work give dopamine.“
resuscitation and management In the traditional approach „one-at-a-
of critically ill patients time“ approach, each abnormality is identi-
fied, documented and then corrected. As
Half of the two million persons in the United soon as the first defect is corrected, the next
States who die annually die acutely of circula- deficit is sought, measured, evaluated, and
tory problems, shock and its sequelae, organ then corrected; the process is repeated as
failure. Since shock is the commonest com- long as abnormalities are found. However,
plication of high risk surgery, trauma, and this may result in uncoordinated, disorgan-
acute illnesses, its early recognition and cor- ized and sometimes contradictory therapeu-
rection are key to its prevention. Thus, it is ex- tic plans.
tremely important to recognize shock as ear- The most commonly monitored variables
ly as possible and to vigorously treat it while in hospitalized patients are the mean arterial
it is still readily reversible. pressure, heart rate, central venous pressure,
In the late stage, it is easy to diagnose and cardiac output. According to the conven-
shock by hypotension, oliguria and collapse, tional therapeutic paradigm, the aim of thera-
but at that stage, therapy is often ineffective. py is to return these values to their normal
The more important question is, „How can range. In several thousand monitored values
we recognize shock in its earliest stage when of patients who died, we were able to restore
therapy is most effective?“ Unfortunately, the 76% of them to the normal range; neverthe-
conventional approach could not be worse. less, these patients still went on to die. Then,
We routinely recognize shock by subjective survivors were found to have essentially the
symptoms and imprecise signs: cold clammy same survival percentages in these variables.
skin, weak thready pulse, unstable vital signs, Obviously, something clearly is wrong with
pallor, and altered mental status. These crite- the conventional approach to monitoring and
ria are subjective, imprecise, and observer-de- treatment of acutely ill patients. First, the
pendent; they are remote from physiologic wrong variables may be monitored. Second,
evidence that could objectively characterize normal values may be appropriate for normal
nonsurvivors’patterns. healthy people, but not appropriate for criti-
The first question to ask in high risk pa- cally ill postoperative patients (1-4).
tients is, „How do you make early practical The conventional etiological approach
therapeutic decisions at the bedside?“ Tradi- characterizes hemorrhagic, cardiogenic, trau-
tionally these decisions are often inadequate matic, neurologic, and septic shock in terms
because we have developed too many simple of: a) signs and symptoms, b) laboratory find-
solutions to complex problems without con- ings, c) pathophysiology, and d) therapy. This
sidering the underlying pathophysiology; for conventional approach is widely accepted,
6 W. C. Shoemaker, M. Beez

simple, straight forward, easy to understand, Invasive monitoring of oxygen

but wrong! It is wrong for three reasons; first, transport to evaluate the
the signs and symptoms are the same for pathophysiologic pattern of
each etiology; they all have oliguria, tachycar- shock
dia, and collapse. Second, the laboratory is
not at all diagnostic. Third and worst of all, it The essential pathophysiology of the shock
has a one-dimension description of the state after high risk surgery may be described
pathophysiology. This one-dimension ap- in terms of oxygen metabolism and oxygen
proach precludes understanding of the inter- balance. The delivery of oxygen is the most
actions of the three circulatory components: clearly defined direct measurement of circula-
cardiac function, pulmonary function and tis- tory function and, more importantly, the ca-
sue perfusion. Tissue perfusion represents the pacity of the circulation to compensate for
important overall purpose of the circulation, exercise, stress, trauma, sepsis, and other
to supply body metabolism. acute illnesses. Oxygen consumption is the
most direct measurement of metabolic activi-
ty. However, it measures what the patient is
Physiologic approach to actually consuming, but not necessarily what
invasive monitoring of high the patient may need. Based on these key
variables which were found to be early out-
risk patients
come predictors. Now it may be hypothe-
The alternative is to observe surviving pa- sized that inadequate delivery of oxygen in
tient’s responses to hemorrhage, high risk sur- the face of increased metabolic demands (for
gery, trauma, and shock. It was found that example, recovery from trauma, surgical op-
survivors developed a pattern of supranormal erations, blood loss, infection, etc.) leaves the
cardiac index (CI), O2 delivery (DO2), and O2 patient with tissue hypoxia, organ dysfunc-
consumption (VO2). The hypothesis is that if tion, organ failure and death.
we prophylactically drive high risk patients to The first problem is to identify the appro-
the survival pattern with early aggressive ther- priate patients to monitor. In the U.S., most
apy that optimizes cardiac output, oxygen de- patients survive major surgery, but 1.5 to 2%
livery and consumption values, outcome may die; since there are about 35 million major
be improved. operations per years this means that over half
Recognition of shock in the earliest peri- a million patients die postoperatively every
od often depends on subjective symptoms, year. If we look at only the high risk group of
rather than precise physiologic measure- patients, about 7% of patients subjected to
ments. Because of this uncertain definition of surgery are high risk and about 20 to 30% of
shock, hemodynamic and oxygen transport them die in the immediate postoperative pe-
variables were monitored in 708 high risk, riod. High risk criteria consist of prior cardiac
surgical patients preoperatively, during sur- infarction or stroke, extensive ablative cancer
gery, and in the immediate post-operative pe- surgery, severe multiple trauma, massive
riod (5). The survivors’ values were compared hemorrhage, septic shock, respiratory failure,
with the values of those who died to describe renal failure, acute abdominal catastrophes
the early temporal patterns. The first and (4); these 7% high risk group contributed to
most dramatic differences occurred in CI, 82% of deaths on our surgical service (1).
DO2 and VO2; nonsurvivors were found to
have relatively normal values, while survivors
had markedly increased values for these vari- Prospective clinical trial
ables (1-17).
Using the survivors oxygen transport values
as goals of therapy, we studied the effects of
Pathophysiology, monitoring, and therapy of shock with organ failure 7

optimizing these values prospectively, first by trol group, if all that was done with the
allocating patients of each of three surgical catheter was to achieve normal values. How-
services to the protocol or control groups. In ever the PA-protocol group had a significant
the first five years, we found that patients reduction in the mortality from 32% to 4%
who entered the study on the second, third, compared with the normal values of the PA-
or fourth postoperative day had successively control group. Also the PA-protocol group
less and less effect from optimizing CI, DO2, had a significant 67% reduction in the venti-
and VO2. In the last year we prospectively lator days, a 30% reduction in ICU days and
performed a pre-operatively randomized con- hospital days and a 25 % reduction in the
trolled trial in the following manner. After hospital costs (1).
each high risk patient was identified preoper- Of interest was a group of high risk pa-
atively, and signed an „informed consent,“ a tients whose doctors didn’t think they were
sealed opaque envelop was opened to allo- sick enough to need invasive monitoring; this
cate the patient to one of three groups: a) the „non-randomized group” had the highest
central venous catheter, b) the Swan-Ganz mortality and highest percentage of organ
pulmonary artery (PA) catheter with normal failures; ironically, sixty percent of these had
values as the goals of therapy, and c) the PA- a PA Catheter placed postoperatively after
protocol group which had a PA catheter with they developed a life-threatening post-opera-
goals of therapy as the supranormal values tive cardiorespiratory event. However, the PA
determined empirically from a previous sur- Catheter at this time did not improve the
vivor group as CI > 4.5 L/min/m2, DO2 as > overall group mortality. The data showed that
600 mL/min/m2, and VO2 > 700 mL/min/m2 the PA catheter can prevent organ failure if
(1). oxygen transport is optimized in the first 8 to
The results of the study are shown in 12 hours, but is not able to reverse lethal or-
Table 2 and Table 5. The central venous gan failure after it occurs.
catheter was as good as the PA catheter-con-

Table 1: Normal values for invasively monitored hemodynamic variables

Variable Formula Normal Units

Value
Cardiac index CI = cardiac output/BSA 3.2 ± 0.2 L/min1 m2
Systemic vascular resistance SVRI = 79.92 X (MAP-CVP)/CI 2180 ± 210 dyne s/cm5 m2
index
Pulmonary vascular resistance PVRI = 79.92 X (MPAP - WP)/CI 270 ± 15 dyne s/cm5 m2
index
Mean transit time Direct measurement 15 ± 1.4 seconds
Central blood volume CBV = MTT X CI X 16.7 830 ± 86 mL/m2
Stroke index SI = CI/HR 46 ± 5 mL/m2
Left ventricular stroke work LVSW = SI X MAP X 0.144 56 ± 6 g m/m2
Right ventricular stroke work RVSW = SI X MPAP X .0144 8.8 ± 0.9 g m/m2
Left cardiac work LCW = CI X MAP X .0144 3.8 ± 0.4 kg m/m2
Right cardiac work RCW = CI X MPAP X .0144 0.6 ± 0.06 kg m/m2
*.0144 and 79.92 are conversion terms.
Abbreviations: BSA = body surface area; MPAP = mean pulmonary artery pressure; WP (or PAOP) = pul-
monary artery occlusion pressure (or wedge pressure); MTT = mean transit time; HR = heart rate.
8 W. C. Shoemaker, M. Beez

Table 2: Oxygen transport variables

Variable Formula Normal Unit

Value
Arterial Hgb O2 satu- Direct measurement 96 ± 1 %
ration
Mixed venous Hgb Direct measurement 75 ± 1 %
saturation
Arterial oxygen con- CaO2 = SaO2 X 1.36 X Hgb* + (.0031 X PaO2) 19 ± 1 mL/dL
tent
Mixed venous O2 CvO2 = SvO2 X 1.36 Hgb* + (.0031 X PvO2) 14 ± 1 mL/dL
content
Oxygen delivery DO2 = CI X CaO2 520 ± 16 mL/min m2
Oxygen consumption VO2 = CI (CaO2 - CvO2) 131 ± 2 mL/min m2
Oxygen extraction O2 ext=DO2/VO2 26 ± 1 %

*Hgb = hemoglobin

Hemodynamic values of the protocol and Measurement of oxygen debt

control groups were described for the preop-
erative, intraoperative, and postoperative pe- These observations lead on to a corollary hy-
riods. The cardiac index values of the proto- pothesis that states that the basic underlying
col group reached optimal values within 8 to problem in all shock is tissue hypoxia and this
12 hour postoperatively, while the control val- can be measured as the net cumulative oxy-
ues remained at the preoperative levels. gen debt. If this oxygen debt is prevented or
There were no significant differences in the corrected early on, there will be a reduction
other hemodynamic variables. The oxygen in organ failure and death. Oxygen debt can
transport values of the control group were be measured directly in surgical patients by
maintained at their preoperative normal val- measuring oxygen consumption preopera-
ues, while the DO2 and VO2 values of the tively and then, correcting for the effects of
protocol group reached their optimal goals in anesthesia and temperature, extrapolating
the first 12 hours postoperatively. There were this value to the intra-operative and immedi-
no significant differences in blood gases, ate postoperative periods. Integrating the
hematocrit, or other oxygen transport vari- area under the curve gives us the net oxygen
ables. The striking difference between the excess or debt at any specified time. Oxygen
groups was the incidence of organ failure. debt was measured in 253 consecutively
The protocol group had only one patient with monitored patients with measurements tak-
ARDS, while the control groups had signifi- en pre-operatively, intra-operatively and post-
cantly more patients with organ failures operatively (17). This is easy to document in
(Table 2). There were no significant differ- surgical patients, where pre-illness baseline
ences between the groups in the incidence of values can be obtained, and the times of the
complications not due to organ failure. That beginning and end of the operation are easi-
is, the oxygen transport optimization proto- ly determined. The patients, who survived
col did not protect against local mechanical without organ failures had the smallest oxy-
or anatomic complications such as wound in- gen debt which lasted an average of 12
fection, dehiscence, postoperative bleeding, hours. By contrast, the nonsurvivors, all of
nosocomial infections, and drug or transfu- whom died with organ failures, had a contin-
sion reactions (1-4). uing oxygen debt during the 48-hour period
Pathophysiology, monitoring, and therapy of shock with organ failure 9

of observation. Intermediate between these shows the physiologic criteria for each major
two groups were the data of survivors with decision node (20). The upper loop describes
organ failures who had oxygen debts lasting criteria for fluid therapy; the second describes
about 24 hours. In the prospective controlled criteria for inotropic therapy; the third for va-
trial described above, the protocol patients sodilators, and the lower for vasopressors.
had less oxygen debt intraoperatively, and The goals of therapy are different in vari-
the oxygen debt became an excess more rap- ous types of illness (Table 3). In postoperative
idly than did the control patients who had patients, the mean cardiac index of survivors
normal values as therapeutic goals (4). was observed to be 4.5 L/min/m2, in trauma
The common denominator in all forms of patients 5.0, in sepsis 5.5, and in cardiogenic
shock is oxygen debt and is limited by re- shock from acute myocardial infarction 2.5
duced oxygen consumption (VO2) to levels L/min/m2. These values are appreciably af-
less than the body needs. Secondly, oxygen fected by age, co-morbid clinical conditions,
debt from reduced VO2 is also the major de- degree of blood loss, length of time in shock,
terminant of outcome. The concept was de- and the reserve compensatory capacity of
scribed over 35 years ago by Guyton et al each vital organ.
(18,19) who anesthetized and bled dogs; When oxygen consumption values are
those that accumulated an oxygen debt of plotted against their corresponding value of
100 ml/kg all survived, while dogs that accu- O2 delivery, there is an initial sloping line
mulated a debt of 140 ml/kg all died; the showing increasing VO2 as the DO2 increases
50% mortality occurred at 120 ml/kg. Our sharply, indicating supply-dependant VO2.
studies on high risk surgical patients (as com- There is a critical point beyond which further
pared with hemorrhage in dogs) corroborat- increase in delivery of oxygen does not in-
ed the conclusions of Guyton (18,19). crease the oxygen consumption, indicating
supply-independent VO2 above this point.
We use the mean values of each etiologic
Therapeutic goals of invasive category of shock as a first approximation to
monitoring the true goals. Then additional fluids or in-
otropic agents are given to increase the DO2
We conclude from these studies that at least in order to determine if VO2 also increases
two-thirds of the patients who die postopera- (as in supply-dependent VO2). Optimization
tively do not die of anatomic reasons, they occurs when further increases in DO2 no
die of physiological problems that can be de- longer increase VO2 (supply-independent
scribed, predicted and prevented. The sec- VO2).
ond important conclusion is that values ob-
served in critically ill high risk patients who
survive ought to be considered as the goals Relative effectiveness of
of therapy. Third, if these goals are achieved crystalloids and colloids in
early, i.e., in the first 8 to 12 hours postoper-
patients with ARDS
atively, there will be a reduction in morbidity
and mortality. This is not only the case in the Invasive monitoring is extremely useful to
surgical patients but also in a wide range of evaluate objectively and scientifically the rel-
medical problems (6-16). ative efficacy of alternative therapies in vari-
ous clinical conditions. For example, it is said
almost as a party-line dictum that albumin
Branch chain decision tree should never be given in adult respiratory dis-
tress syndrome (ARDS) patients, because col-
A Branch Chain Decision Tree or Clinical Al- loids leak into the lung tissues through the
gorithm for management of high risk patients pulmonary capillaries and drag a lot of water
10 W. C. Shoemaker, M. Beez

Table 3: Invasive and noninvasive hemodynamic values for survivors and nonsurvivors

Variable, unit Optimal Value Survivors (N=103) Nonsurvivors (N=48) P value

Mean ± SEM Mean ± SEM
CI, L/min/m2 4.0 4.14 ± 0.02 3.57 ± 0.03 0.001
MAP, mmHg 85 88 ± 0.37 83 ± 0.69 0.08
SapO2, % 98 99 ± 0.05 96 ± 0.26 0.001
O2 delivery 600 624 ± 39 520 ± 47 0.01
O2 consumption 140 170 ± 76 127 ± 67 0.01
PtcO2/FiO2, torr 200 206 ± 2.9 93 ± 2.6 0.001
CI cardiac index, MAP mean arterial pressure, SapO2 arterial hemoglobin saturation by pulse oximetry, Pt-
cO2/FiO2 transcutaneous oxygen tension indexed to FiO2

into the lungs with it. However, this concept 10% Dextran-40, 1000 ml Ringer’s lactate,
had not been experimentally verified and, 500 ml of whole blood (l unit), or 500 ml of
therefore, we tested it by a prospective, ran- packed red blood cells according to appropri-
dom-order, cross-over designed study of pa- ate indications. Data have shown that col-
tients in early ARDS (defined as within 24 to loids increase cardiac index, but RL did not.
48 hours after the diagnosis was established) The colloids, whole blood and packed red
given 1.0 liters of Ringer’s lactate (RL) and cells all improved DO2 and VO2, but Ringer’s
100 mL of 25% albumin (ALB); each agent lactate (RL) did not significantly change either
was given to each patient, half of the patients of these variables. However, in the late stage
receiving RL first and then crossed over to of ARDS and septic shock, capillary leak does
ALB, the other half receiving ALB first and occur, and there were minimal improvements
then RL. These data show that 1000 ml of RL in response to fluid interventions.
increased blood volume only 200 mL at the When the changes in VO2 plotted against
end of the infusion, which was the maximum the corresponding changes in DO2, there are
volume effect, while 100 ml of 25% ALB in- increases in both DO2 and VO2 after the col-
creased blood volume an average of 450 mL. loids and blood, but not after crystalloids.
In essence, ALB didn’t leak, rather it dragged Since O2 can not be stored, increases in DO2
350 mL of interstitial water back into the plas- that increase VO2 indicate that there had
ma volume and with this increase in plasma been preexisting oxygen debts which are par-
volume there was increased cardiac index, tially restored
colloidal osmotic pressure, and O2 consump-
tion without worsening the P(A-a)O2 gradient
or pulmonary shunt. Albumin as well as Future directions for
starch significantly improved both DO2 and noninvasive monitoring of
VO2, while RL only transiently increased DO2
and actually decreased VO2 probably be-
acutely ill emergency and
cause of increases in the diffusion pathway critically ill patients
and the diffusion time.
We subsequently expanded this series to Finally, we would like to briefly outline the fu-
over 400 studies in 212 patients in early ture direction of both critical care and emer-
ARDS (24 to 48 hours after diagnostic criteria gency care in noninvasive circulatory moni-
were met) given 500 ml of 5% plasma protein toring because these are easier, cheaper and
fraction, 100 ml 25% albumin, 500 ml of more feasible to use. Although previous
Pathophysiology, monitoring, and therapy of shock with organ failure 11

bioimpedance instruments have been unreli- can be used anywhere in the hospital includ-
able, new high tech hardware and software ing the emergency department, prehospital
innovations from defense industries have area or physician’s offices. Irrespective of the
greatly improved bioimpedance technology. precipitating event in shock, the cardiac, pul-
These newer technologies provide more reli- monary and tissue perfusion functions should
able data that allow us to titrate therapy more be evaluated in terms of the primary prob-
closely according to physiological criteria. lem, pathophysiology, compensation mecha-
We compared the new bioimpedance nisms, decompensations, and therapy (Table
method (Noninvasive Medical Technologies, 4). This allows a more comprehensive ap-
Auburn MI) to the standard thermodilution proach to acute circulatory problems irre-
technique in 2192 paired measurements; spective of the precipitating etiological event,
there was a close correlation (r=0.86). These ie, cardiogenic, hemorrhagic, traumatic, sep-
data demonstrated satisfactory agreement tic, or post-surgical. This offers an integrated
between the two methods and the ability of physiologic approach for treatment of the
this bioimpedance method to track and trend critically ill patient.
thermodilution changes. The temporal patterns of noninvasive cir-
Most importantly, the new impedance culatory variables of the survivors and non-
cardiac output system may be combined with survivors were described in a series of 139
other noninvasive monitoring systems includ- blunt trauma patients beginning with the ini-
ing pulse oximetry to assess pulmonary func- tial measurements after admission to the ED.
tion and transcutaneous oximetry to assess The nonsurvivors’ mean arterial pressure
tissue oxygenation. These noninvasive moni- (MAP) was higher than normal and higher
tors evaluate the interactions of the three cir- than those of the survivors in the first hour af-
culatory components to identify, diagnose, ter admission consistent with greater
and recommend treatment in early stages adrenomedullary stress responses. The non-
when circulatory problems are easily re- survivors’ MAP fell abruptly about the 3rd or
versible. 4th hour after admission. Cardiac index val-
ues were initially higher in the survivors indi-
cating lesser blood volume deficits or better
Noninvasive monitoring of physiologic compensations. Nonsurvivors’
emergency patients SapO2 were significantly lower than the sur-
vivors’ values. The nonsurvivors’ transcuta-
The advantages of these noninvasive monitor- neous oxygen tensions indexed to the FiO2,
ing systems are that they provide continuous PtcO2/FiO2, were markedly lower than the
on-line real-time display of cardiac function, survivors’ values and lower than normal
pulmonary function, and tissue perfusion that throughout the observation period.

Table 4: Mean net cumulative deficits or excesses of monitored values of survivors and
nonsurvivors throughout the period of observation

Variable SURVIVORS NONSURVIVORS P-Value

Mean SEM Mean SEM
CI, L/m 2 +81 52 -232 138 0.007
MAP, mmHg.h -10 13 -57 24 0.078
SapO2, %.h -1 0.3 -8 2.6 0.006
PtcO2/FiO2, torr.h +313 87 -793 175 0.001
12 W. C. Shoemaker, M. Beez

Table 5: Outcome of various groups treated to normal and supranormal goals of PA (1). (W.
Shoemaker et al. Prospective trial of supranormal values of survivors as therapeutic goals
in high-risk surgical patients).

Groups Nonrandomized CVP-Control PA-control PA-protocol

No.patients (N=45) (N=30) (N=30) (N=28)
Years 56.9 ± 2.5 55.2 ± 3.0 53.4 ± 2.5 56.4 ± 3.1
Males/females (%) 45/55 64/36 39/61 75/25
Hospital days 21.9 ± 1.7 22.2 ± 2.8 25.2 ± 3.4 19.3 ± 2.4
ICU days 14.0 ± 1.7 11.5 ± 1.7 15.9 ± 3.1 10.2 ± 1.6*
Ventilator days 6.5 ± 1.3 4.6 ± 1.4 9.4 ± 3.4 2.3 ± 0.5*
Intraop.death 0 0 1 0
Postop.deaths No.(%) 17 (23%) 7 (23%) 10 (33%) 1 (4%) +

Net cumulative amount of but recent studies showed no advantage of

deficit (or excess) of monitored the PA catheter in cardiac and other medical
variables and outcome conditions nor in postoperative patients ad-
prediction mitted to the ICU after organ failures had de-
veloped (17-22). Two recent consensus con-
The circulatory measurements were used to ferences found insufficient evidence to deter-
quantify the amount of deficits in cardiac, mine whether PA catheter-guided therapy sig-
pulmonary, and tissue perfusion functions in nificantly alters outcome; they did not consid-
a series of 151 high risk emergency patients er time factors, but mixed early and late stud-
on admission to the ED; the net cumulative ies together, and arrived at ambiguous con-
deficit or excess of each monitored variable clusions (22,23). However, an incisive meta-
of each patient was calculated as the area be- analysis by Boyd and Bennett (24) found sig-
tween the continuously monitored values nificant outcome improvement in seven
and the normal or optimal values. Table 5 lists prospectively randomized studies when PA
differences in the net cumulative deficits or catheter-directed therapy was given early or
excesses in survivors and nonsurvivors for prophylactically, but they found no outcome
cardiac index, MAP, pulse oximetry, and tran- improvements in seven other randomized
scutaneous O2 during the period of monitor- studies of patients who entered the ICU after
ing. These differences evaluated by discrimi- organ failure or sepsis had already occurred.
nant analysis to quantify the capacity of each Monitoring and therapy is often ineffective in
variable to affect outcome; 95 % of the sur- the late stages of shock after organ failure has
vivors and 62% of the nonsurvivors were cor- occurred because no amount of oxygen will
rectly classified in the early post-admission re- restore irreversible oxygen debts, failed or-
suscitation period. gans, or dead cells. Optimal therapeutic goals
must be achieved in the first 8 to 12 hours af-
ter surgery or trauma to be effective
Importance of time (1,2,15,24,35-38,42,43).
relationships for improved
outcome
Invasive PA catheters have been used for de-
finitive monitoring of critically ill ICU patients,
Pathophysiology, monitoring, and therapy of shock with organ failure 13

Summary ic breakdowns late in their hospital course

that could not be suspected in the early re-
Early or prophylactic therapy is a useful ap- suscitation period.
proach to improving outcome in acute illness, The conceptual basis of this approach
trauma, and high risk surgery. Noninvasive was that if these methods could be imple-
monitoring provides a way to recognize circu- mented in emergency trauma patients enter-
latory deficiencies earlier, to evaluate the ing the ED in a large inner city county hospi-
evolving temporal patterns in acute illnesses, tal as a worst case scenario, they could be ap-
and to quantify deficits of monitored vari- plied more widely in the ED, OR, and
ables representing cardiac, pulmonary, and throughout the hospital. The major assump-
tissue perfusion functions, and to titrate ther- tion in this approach is that early circulatory
apy to predetermined physiolgic end points deficiencies that ultimately lead to shock, or-
(25,26,34-38,42,43). Our present approach gan failure, and death can be identified short-
quantitatively compares continuous displays ly after onset of trauma, hemorrhage, or sur-
of monitored data with normal values, calcu- gery. Semi-quantitative estimates of the
lates their net cumulative deficits or excesses, amount of deficits may be calculated from
and uses of discriminant analysis to predict the area between normal and the continuous-
outcome. ly monitored data; discriminant analysis pro-
Noninvasive monitoring systems were vides the mathematical basis for prediction of
found to be feasible early in acute emergency outcome. When identified early, these defi-
conditions in patients with potential or sus- ciencies may be more effectively treated than
pected circulatory alterations and significant after occurrence of multiple organ failures
risk of death or organ failures. Noninvasive from irreversible oxygen debt. In essence,
systems are able to identify the time course noninvasive monitoring begun in the ED was
of hemodynamic patterns and to provide the easier, cheaper, faster, safer, and equally sen-
means to calculate quantitatively the accumu- sitive to the invasive PA catheter (25,26,
lated amount of deficit or excess of each 42,43).
monitored variable. For example, the cardiac
index of severely traumatized patients who
lived pumped an average of 81 liters/m2 References
more than would have been provided by the
1. Shoemaker WC, Appel PL, Kram HB et al.
optimal value of 4.0 L/min/m2; this was equiv-
Prospective trial of supranormal values of
alent to 140 liters of cardiac output per pa- survivors as therapeutic goals in high risk
tient. Those who died pumped an average of surgical patients. Chest 1988; 94: 1176
232 liters/m2 or 402 liters per patient less 2. Boyd O, Grounds M, Bennett D. Preopera-
than optimal, a difference of 542 liters. tive increase of oxygen delivery reduces
Noninvasive monitoring provides a plat- mortality in high risk surgical patients. JA-
form to develop a coherent seamless thera- MA 1993; 270: 2699
peutic plan based on physiologic criteria as 3. Yu M, Levy MM, Smith P et al. Effect of max-
the patient proceeds from the ED to the OR, imizing oxygen delivery on mortality and
and to the ICU. Linear discriminant function mortality rates in critically ill patients: A
predicted outcome correctly in 95% of the prospective randomized controlled study.
Crit Care Med 1993; 21: 830-838
survivors and 62% of the nonsurvivors in the
4. Shoemaker WC, Kram HB, Appel PL, Flem-
early period after admission. This is probably ing AW. The efficacy of central venous and
as much as should be expected for nonsur- pulmonary artery catheters and therapy
vivors since many patients develop late lethal based upon them in reducing mortality and
complications rather such as resistant nosico- morbidity. Arch Surg 1990; 125:1332-38
mial infections, drug reactions, wound disrup- 5. Shoemaker WC, Appel PL, Kram HB. He-
tions, postoperative bleeding, and anastomot- modynamic and oxygen transport respons-
14 W. C. Shoemaker, M. Beez

es in survivors and nonsurvivors of high risk 17. Shoemaker WC, Appel PL, Kram HB. Role
surgery. Crit Care Med 1993; 21: 977 of oxygen debt in the development of organ
6. Hankeln KB, Senker R, Schwarten JN et al. failure, sepsis, and death in high risk surgical
Evaluation of prognostic indices based on patients. Chest 1992: 102: 208-15
hemodynamic and oxygen transport vari- 18. Guyton AC, Crowell JW. Dynamics of the
ables in shock patients with ARDS. Crit heart in shock. Fed Proc 1961; 10: 51-54
Care Med 1987; 15: 1 19. Crowell JW, Smith EE. Oxygen deficit and ir-
7. Berlauk JF, Abrams JH, Gilmour IJ et al. Pre- reversible hemorrhagic shock. Am J Physiol
operative optimization of cardiovascular he- 1964; 106: 313-319
modynamics improves outcome in periph- 20. Shoemaker WC, Appel PL, Kram HB, Bishop
eral vascular surgery. Ann Surg 1991; 214: M, Abraham E. Hemodynamic and oxygen
189 transport monitoring to titrate therapy in
8. Moore FA, Haemel JB, Moore EE et al. In- septic shock. New Horiz 1993; 1: 145-159
commensurate oxygen consumption in re- 21. Conners AF,Jr, Speroff T, Dawson NV,
sponse to maximal oxygen availability Thomas C et al. The effectiveness of right
predicts postinjury oxygen failure. J Trauma heart catheterization in the initial care of
1992; 33: 58 critically ill patients. JAMA 1996; 276: 899-
9. Creamer JE, Edwards JD, Nightingale P. He- 97
modynamic and oxygen transport variables 22. Guyatt G, Ontario Intensive care group. A
in cardiogenic shock secondary to acute randomized control trial of right heart
myocardial infarction. Am J Cardiol 1990; catheterization in critically ill patients. J In-
65: 1287 tensive Care Med 1991; 6 :91-95
10. Yu M, Takanishi D, Myers SA, Takiguchi SA 23. Zion MM, Balkin J, Rosenmann D et al: Use
et al. Frequency of mortality and myocardial of pulmonary artery catheters in patients
infarction during maximizing oxygen deliv- with acute myocardial infarction: analysis of
ery: a prospective, randomized trial. Crit experience in 5841 patients in the SPRINT
Care Med 1995; 23: 1025-32 registry. Chest 1990; 98: 1331-35
11. Rady MY, Edwards JD, Rivers EP et al. Mea- 24. Hayes MA, Timmins AC, Yau EHS, Pallazo
surement of oxygen consumption after un- M et al. Elevation of systemic oxygen deliv-
complicated acute myocardial infarction. ery in the treatment of critically ill patients.
Chest 1993; 103: 886-95 New Engl J Med 1994; 330: 1717-22
12. Abraham E, Bland RD, Cobo JC et al. Se- 25. Gattinoni L, Brazzi L, Pelosi P et al. A trial of
quential cardiorespiratory patterns associat- goal-oriented hemodynamic therapy in criti-
ed with outcome in septic shock. Chest cally ill patients. N Engl J Med 1995; 333:
1984; 85: 75-80 1025-32
13. Scalea TM, Simon HM, Duncan AO et al. 26. Taylor RW, and the Pulmonary Artery
Geriatric blunt multiple trauma: Improved Catheter Consensus Conference Partici-
survival with early invasive monitoring. J pants: Pulmonary Artery Catheter Consen-
Trauma 1990; 30: 129-136 sus Conference. Crit Care Med 1997; 25:
14. Tuchschmidt J, Fried J, Astiz M et al. Evalua- 910-25
tion of cardiac output and oxygen delivery 27. Artegas A, Bernard GR, Brigham RL, Carlet
improves outcome in septic shock. Chest J, Falke K et al: The American-European
1992; 202: 216 Consensus Conference on ARDS. Parts 1
15. Bishop MW, Shoemaker WC, Appel PL et al. and 2. Am J Respir Crit Care Med 1994;
Relationship between supranormal values, 149: 818-24 and 1998; 157: 1332-47
time delays and outcome in severely trau- 28. Boyd O, Bennett D. Enhancement of periop-
matized patients. Crit Care Med 1993; 21: erative tissue perfusion as a therapeutic
56 strategy for major surgery. New Horiz 1996;
16. Bishop MH, Shoemaker WC, Kram HB, Or- 4: 453-65
dog GJ et al: Prospective randomized trial of 29. Shoemaker WC, Belzberg H, Wo CCJ,
survivor values of cardiac index, oxygen de- Milzman DP et al. Multicenter study of non-
livery, and oxygen consumption as resusci- invasive monitoring systems as alternatives
tation endpoints in severe trauma. J Trauma to invasive monitoring of acutely ill emer-
1995; 38: 780-7 gency patients. Chest 1998; 114: 1643-52
Pathophysiology, monitoring, and therapy of shock with organ failure 15

30. Shoemaker WC, Wo CCJ, Belzberg H, 42. Berlauk JF, Abrams JH, Gilmour IJ et al. Pre-
Demetriades D et al. Early physiologic pat- operative optimization of cardiovascular he-
terns in acute illnesses and accidents. New modynamics improves outcome in peripher-
Horiz 1996; 4: 395-412 al vascular surgery. Ann Surg 1991; 214:
31. Tremper KK, Waxman K, Shoemaker WC. 289-97
Effects of hypoxia and shock on transcuta- 43. Wang X, Sun H, Adamson D et al. An im-
neous PO2 values in dogs. Crit Care Med pedance cardiography system: A new de-
1979; 7: 526 sign. Ann Biomed Eng 1989; 17: 535-56
32. Tremper KK, Shoemaker WC. Transcuta- 44. Wang X, Van De Water JM, Sun H et al. He-
neous oxygen monitoring of critically ill modynamic monitoring by impedance car-
adults with and without low flow shock. Crit diography with an improved signal process-
Care Med l981; 9: 706-9 ing technique. Proc IEEE Eng Med and Biol
33. Lubbers DW. Theoretical basis of the tran- 1993; 15: 699-700
scutaneous blood gas measurements. Crit 45. Shoemaker WC, Wo CCJ, Bishop MH et al.
Care Med 1981; 9: 721-30 Multicenter trial of a new thoracic electric
34. Venus B, Patel KC, Pratap KS, Konchigeri H, bioimpedance device for cardiac output es-
Vidysager D. Transcutaneous oxygen moni- timation. Crit Care Med 1994; 22: 1907-14
toring during pediatric surgery. Crit Care 46. Shoemaker WC, Wo CCJ, Chan L, Ramcone
Med 1981; 9: 714-16 E et al. Outcome prediction of emergency
35. Tremper KK, Huxtable RF. Dermal heat patients by noninvasive hemodynamic mon-
transport analysis for transcutaneous O2 itoring. Chest 2001; 120: 528-37
measurements. Acta Anesth Scand 1978; 47. Kern JW, Shoemaker WC. Meta-Analysis of
supp1 68: 4-9 hemodynamic optimization in high-risk pa-
36. Tremper KK, Waxman K, Bowman R, Shoe- tients. Crit Care Med 2002; 30: 1686-92
maker WC. Continuous transcutaneous oxy- 48. Shoemaker WC. Effects of transfusion on
gen monitoring during respiratory failure, surviving and nonsurviving postoperative
cardiac decompensation, cardiac arrest, and patients. Surg Gynecol Obstet 1976; 143:
CPR. Crit Care Med 1980: 8: 337 33-40
37. Rowe MI, Weinberg G. Transcutaneous oxy- 49. Shoemaker WC, Wo CCJ. Circulatory ef-
gen monitoring in shock and resuscitation. J fects of whole blood, packed red cells, albu-
Pediatr Surg 1979; 14: 773 min, starch, and crystalloids in resuscitation
38. Shoemaker WC, Appel PL, Kram HB. Inci- of shock, acute critical illness. Vox Sang
dence, physiologic description, compensa- 1998; 74 (Suppl 2): 69-74
tory mechanisms, and therapeutic implica-
tions of monitored events. Crit Care Med
1989; 17: 1277
39. Schultz RJ, Whitfield GF, La Mura JJ et al.
The role of physiologic monitoring in pa-
tients with fractures of hip. J Trauma 1985;
25: 309-16
Correspondence address:
40. Fleming AW, Bishop MH, Shoemaker WC et William C. Shoemaker, M.D.
al. Prospective trial of supranormal values as Department of Surgery
goals of resuscitation in severe trauma. Arch Division of Trauma/Critical Care
Surg 1992; 127: 1175-81 Los Angeles County and
41. Tuchschmidt J, Fried JC, Astiz M et al. Eleva- University of Southern California
tion of cardiac output and oxygen delivery 1200 N State Street
improves outcome in septic shock. Chest Los Angeles, CA 90033
1992; 102: 216-220 USA