Correlation of Blood Gas Parameters with Central Venous Pressure in Patients with Septic Shock; a Pilot Study

Abstract

Objective:

To determine the correlation between blood gas parameters and central venous pressure (CVP) in patients suffering from septic shock.

Methods:

Forty adult patients with diagnosis of septic shock who were admitted to the emergency department (ED) of Shohadaye Tajrish Hospital affiliated with Shahid Beheshti University of Medical Sciences, and met inclusion and exclusion criteria were enrolled. For all patients, sampling was done for venous blood gas analysis, serum sodium and chlorine levels. At the time of sampling; blood pressure, pulse rate and CVP were recorded. Correlation between blood gas parameters and hemodynamic indices were.

Results:

A significant direct correlation between CVP with anion gap (AG) and inversely with base deficit (BD) and bicarbonate. CVP also showed a relative correlation with pH, whereas it was not correlated with BD/ AG ratio and serum chlorine level. There was no significant association between CVP and clinical parameters including shock index (SI) and mean arterial pressure (MAP).

Conclusion:

It seems that some of non invasive blood gas parameters could be served as alternative to invasive measures such as CVP in treatment planning of patients referred to an ED with septic shock.

Introduction

The early recognition and management of shock is one of the most important clinical aspects of critical care in an emergency department (ED). The management and prognosis of shock certainly depends  on rapid identification  and also careful assessment  corresponding  with  treatment  of the underlying cause. The initial evaluation of patients with possible hypoperfusion  are guided by some clinical and paraclinical factors. It has been reported that early stage of hypoperfusion is insufficient to make changes in vital signs in previously healthy individuals; however, quantifying the hypoperfusion degree remains a challenging task [1]. Assessment of the possibility to replace invasive parameters with less invasive paraclinic variables and some clinical ones for the discrimination of shock has been tried in recent years. The evaluation and therapy is commonly guided by shock index (SI), base deficit (BD), lactate clearance,  anion gap (AG), pH, bicarbonate  and also saturation of central venous oxygen (ScvO ). Improved understanding of the correlation between these variables and central venous pressure (CVP) may also reveal new strategies for assessment of septic shock in patients who need critical care at arrival to an emergency department (ED) [1-4]. The SI is normally 0.5 to 0.7 and has been almost likely be elevated in the setting of acute hypovolemia. Reviewing previous studies showed that abnormal BD can be an indicator for the increased transfusion requirement and also is related to higher incidence of shock-related complications, intensive care unit (ICU),  in-hospital  lengths  of stay and increased mortality. Furthermore, it was suggested as an indicator and monitoring parameter for the success of resuscitation efforts [5-8]. Other studies showed that early therapy directed to relieve tissue hypoxia and repay oxygen improves clinical outcome in shock irrespective of its etiology. Mixed venous oxygen saturation obtained from pulmonary artery (SvO ) and central venous oxygen saturation (ScvO ) were shown to reflect the balance of systemic oxygen delivery (supply) to consumption (demand) [5,7-9]. Also, there are literature reviews that discussed the use of adjuncts such as blood lactate level in the initial assessment that may increase the ability to accurately recognize patients with occult hypoperfusion or early shock secondary to hemorrhage [9-11]. Base on the above mentioned,  this study aimed to assess the correlation between venous blood gas parameters and CVP in the setting of septic shock, trying to develop a non invasive method for quantifying the hypoperfusion degree.

Materials and Methods

Study Design and Setting

This study was conducted from September to November 2013 in ED of Shohadaye Tajrish Hospital affiliated with Shahid Beheshti University of Medical Sciences, a major referral hospital in Tehran, Iran. The principles of the Declaration of Helsinki were applied throughout the study. The study protocol was approved by institutional review board (IRB) and ethics committee of Shahid Beheshti University of Medical Sciences. All the participants provided their informed written consents before inclusion in the study.

Participants

Forty adult patients (Age>18 years) with diagnosis of septic shock who were admitted  to ED were included, regardless of sepsis source. Septic shock was defined as systemic  inflammatory  response syndrome (SIRS) along with evidence of infection, including positive blood culture, signs of pneumonia on chest x-ray, or other radiologic  or laboratory evidence of infection; signs of end-organ dysfunction are required such as renal failure, liver dysfunction, changes in mental status, or elevated serum lactate; refractory hypotension (low blood pressure that does not respond to treatment). SIRS was also defined as presence of 2 out of these 4 criteria: tachypnea (>20 breaths per minute), or on blood gas, a PCO less than 32 mmHg signifying hyperventilation; leukopenia (<4000 cells/mm3) or leukocytosis (>12000 cells/ mm3); tachycardia  (>90 beats per minute); Fever (>38.0°C) or hypothermia (<36.0°C). Each patient that had at least four criteria of six, as mentioned below was enrolled as candidate for central venous catheterization:  Ill appearance  or altered mental status,  heart rate >100 beats/minute,  respiratory rate >20 breaths/minute or PaCO<32 mmHg, base deficit <−4 mEq/Liter or lactate >4 mmol/Liter, urine output <0.5 mL/kg/hour, and arterial hypotension more than 30 minutes continuous duration. Exclusion criteria were defined as CVP catheterization contraindication, ∆CVP more than 4 cmH O with impression of obstructive shock.

Data Collection

After  diagnosis  of circulatory  shock,  a central venous catheter was inserted by emergency physicians.  For  all these  patients,  venous  blood sampling was requested for gas analysis, sodium (Na) and chlorine (Cl) levels. At the time of sampling, non-invasive blood pressure, pulse rate and also CVP were recorded.

Statistical Analysis

The collected data were analyzed using the SPSS statistical software package (Version 21, SPSS Inc, Chicago, IL, USA). Correlation coefficient (r) was calculated with Pearson’s correlation test to analyze the significance of relation between CVP with blood gas, Cl, mean arterial pressure (MAP), and SI. A p-value of <0.05 was considered to be statistically significant.

Results

A  total  of  40  patients  who  were  diagnosed with septic shock included 15 (37.5%) females and 25 (62.5%) males, ranging in age from 34 to 94 (mean age=70.45±16.0) years were enrolled. The mean SI was 1.16±0.4 and ranged from 0.23 to 2.2 (normal range=0.5-0.7). Pearson’s correlation tests showed a direct correlation between CVP and AG (r=0.388, p=0.013,) (Figure 1) and inverse relation with HCO (r=-0.465, p=0.002,) and BE (r=-0.391, p=0.013,) (Figure 2 and 3). Association  between CVP and paraclinical and clinical parameters in patients were summarized in Table 1.

Fig. 1.

Central venous pressure–Anion Gap correlation in patients with septic shock

Fig. 2.

Central venous pressure–base deficit correlation in patients with septic shock

Fig. 3.

Central venous pressure–HCO3 correlation in patients with septic shock

Table 1.

Central venous pressure and paraclinical and clinical parameters association in patients with septic shock

	AG	BD	BD b /AG c	pH	HCO3 d	Cl e	MAP f	SI g
Correlation Coefficient	0.388a	-0.391a	-0.139	-0.273	-0.465a	-0.177	-0.041	-0.095
P -value	0.013	0.013	0.391	0.088	0.002	0.300	0.804	0.561

^a Significant correlation;

^b BD: Base Deficit;

^c AG: Anion Gap;

^dHCO3: Bicarbonate;

^eCl: Cholaride;

^fMAP: Mean Arterial Pressure;

^gSI: Shock Index

Discussion

The results of the present study showed indirect correlation between BE and HCO3, and direct correlation  between  anion gap with CVP in the setting  of septic  shock.  However,  there  was not any significant correlation between CVP and other studied parameters. There are few reports regarding the assessment of relation among the physiologic and blood gas variables in treatment of septic shock. The increased rate of SI was observed in 94% of patients even after normalization of heart rate (HR), MAP and CVP. There are several evidences regarding the morbidity or mortality after normalization of physiologic  variables  including  MAP,  CVP  and HR, so it seems that these physiologic  variables cannot serve as consistent indexes in shock state [12-14]. It has been reported that SI could be served as adjunct monitoring tools in identifying systemic oxygenation, cardiac function, hypoperfusion during initial therapy of shock [15]. In contrast to, the closely relation of base deficit with SI has been observed in several studies and it has also been suggested as predicting factor in multi organ damage death and hemorrhage [16-19]. In our study, we found a good correlation between CVP and blood gas variables. Our results were consistent with the findings of Kaplan study in traumatic patients [2,20]. This suggests that non-invasive monitoring and prompt intervention are necessary to identify and control of shock using several paraclinical factors such as AG, BD, pH and HCO3. In a study recently performed by Drakhshanfar et al., [5] among Iranian pediatric patients, BD caused by metabolic acidosis was introduced as a predicting factor in blunt abdominal trauma. As we mentioned, there  are  still  some  controversies  regarding  the correlation between SI, pH and MAP with CVP in the current study. The prospective and retrospective studies should be required to validate and expand these findings. We proposed that these indices can be tightly linked to septic shock patients and can be applied in a routine clinical practice for rapid critical treatment. Currently there is no consensus on the accurate markers of hemodynamic instability in hypovolemia and   septic   shock.  There   is   also   controversy regarding the endpoints for optimal fluid therapy in hypovolemic  or septic shock [21,22]. However recent lines of evidence suggest that fluid therapy should be based on dynamic (such as cardiac output, pulse pressure variation and stroke volume variation) rather than static hemodynamic variables (such as CVP, pulmonary artery occlusion pressure). This is because the dynamic markers are better predictors of fluid resuscitation especially in critically ill patients [23,24]. In the same way several studies have revealed the role of O saturation monitoring in predicting the outcome of patients with septic shock [25,26]. It has been observed that septic patients have lower O2 saturation  values  when  compared  to healthy individuals or less severe septic patients [25]. In a study by Georger et al. [27] it was shown that O saturation can be increased by administration  of norepinephrine  in septic shock patients. In these patients the O2 saturation has been shown to be correlated with MAP [28]. Taking all these findings together it can be postulated that O2 saturation is an indicator of both endothelial function [29] and regional tissue perfusion pressure; the latter is affected by MAP. Capillary recruitment secondary to local vasodilatation, as well as perfusion pressure of the tissue bed is effective factors in reoxygenation of the tissue in response to an ischemic challenge especially in septic  shock.  Thus  reoxygenation  happens  in the interaction between the macrohemodynamics (perfusion pressure) and microcirculation (endothelial function). The present lines of evidence suggest that there is a correlation between the O2 saturation slope and the prognosis [30]. Taking all these together, it could be concluded that the oxygenation of the tissues in septic shock determined by O2 saturation is a valuable indicator and predictor of response to therapy [27]. However the fact that improvement of O2 saturation may alleviate the prognosis of patients with septic shock remains the matter of debate. We note some limitations to our study. We measured the baseline blood gas parameters in septic shock patients and we did not follow that changes of these variables. We also did not correlate these parameters with patients’ clinic and outcome. The main limitation of the study is the lack of power to analyze mortality prediction. As we have not measured the blood gas parameters in different stages of the shock, we cannot comment on their role and correlation with the clinical entity. Future studies with continuous measurement of blood gas parameters and correlation with clinical parameters is recommended.

Another potential limitation derives from the fact that VOT-derived variables were obtained once. In conclusion, it seems that some of non-invasive blood gas parameters could be served as alternative to invasive measures such as CVP in treatment planning of patients referred to an ED with septic shock.

Conflict of Interest: None declared.

Funding/Support

All authors declare that this study was accomplished without any funding or support and the authors were responsible for all expense. This study was part of Dr. Fariba Amini’s thesis as general physician course at the Shahid Beheshti University of Medical Sciences.