Resuscitation, 23 (1992) 221-234 Elsevier Scientific Publishers Ireland Ltd.

Shock index: a re-evaluation failure*

221

in acute circulatory

Mohamed Y. Rady”, Peter Nightingalea, Roderick A. Littleb and J. Denis Edwardsb uIntensive Care Unit, University Hospital of South Manchester and “North Western Injury Research Centre. Manchester. Manchester (UK)

(Received April 30th, 1992; accepted May I1th, 1992)

Study Objective: To evaluate the relationship between the shock index SI (ratio of heart rate to systolic arterial pressure) and cardiac function and oxygen transport in an experimental model of hemorrhage and clinical septic shock. Methods and Results: This study was conducted in a hypovolemic circulatory failure model; 40?/ hemorrhage in the anesthetized pig and normovolemic hyperdynamic septic patients in the intensive care unit (ICU). Hemodynamic and oxygen transport variables were measured and their relationships to SI was examined. SI was inversely related to blood loss, cardiac index (CI), stroke volume (SV), mean arterial pressure (MAP) and left ventricular stroke work (LVSW) (r = -0.73, -0.75, -0.89 and -0.75, respectively P < 0.01) following hemorrhage in the anesthetized pig. Oxygen transport variables, i.e. oxygen delivery (Do*) and mixed venous oxygen saturation (Svoz) (r = -0.68 and -0.74, respectively, P < 0.01) were also inversely related to the SI. Oxygen consumption (VO,) increased initially with increasing SI and fell when SI was greater than 3.0. In clinical septic shock and following blood volume expansion, the SI was not correlated to CI, SVI, MAP or systemic vascular resistance (SVR) (r = -0.01, -0.47, -0.34 and -0.14, respectively, P-value NS) but was inversely related to LVSWI (r = -0.68, P < 0.01). There were no relationships between the SI and oxygen transport variables (DOZ. $0,) (r = -0.02 and -0.17, P-value NS) in septic shock. Conclusion: SI provides a non-invasive means to monitor deterioration or recovery of LVSW during acute hypovolemic and normovolemic circulatory failure and its therapy. SI may be of limited value in the assessment of systemic oxygen transport and response to therapy in clinical shock. Key words: shock index; hemorrhage; hyperdynamic sepsis; left ventricular stroke work; oxygen transport

INTRODUCTION

The SI, which is defined as the ratio of heart rate to systolic arterial pressure, was first used by Allgower and Buri in 1967 [ 11. They observed that the normal range was 0.5-0.7 in a healthy adult and increased to values as high as 2.5 following acute gastrointestinal hemorrhage. There was a proportional rise in the SI with progressive Correspondence to: M.Y. Rady, Department of Emergency Medicine, Henry Ford Hospital, 2799 West Grand Blvd, Detroit, MI 48202, USA. *Presented April 1990 at the annual conference of British Society of Intensive Care Medicine, Southampton. UK.

0300-9572/92/$05.00 0 1992 Elsevier Scientific Publishers Ireland Ltd. Printed and Published in Ireland

228

loss in circulating blood volume in these patients. It was assumed that the SI was a sensitive guide to the degree of hypovolemia following hemorrhage. However, Shippy et al. found no correlation between heart rate or mean arterial pressure and the circulating blood volume in critical illness [2]. We have re-evaluated the physiological significance of the SI in acute circulatory failure with reduced (i.e. simple hemorrhage) or expanded circulating blood volume (i.e. hyperdynamic sepsis). MATERIALS,

PATIENTS

AND

METHODS

Experimental hemorrhagic shock

The following experiments were conducted in accordance with the Animals (Scientific Procedures) Act, 1986 (UK), which adequately covers the NIH guidelines for experimentation on animals (USA). Large White pigs (n = 21, 15-27 kg) were anesthetized with ketamine 20 mg kg-’ i.m. and mechanically ventilated with a mixture of oxygen, nitrous oxide (F102 = 0.5) and isoflurane (1.5%). Instrumentation included a right and left femoral arterial cannulas for bleeding and monitoring arterial pressure (Portex ID 2.5 mm). A French 7 gauge thermodilution pulmonary artery catheter (American Edwards, Baxter Healthcare Ltd., UK) was inserted via the right femoral vein to measure cardiac output (Cardiac output computer COM-I 9310, American Edwards Laboratories, CA, USA), right atria1 pressure and pulmonary arterial and wedge pressures. Core temperature was maintained at 38.5”C. Pulmonary and systemic arterial pressure and electrocardiogram were recorded on a six channel recorder (M-19 Devices, UK) with thermal pen. Systemic and pulmonary arterial blood gases and oxyhemoglobin saturations were measured with ABL 330 and 0SM3 Hemoximeter (Radiometer, Copenhagen). Hemorrhage was induced by bleeding at a rate of 0.75 ml kg-’ mm-’ with a calibrated pump (Watson and Marlow 503 S, UK) until 30 ml kg-’ of blood was withdrawn. Hemodynamic and oxygen transport variables were measured at 15 ml kg-‘, 30 ml kg-’ and 30 min later. Derived variables were calculated using standard formulae [3]. Pearson correlation coefftcients between the SI and different physiological variables were calculated using simple regression analysis with statistical significance at P < 0.01. Clinical septic shock

This clinical study was approved by the Human Research ethical committee of University Hospital of South Manchester. Sixteen consecutive patients admitted to the Intensive Care Unit (ICU) at UHSM in septic shock were studied. The criteria for septic shock were: systemic hypotension (MAP < 60 mmHg) after blood volume oliguria, pyrexia (> 38”(Z), leucocytosis (> 11 000 dl-‘) and expansion, bacteriological evidence of a source of sepsis from the bloodstream or a localized site. All patients required mechanical ventilation and femoral and pulmonary arterial catheterization. Cardiac output was measured by thermodilution with cold 10 ml injectate (< 10°C) of 5% dextrose in triplicates and vascular pressures were recorded simultaneously (Marquette, Milwaukee). Systemic and pulmonary arterial blood haemoglobin, oxyhemoglobin saturation and oxygen partial pressure were

229 Table 1. The relationships between oxygen delivery (Do$ and percent blood volume loss (% BVL) with mean arterial pressure (MAP), heart rate (HR) and shock index (St) during hemorrhage in isoflurane anesthetized pigs. Values Pearsons correlation coefficients.

MAP (mmHg) HR (beats min-‘) SI

0.60* -0.57* -0.68;

-0.81* 0.97* 0.84:

*P c 0.01.

measured with oximeter-282 and IL-l 3 12 (Instrumentation Laboratories). Controlled blood volume expansion with modified fluid gelatin (MFG) and blood was guided by frequent measurement of wedge pressure (not exceeding 18 mmHg) until maximal LVSWI was achieved [4]. The arterial hematocrit was maintained between 33 and 35%. Inotropic and vasopressor drugs were administered according to a standard protocol [5]. Hemodynamic and oxygen transport variables and SI were measured after blood volume expansion and during subsequent therapy. Pearson correlation coefficients between the SI and different physiological variables were calculated for pooled (294 measurements in sixteen patients) and individual patients data (no less than 6 measurements) using simple regression analysis. Statistical significance was chosen at P c 0.01. RESULTS

The results are presented in Tables I, II and III and Figs 1, 2 and 3. The shock index (SI = HR/systolic BP) rose progressively as blood loss increased (Table I) and values greater than 3.0 predicted an impending cardiovascular collapse and a fall in the VOW(Fig. 1). There were significant negative correlations between the cardiac

Table II. The relationships between the shock index (St) and mean arterial pressure (MAP). cardiac index (CI), stroke volume (SV), left ventricular stroke work (LVSW), systemic vascular resistance (SVR), oxygen delivery (DON) and mixed venous oxygen saturation (SO,) during hemorrhage in isoflurane anesthetized pigs. Values are Pearson correlation coefficients.

MAP (mmHg) CI (ml min-’ kg-‘) SV (ml beat-‘) LVSW (g) SVR (dynes s-t cmes) Do, (ml min-’ kg-‘) svq

WI

*P c 0.01.

Sl

I/S1

Log SI

-0.75* -0.73* -0.75* -0.75’ 0.37 -0.68* -0.72;

0.67* 0.68* 0.91* 0.93* -0.37 0.62’ 0.69*

-0.73; -0.74* -0.87* -0.89* 0.37 -0.68* -0.75*

230 Table III. The relationships between the shock index (SI) and mean arterial pressure (MAP), cardiac index (CI), stroke volume index (SVI), left ventricular stroke work index (LVSWI), systemic vascular resistance (SVR), oxygen delivery (DO,) and mixed venous oxygen saturation (S,oJ in sixteen patients with septic shock following controlled blood volume expansion and in sixteen patients with reduced circulating blood volume (hemorrhagic shock, data derived from Cournand et al., 1943). Values are Pearson correlation coefficients.

MAP (mmHg) CI (I min-t me2) SVI (ml beat-’ me2) LVSWI (g mb2) SVR(dynes s-’ cme5) Do2 (ml min-’ me2) 5402

VW

SI (Septic shock)

FI!Iemorrhagic shock)

-0.34 -0.01 -0.47 -0.68* -0.14 -0.02 -0.17

-0.86, -0.64* -0.82* -0.82* 0.35 -0.62* -0.73*

*P c 0.01.

flow (CI, SV and LVSW) and oxygen transport (DO* and SO*) variables and the SI during hemorrhage (Table II). Although some of these relationships (e.g. LVSW) seemed to be improved by taking an inverse function or a logarithmic transformation of SI, the changes were not significant (Table II). Following controlled blood volume expansion in clinical septic shock, the SI varied from 0.4 to 2.4 with a mean (S.D.) of 0.8 (0.3). The SI did not correlate with CI, SV, MAP or SVR but was inversely related to LVSWI (Table III). There were no correlations between the SI and oxygen transport variables (DON, VOW,SO,).

II

0.5

210

1.0 Shock

410

8

(Log. scale1

index (SII

Fig. 1. The relationship between the shock index (SI) and oxygen consumption (VqI) in eight isoflurane anesthetized pigs during simple hemorrhage. Each line represents data from individual animals.

231 250

0.4

0:5

I:0

115

214

(Log scale1

shock index

Fig. 2. The relationship between the shock index (SI) and oxygen consumption (VoJ) in eight patients following controlled blood volume expansion in septic shock. Each line represents data from individual patients.

400

1 .

350 %

300 -

. *z

250 -

‘1

. . .

.

0.7

shock

.

1.4

index

Fig. 3. The relationship between the shock index (SI) and oxygen consumption (Vo,I) from pooled data from sixteen patients in hypovolaemic shock (Data from Cournand et al.).

232 DISCUSSION

In simple hemorrhage, the cardiac flow (CI, SV and LVSW), MAP and oxygen transport variables (Do2 and Svo2) were reduced with progressive blood loss. The blood loss and DO, were significantly related to HR, MAP and SI in the anesthetized pig. The SI was inversely related to left ventricular function variables, i.e. CI, SV, LVSW and MAP. There was a similar relationship between DO, and S,O, and the SI during hemorrhage. Cournand et al. measured circulating blood volume, hemodynamic and oxygen transport variables in clinical hypovolaemic shock before resuscitation [6]. It was possible to calculate the SI from the data provided and examine the relationship between the SI and blood volume, hemodynamic and oxygen transport variables. There were similar inverse relationships between the SI and circulating blood volume, left ventricular function, i.e. CI, SVI, LVSW and MAP and oxygen transport variables (Do2 and S,O,) similar to that in the anesthetized pig (Table III). It seemed that SI indicated the severity of reduction in systemic blood flow and oxygen transport in simple hemorrhage. The relationship between the SI and Vo2 was unusual during hemorrhage in the anesthetized pig. Although DO* was reduced with concomitant rise in the SI, the VO, was increased initially above control values (by about 7%) in 16 out of 21 animals studied. This is consistent with Weiskopft et al. observation of an increase in Vo2 in mechanically ventilated pigs following hemorrhage [7]. Later when the SI increased above 3.0 the VO, returned to control values and any further increase in the SI was associated with a sharp fall in Vo2. This reduction in Vo, preceded the hemodynamic collapse and death in the anesthetized pig. Therefore, it seemed that the critical Do2 (defined as the lowest Do2 below which VOWfell sharply) [8] was reached when the SI was greater than 3.0 in simple hemorrhage in isoflurane anesthetized pigs. This coincided with more than 50% and 70% reductions in CI and LVSW respectively. Analysis of the SI and Vo2 (from Cournand et al. data) failed to show any critical SI value (in the range 0.4-2.4) above which VO, decreased during hemorrhage. The pooling of small number of measurements from different patients studied could account for the lack of a critical SI being observed in that clinical study [6]. In clinical septic shock, peripheral vascular failure (i.e., reduced SVR) and myocardial depression manifest when optimal blood volume has been restored [g-13]. After controlled blood volume expansion in septic shock, the SI was still inversely related to the LVSWI. There were no significant relationships between the SI and systemic blood flow (CI and SVI), systemic oxygen transport (Do2 and SvOz),, MAP or SVR. The SI did not reflect the severity of peripheral vascular failure (i.e. SVR) observed in sepsis. The lack of such relationships was not altered by the use of vasopressor and/or inotropes in septic shock. Again, there was no critical SI value observed to indicate a decrease in VO, in either individual or pooled patients data. This can be explained by the lack of correlation between Do2 and SI in septic shock. In simple hemorrhage the SI was inversely related to SV (i.e. CO/HR) and MAP and therefore LVSW which is derived from both variables. However, this explanation does not hold in septic shock since neither SV nor MAP were significantly related to the SI but it was still inversely related to LVSWI. This

233

suggested that such a relationship was not solely due to mathematical coupling because of shared physiological variables (i.e. HR and systolic arterial pressure). Several clinical studies have shown that LVSWI was an important determinant of final outcome in critical illness [4,14-191. Since there is an inverse relationship between the SI and LVSW in shock, it may be assumed that the SI can be of use as a prognostic variable in critical illness. Allogower and Buri reported a mortality rate of 40% at a SI > 1.0 following blunt abdominal trauma [l]. Later Oslen et al. observed that persistent elevation of the SI (> 1.0) for several hours following trauma was related to poor outcome [20]. A persistent elevation of the SI > 1.0 (or LVSWI < 25 g rnm2)for several hours would significantly increase the morbidity and mortality in acute circulatory failure. The LVSWI is dependent on preload (or circulating blood volume), afterload, heart rate and LV mechanical performance (diastolic compliance and systolic contractility) [21,22]. The SI may reflect changes in the forementioned factors affecting the LVSW, however, this can limit its use to decide upon the appropriate therapeutic regime (i.e. blood volume expansion or inotropes etc.) in acute circulatory failure. In conclusion, the SI is related to LVSW in acute circulatory failure with reduced or normal circulating blood volume. It may be of use to assess the severity and initial response to resuscitation in shock. The persistent elevation of the SI (> 1.O) indicates an impaired left ventricular function (due to blood loss and/or cardiac depression) and a high mortality rate and therefore the necessity for monitoring and optimization of hemodynamics and oxygen transport. REFERENCES M. Allgower and C. Buri, Schockindex, Deutsche Medizinische Wochenschrift, 46 (1967) I-IO. CR. Shippy, P.L. Appel and WC. Shoemaker, Reliability of clinical monitoring to assess blood volume in critically ill patients, Crit. Care Med., I2 (1984) 107-I IO. R. Bland, W.C. Shoemaker, E. Abraham and J.C. Cobo, Hemodynamic ard oxygen transport patterns in surviving and nonsurviving postoperative patients. Crit. Care Med., I3 (1985) 85-90. M.Y. Rady, J.D. Edwards, P. Nightingale and A.J. Mortimer, Cardiorespiratory patterns in survivors and nonsurvivors following blunt chest trauma, B. J. H. M., 42 (1989) 142. J.D. Edwards, G.C. Brown, P. Nightingale, R. Slater and E.B. Faragher, Use of cardiorespiratory values as therapeutic goals in septic shock, Crit. Care Med., I7 (1989) 1098-I 103. A. Coumand, R.L. Riley, SE. Bradley, ES. Breed, R.P. Noble, H.D. Lawson, M.I. Gregersen and D.W. Richards, Studies of the circulation in clinical shock, Surgery, I3 (1943) 964-995. R.B. Weiskopf, MS. Bogetz, I.A. Reid, M.F. Roizen and L.C. Keil, Cardiovascular, endocrine and metabolic responses of conscious swine to hemorrhage. In: Swine In Biomedical Research, Editor: M.E. Tumbleson, Plenum Press, New York, 1986, pp. 1405-141 I. P.T. Schumacker and S.M. Cain, The concept of a critical oxygen delivery, Inten. Care Med., I3 (I 987) 223-229. H.R. Adams, J.L. Parker and M.H. Laughlin, Intrinsic myocardial dysfunction during endotoxemia: dependent or independent of myocardial ischaemia? Circ. Shock, 30 (1990) 63-76. C.H. Baker and F.R. Wilmoth, Microvascular response to E. coli endotoxin with altered adrenergic activity, Circ. Shock, I2 (1984) 165. G. Carroll and J. Synder, Hyperdynamic severe intravascular sepsis depends on fluid administration in cynomolgus monkeys, Am. J. Physiol., 243 (1982) Rl3l-Rl4l. V. Dorio, C. Wahlen, M. Naldi, A. Fossion, J. Juchmes and R. Marcelle, Contribution of peripheral pooling to central hemodynamic disturbances during endotoxin insult in intact dogs, Crit. Care Med., I7 (1989) 1314-1319.

234 13 H.J. Lubbesmeyer, J.L. Theissen, S. Doty, R. Kimura, L., Traber, D.N. Herndon and D.L. Traber, Myocardial depression in hyperdynamic endotoxaemia, Circ. Shock, 26 (1988) 139-146. 14 F.P. Ognibene, M.M. Parker, C. Natanson and J.E. Parrillo, Depressed left ventricular performance in response to volume infusion in patients with sepsis and septic shock, Chest, 93 (1988) 903-910. 15 SM. Jafri, S. Lavine, B.E. Field, M.T. Bahorozian and R.W. Carlson, Left ventricular diastolic function in sepsis, Crit. Care Med., 18 (1990) 709-714. 16 M.M. Parker and J.E. Parrillo, Myocardial function in septic shock, J. Crit. Care, 5 (1990) 47-61. 17 WC. Shoemaker, P.L. Appel, R. Bland, J.A. Hopkins and P. Chang, Clinical trial of an algorithm for outcome prediction in acute circulatory failure, Crit. Care Med., 10 (1982) 390-397. 18 W.C. Shoemaker, E.S. Montgomery, E. Kaplan and D.H. Elwyn, Physiologic patterns in surviving and nonsurviving shock patients. Use of sequential cardiorespiratory variables in defining criteria for therapeutic goals and early warning of death, Arch. Surg., 106 (1973) 630-636. 19 W.C. Shoemaker and J.M. Reinhard, Tissue perfusion defects in shock and trauma states, Surg. Gyn. & Obst., 137 (1973) 980-986. 20 H.J. Ostern, 0. Trentz, G. Hempelmann, O.A. Trentz and J. Sturm, Cardiorespiratory and metabolic patterns in multiple trauma patients, Resuscitation, 7 (1980) 169-184. 21 R.D. Goldfarb, Cardiac mechanical performance in circulatory shock, Circ. Shock, 9 (1982) 633-653. 22 W.J. Sibbald, Myocardial function in the critically ill: Factors influencing left and right ventricular performance in patients with sepsis and trauma, Surg. Clin. North Am., 65 (1985) 867-893.