Reliability of central venous pressure to assess left ventricular preload for fluid resuscitation in patients with septic shock

Abstract

Background: Initial fluid resuscitation is an important hemodynamic therapy in patients with septic shock. The Surviving Sepsis Campaign Guidelines recommend fluid resuscitation with volume loading according to central venous pressure (CVP). However, patients with septic shock often develop a transient decrease in cardiac function; thus, it may be inappropriate to use CVP as a reliable marker for fluid management.

Methods: We evaluated 40 adult patients with septic shock secondary to intra-abdominal infection who received active treatment and were monitored using transthoracic echocardiography (TTE) and CVP for 2 days after admission to our intensive care unit (ICU). We measured left ventricular end-diastolic diameter (LVEDD), left atrial diameter (LAD), and the pressure gradient of tricuspid regurgitation (TR∆P). The shock status was treated with volume loading and inotrope/vasopressor administration according to the TTE findings. We assessed left ventricular fractional shortening (LVFS) as an index of left ventricular contractility and TR∆P as an index of right ventricular afterload and then examined the correlation between CVP and LVEDD/LAD/TR∆P.

Results: LVFS decreased to ≤30% in 42.5% and 27.5% of patients with septic shock, and severe left ventricular dysfunction with LVFS ≤20% developed in 12.5% and 15.0% of patients on the first and second ICU days, respectively, despite the use of inotropes/vasopressors. Mild pulmonary hypertension as indicated by TR∆P ≥30 mmHg was present in 27.5% and 30.0% of patients on their first and second ICU days, respectively. There was no significant correlation between CVP and LVEDD/LAD/TR∆P. The hospital mortality rate in this study was 10.0%, although the predicted mortality based on the Acute Physiology and Chronic Health Evaluation II score was 58.7%.

Conclusions: Our results suggest that CVP is not a reliable marker of left ventricular preload for fluid management during the initial phase of septic shock. Assessment of left ventricular preload, right ventricular overload, and left ventricular contractility using TTE seems to be more informative than the measurement of CVP for fluid resuscitation since some patients developed left ventricular dysfunction and/or right ventricular overload.

Keywords: Central venous pressure; Echocardiography; Septic shock.

Figure 1

Screening of study patients.

Figure 2

Left ventricular fractional shortening in patients with septic shock. Left ventricular fractional shortening decreased to ≤30% in 42.5% and 27.5%, and to ≤20% in 12.5% and 15.0% of patients on the first and second intensive care unit (ICU) days, respectively. There was no significant difference (P = 0.72) in median value between the first and second ICU days.

Figure 3

Pressure gradient of tricuspid regurgitation in patients with septic shock. Mild pulmonary hypertension with a pressure gradient of tricuspid regurgitation of ≥30 mmHg occurred in 27.5% and 30.0% on the first and second intensive care unit (ICU) days, respectively. There was no significant difference (P =0.17) in median value between the first and second ICU days.

Figure 4

Central venous pressure versus left ventricular end-diastolic diameter in patients with septic shock. There was no significant correlation between central venous pressure and left ventricular end-diastolic diameter on the first and second intensive care unit days.

Figure 5

Central venous pressure versus left atrial diameter in patients with septic shock. There was no significant correlation between central venous pressure and left atrial diameter on the first and second intensive care unit days.

Figure 6

Central venous pressure versus the pressure gradient of tricuspid regurgitation in patients with septic shock. There was no significant correlation between central venous pressure and the pressure gradient of tricuspid regurgitation on the first and second intensive care unit days.