Resuscitation of haemorrhagic shock with normal saline vs. lactated Ringer's: effects on oxygenation, extravascular lung water and haemodynamics

Abstract

Introduction: Pulmonary oedema and impairment of oxygenation are reported as common consequences of haemorrhagic shock and resuscitation (HSR). Surprisingly, there is little information in the literature examining differences in crystalloid type during the early phase of HSR regarding the development of pulmonary oedema, the impact on oxygenation and the haemodynamic response. These experiments were designed to determine if differences exist because of crystalloid fluid type in the development of oedema, the impact on oxygenation and the haemodynamic response to fluid administration in early HSR.

Methods: Twenty anaesthetised swine underwent a grade V liver injury and bled without resuscitation for 30 minutes. The animals were randomised to receive, in a blinded fashion, either normal saline (NS; n = 10) or lactated Ringer's solution (LR; n = 10). They were then resuscitated with study fluid to, and maintained at, the preinjury mean arterial pressure (MAP) for 90 minutes.

Results: Extravascular lung water index (EVLWI) began to increase immediately with resuscitation with both fluid types, increasing earlier and to a greater degree with NS. A 1 ml/kg increase in EVLWI from baseline occurred after administartion of (mean +/- standard error of the mean) 68.6 +/- 5.2 ml/kg of normal saline and 81.3 +/- 8.7 ml/kg of LR (P = 0.027). After 150 ml/kg of fluid, EVLWI increased from 9.5 +/- 0.3 ml/kg to 11.4 +/- 0.3 ml/kg NS and from 9.3 +/- 0.2 ml/kg to 10.8 +/- 0.3 ml/kg LR (P = 0.035). Despite this, oxygenation was not significantly impacted (Delta partial pressure of arterial oxygen (PaO2)/fraction of inspired oxygen (FiO2) <or= 100) until approximately 250 ml/kg of either fluid had been administered. Animals resuscitated with NS were more acidaemic (with lower lactates), pH 7.17 +/- 0.03 NS vs. 7.41 +/- 0.02 LR (P < 0.001).

Conclusions: This study suggests that early resuscitation of haemorrhagic shock with NS or LR has little impact on oxygenation when resuscitation volume is less than 250 ml/kg. LR has more favourable effects than NS on EVLWI, pH and blood pressure but not on oxygenation.

Figures

Figure 1

Blood pressure curves for experimental groups. This graph is a composite of the mean arterial pressures of the two resuscitation groups. The liver injury is created at time zero. Fluid resuscitation was begun at 30 minutes into the experiment. There is no difference in mean arterial pressure (MAP) between the two groups until resuscitation. Despite being given significantly more fluid, mean arterial pressures of the animals five the normal saline (NS) was significantly lower than the lactated Ringer's solution (LR) group beginning 39 minutes into the resuscitation (NS 56.9 ± 1.6 mmHg vs. LR 64.0 ± 2.0 mmHg; * p = 0.01)) and remaining so until close to study end.

Figure 2

EVLWI and PaO2/FiO2 per volume of resuscitation fluid. (a) Extravascular lung water index (EVLWI) and partial pressure of arterial oxygen (PaO2)/fraction of inspired oxygen (FiO2) as a function of resuscitation volume over the entire study. Linear regression analysis reveals a larger increase in EVLWI in the normal saline (NS) group as compared with the lactated Ringer's solution (LR) group. An increase of 1 ml/kg of EVLWI occurred at a resuscitation volume of 55.6 ± 6.3 ml/kg for normal saline and 76.9 ± 12.1 ml/kg for LR (p = 0.02). A significant change in oxygenation defined as a drop in PaO2/FiO2 of 100 or more (roughly corresponding to a drop in arterial oxygen saturation below 90% on room air) did not occur until more than 250 ml/kg of either fluid had been administered. (b) EVLWI and PaO2/FiO2 as a function of resuscitation volume of 250 ml/kg or less. To examine fluid type-specific effects we examined changes in oxygenation and EVLWI at similar total volumes of resuscitation. A limit of 250 ml/kg was chosen as none of the LR animals required more than this volume to maintain goal mean arterial pressure (MAP) and to allow examination of the effects of fluid type early in the resuscitation at similar volumes. Linear regression analysis revealed a larger increase in EVLWI in the NS group as compared with the LR group at similar volumes infused as shown by differences in the slopes of the regression lines (p = 0.027). An increase of 1 ml/kg of EVLWI occurred at a resuscitation volume of 68.6 ± 5.2ml/kg for NS and 81.3 ± 8.7 ml/kg for LR (p = 0.027). A significant change in PaO2/FiO2 (≥ 100) was not seen over this range of volume.

Figure 3

Haemodynamic data during resuscitation. Boxplots for the means of the values at the time the difference in mean arterial pressure (MAP) between the two groups first became significant. (a) Significant differences in MAP, cardiac output (CO) and systemic vascular resistance (SVR). There were no differences in baseline values immediately prior to resuscitation between the groups. During the resuscitation stage MAP was lower in the normal saline (NS) group (NS 56.9 ± 1.6 mmHg vs. lactated Ringer's solution (LR) 64.0 ± 2.0 mmHg; p = 0.01) despite this group having higher COs (NS 5.2 ± 0.3 l/minute vs. LR 4.4 ± 0.2 l/minute; p = 0.016). This was due to a significant difference in SVR (NS 923 ± 51.4 dyne × sec/m3vs. 1177.6 ± 34.6 dyne × sec/m3; p < 0.001). (b) There were no differences in the preload metric global end-diastolic volume (GEDV) or in stroke volume (SV). The differences in CO seen in (a) were due to a significant increased heart rate (HR) in the NS group (NS 114.9 ± 6.5 beats/minute vs. LR 93.6 ± 3.6 beats/minute; p = 0.012).

Figure 4

Mean PVPI during resuscitation. Boxplots for mean pulmonary vascular permeability index (PVPI) at the time when the difference in extravascular lung water index between the two groups first became significant. A significant difference between the two groups was found (normal saline (NS) 3.6 ± 0.25 vs. lactated Ringer's solution (LR) 2.9 ± 0.13; p = 0.014).