Cardiac failure in sepsis is characterized by acute left ventricular (LV) systolic dysfunction, LV diastolic dysfunction, and right ventricular (RV) dysfunction, either alone or in combination (1). The prognosis associated with sepsis-related myocardial dysfunction is not clear, but there are many studies that found this condition to be predictive of an impaired outcome (2). However, formal criteria to diagnose sepsis-related myocardial dysfunction seem to be lacking (3). To improve the prognosis for septic patients, many studies investigated potential parameters that could elucidate the link between cardiac performance and outcome in patients with severe sepsis and septic shock. Historically, the main focus has been on LV function, where studies demonstrated a variable correlation with mortality (4, 5). In addition, studies investigating the correlation between RV function and outcome have also produced conflicting results (6–9). Despite a current lack of evidence, the right ventricle potentially holds a key to define a reliable cardiac parameter that can be used to guide treatment. Clearly, the right ventricle is facing great challenges in sepsis. In addition to direct myocardial depression, the right ventricle is often exposed to acute elevations in afterload, as the result of acute lung injury, hypercapnia, acidosis, positive airway pressure from mechanical ventilation, sepsis-induced pulmonary arterial thrombosis, and LV dysfunction (10). However, by design the right ventricle is poorly adapted to compensate for these conditions in the acute setting.
Pulmonary artery catheter (PAC)-derived assessment of right ventricular ejection fraction (RVEF) has been associated with mortality in sepsis, albeit in a limited number of publications with emphasis on short-term mortality (11–14). The technique itself is considered to be reliable, but implementation is limited by its invasive nature.
The present study was aimed to explore the association between RVEF and long-term mortality in a noninterventional setting in patients with severe sepsis and septic shock. We hypothesized that patients with a reduced RVEF would have increased rates of 1-year all-cause mortality in comparison to patients with a normal RVEF. Secondary endpoints were the association between RVEF and ICU mortality and morbidity.
PATIENTS AND METHODS
This single-centre retrospective cohort study was performed in a tertiary teaching hospital with a closed-format, 20-bed, mixed ICU. A local ethical and scientific committee waived the need for informed consent in accordance with applicable laws (Regionale Toetsingscommissie Patiëntgebonden Onderzoek Leeuwarden, registration number nWMO 334). Patient data were deidentified before the study.
All patients admitted at the study site from January 2011 through December 2017 with suspected sepsis as the principal reason for ICU admission were eligible for assessment. Patients were included if they fulfilled the following criteria: 18 years of age or older; presence of severe sepsis or septic shock as defined by international criteria (15); and equipped with a PAC within 24 h after ICU admission. Septic shock was defined as severe sepsis with persistent hypotension (systolic blood pressure <90 mmHg) despite adequate volume resuscitation in the absence of other causes for hypotension and/or use of vasopressor agents (norepinephrine any dose, dopamine >5 μg/kg/min). The decision to insert a PAC was to the discretion of the attending physician. Exclusion criteria included: pregnancy, documented history of congenital heart disease, metastatic malignant disease with futile prognosis, recent acute coronary syndrome (<1 week ago), severe sepsis present for more than 24 h before ICU admission, and/or a malfunctioning PAC during the first 24 h after ICU admission.
Patients eligible for assessment were identified by an Acute Physiology and Chronic Health Evaluation II (APACHE II) diagnosis of sepsis. After this initial screening, the definitive inclusion was determined by the abovementioned criteria as extracted from available medical correspondence and clinical information. Demographic and clinical data were extracted from MediScore (version 126.96.36.19911, ItéMedical, Tiel, The Netherlands) and electronic health records saved in Metavision (version 5.45.62; iMDsoft, Needham, Mass) and Epic (1979–2018; Epic Systems Corporation, Verona, Wis). Comorbidities and reason for admission were determined in line with the national NICE-dictionary (16). The severity of illness was assessed as APACHE II and Sequential Organ Failure Assessment (SOFA) scores (17, 18). Mortality data were derived from the Municipal Personal Records database.
RVEF, cardiac index, stroke volume index (SVi), end-diastolic volume index (EDVi), mixed venous oxygen saturation (SvO2), and mean pulmonary artery pressure (mPAP) were derived from a 7.5 F Swan-Ganz CCOmbo V (Model 774F75; Edwards Lifesciences Corporation, Irvine, Calif) interfaced with a computerized monitoring system (Vigilance II CCO/SvO2/CEDV Monitor; Edwards Lifesciences Corporation). A thermal filament on the catheter enabled near-continuous measurement at a sample rate of 1 per minute. All PAC-derived measurements were averaged over the first 24 h after ICU admission, or until ICU discharge in case patients died or were transferred to the ward within 24 h after ICU admission. The start of measurements made by PAC was defined as the first time point where the RVEF was obtained after admission.
According to a previous publication, patients were separated a priori into subgroups according to their RVEF: group A: RVEF less than 20%, group B: RVEF 20% to 30%, and group C: RVEF more than 30% (19).
The primary endpoint of our study was 1-year all-cause mortality. Secondary endpoints included the correlation between RVEF and ICU mortality and morbidity (as determined by the length of stay [LOS] in the ICU; duration of mechanical ventilation; usage of vasopressor and inotropic agents [norepinephrine, dopamine, dobutamine, enoximone, and vasopressin]; and hemodynamic variables, including cardiac index, SvO2, and fluid balance).
Statistical analysis was performed using the Statistical Package for Social Science (SPSS version 24.0 for Windows; SPSS Inc., Chicago, Ill). Continuous data are described as median with interquartile range (IQR) and categorical data are described as counts with percentages (%). The Kolmogorov–Smirnov test was used to test for normal distribution. For continuous data, differences between groups were compared using a one-way analysis of variance or Kruskal–Wallis test for normally and not normally distributed data respectively. For categorical data, this comparison was made using the chi-square test, whereas Fisher exact test was performed when expected counts were 5 or less. Multivariate logistic regression analysis of the association between RVEF and mortality was performed using the backward likelihood ratio method. Baseline characteristics and outcome variables with a P <0.25 in the univariate analysis were included. For multicategorical variables, the group with the highest number of patients served as reference category. Outcomes are reported as odds ratio (OR) with corresponding 95% confidence intervals (CIs). Cumulative survival curves (censored endpoint at day 365) were estimated with the Kaplan–Meier procedure and the effect of RVEF on survival probability was compared between groups with the use of a log rank test. Subsequently, a Cox regression survival analysis was performed to correct for confounders identified by the multivariate analysis. Receiver-operating characteristic (ROC) curve analysis evaluated the discriminability of RVEF between survivors and nonsurvivors after 1 year expressed as area under the curve (AUC). All statistics were two-tailed and considered statistically significant if the P value was <0.05.
During the 7-year study period, 858 patients were found eligible for assessment. Ninety-eight patients were studied; 760 were excluded for various reasons (Fig. 1). No patients were lost to follow-up. Twenty-one patients were assigned to group A (RVEF <20%), 55 patients to group B (RVEF 20%–30%), and 22 patients to group C (RVEF >30%). Baseline characteristics are listed in Table 1. There were no significant differences between the three groups with regard to most of the baseline characteristics, including medical history, reason for admission, source of infection, and SOFA score. However, patients in group A had a significantly higher APACHE II score, had a higher incidence of septic shock, and were more often mechanically ventilated in comparison with other groups.
One-year all-cause mortality was 57% in group A, 18% in group B, and 23% in group C (P = 0.003; Table 2). Kaplan–Meier curves for estimated survival showed that 1-year all-cause mortality was significantly higher in patients with RVEF less than 20% compared with the other groups, whereas there was no significant difference between groups B and C (X2 = 14.00, P = 0.001; Fig. 2).
After adjustment for APACHE II score, chronic obstructive pulmonary disease, mechanical ventilation, mPAP, and septic shock, multivariate logistic regression analysis showed that RVEF less than 20% remained an independent predictor of 1-year all-cause mortality (OR 4.13; 95% CI, 1.28–13.35; P = 0.018). In an additional model, adjusting for the same abovementioned variables, RVEF as a continuous variable was also found to be an independent predictor of the primary endpoint (OR 0.92; 95% CI, 0.85–0.99; P = 0.018). In both models, APACHE II score remained independently predictive of outcome as well (Table 3).
In a Cox regression survival analysis, with correction for APACHE II score, the difference in survival remained significant between patients with RVEF less than 20% and the other groups (X2 = 20.51, P < 0.001).
ROC curve analysis revealed that RVEF as a continuous variable was moderately accurate as a discriminator of 1-year all-cause mortality (AUC 0.70; 95% CI, 0.57–0.83; P = 0.003; Fig. 3). The Youden index was 25%, as presented in the figure, with a sensitivity of 78% and a specificity of 68%. Repeating the multivariate logistic regression analysis using this cutoff value, confirmed RVEF less than 25% was an independent predictor of long-term mortality (OR 6.16; 95% CI, 2.12–17.90; P = 0.001).
Patients in group A (RVEF <20%) had a significantly higher ICU mortality rate compared with other groups (P = 0.009; Table 2). There were no significant differences across groups in LOS ICU and duration of mechanical ventilation. Although the use of norepinephrine was significantly higher among patients in groups A and B, their highest dose did not significantly differ. Patients with RVEF less than 20% had a significantly lower SvO2, cardiac index, SVi, and lowest arterial pH compared with the other groups. Furthermore, heart rate, mPAP, highest lactate, and fluid balance were significantly higher (Table 2). Subgroup analyses are presented in the Supplemental Digital Content section (see Figures, Supplemental Digital Content 1 and 2, which shows Kaplan–Meier curves for estimated survival based on a history of cardiac disease and presence of pulmonary sepsis, http://links.lww.com/SHK/A916 and http://links.lww.com/SHK/A917).
In the present study, RVEF, both as a categorical variable (RVEF <20%) and as a continuous variable, was an independent predictor of 1-year all-cause mortality in patients with severe sepsis and septic shock. From the Kaplan–Meier curve, it becomes clear that these findings are fully attributable to differences in short-term mortality. RVEF less than 20% was additionally associated with greater severity of illness, suggested by a higher APACHE II score, more septic shock and a higher incidence of mechanical ventilation, and higher indices of circulatory failure, as indicated by a higher heart rate, lower cardiac index, lower SvO2, more use of norepinephrine, lower arterial pH, higher lactate, and a higher fluid balance. The optimal cutoff value of RVEF for predicting an adverse outcome was 25%.
To the best of our knowledge, this was the first study that evaluated the association of PAC-derived RVEF with long-term mortality. Previous studies evaluated RV function with echocardiography in patients with severe sepsis and septic shock. However, inconsistent results on the relation between RV function and long-term mortality were reported (1, 20, 21). In a recent study (20), isolated RV dysfunction was found to be independently associated with worse 1-year survival (hazard ratio [HR] 1.6; 95% CI, 1.2–2.1; P = 0.002). Remarkably, this was not the case when patients with combined RV/LV dysfunction were added. In contrast to our study, RV function was assessed using a multimodality approach, which did not include RVEF, and RV function was only evaluated once in the first 72 h after ICU admission by transthoracic echocardiography. The prognostic importance of RV dysfunction was also reported in an earlier study (21), which demonstrated that RV free wall longitudinal strain, assessed by speckle tracking echocardiography, was an independent predictor of outcome at 6 months (OR 1.1; 95% CI, 1.0–1.3; P = 0.020). Multivariate analysis showed that severe RV free wall longitudinal strain dysfunction was an even stronger predictor (OR 11.9; 95% CI, 1.9–232; P = 0.030). Interestingly, the Kaplan–Meier curve for 1-year survival showed strong similarities with our study. Patients with normal RV strain function and mild–moderate RV strain dysfunction had similar 1-year survival estimates, whereas patients with severe RV strain dysfunction all died before 6 months. In contrast to our study, RV measurements were only performed once within 24 h after ICU admission. A third study found no difference in 1-year mortality between patients with normal myocardial function and those with RV dysfunction (1). This result may be explained by the fact that only 6 patients (6%) had severe RV dysfunction, according to a unique multimodal approach which did not include RVEF.
A finding of interest from the present study was that RVEF was also associated with short-term mortality. This is in concordance with several other studies that associated a reduced PAC-derived RVEF with higher short-term mortality (11–14). These studies were, however, limited by their small sample sizes and single-point measurements. Interestingly, studies that were unable to find this association made use of another assessment modality. Two studies performed RVEF measurements with traditional transthoracic echocardiography (22, 23). However, it remains challenging to evaluate the right ventricle with this method, due to the complex geometry, lack of clear landmarks, and poor visibility during mechanical ventilation (24, 25). An early study performed radionuclide ventriculography at the patient's bedside (26). Remarkably, they found initial RVEF to be lower in survivors. This finding has not been reproduced thus far and might be the result of their heterogenic study population, which included both children and adults and consisted mainly of patients with cancer.
It is of note that the percentage of patients on mechanical ventilation in our study were significantly higher among patients with the lowest RVEF. This may be due to the intervention itself, that is, positive pressure ventilation increases RV afterload. Alternatively, both mechanical ventilation and RV dysfunction are concomitant markers of the same disease process. The design of the study does not allow to unravel this mechanism.
With respect to the other secondary endpoints, it becomes clear that patients with the lowest RVEF had a lower cardiac index and SvO2, and norepinephrine was administered more often to maintain a minimal perfusion pressure. This suggests that the observed effects extend beyond an isolated loss of RV efficiency, but rather represent a true inability of the heart to maintain sufficient circulation. The fact that the fluid balance is also significantly more positive in patients with the lowest RVEF may either represent a cause or an effect. In other words, inadequate fluid load may cause RV failure, but RV failure may also lead to extra fluid administration. These data are in line with a previous publication in which a low RVEF was identified as an important determinant for the loss of fluid responsiveness and the need for vasopressor therapy in early septic shock states (27). Lastly, it must be stressed that 51% of the patients had a mean PAP at least 25 mmHg, fulfilling the criteria for pulmonary hypertension (28). This underlines the idea that increased RV afterload may play a role, but cannot be seen as sole determinant of RV failure. This finding is in concordance with previous reports (11, 29), which suggested that RV dysfunction may be caused by abnormalities in RV afterload in some patients and depressed myocardial contractility in others.
The design of the study had inherent limitations. The main weakness of this study was the absence of LV measurements. Therefore, the impact of LV function on the results remains unclear. Furthermore, this is a highly selected population because only patients that required insertion of a PAC were included. Consequently, our study population was not representative of all patients with severe sepsis and septic shock. Another limitation was the potential variability in time of disease onset because it is a challenge to pinpoint the exact beginning of sepsis. However, as soon as patients were admitted to the ICU, PAC instrumentation was initiated within hours. Finally, this study was limited by its single-centre retrospective nature. This study design is susceptible to the effects of bias due to the inability to control for potential confounders and does not allow to elaborate on a cause–effect relationship. The strength of the study is inherently related to the invasive, yet robust measurement of RV function in a near-continuous way and the excellent follow-up on long-term mortality.
RVEF was independently associated with 1-year all-cause mortality in a highly selected group of patients with severe sepsis and septic shock. In addition, a low RVEF was associated with an increase in signs of circulatory failure, suggesting maladaptation. Future studies are needed to establish a potential cause-and-effect relationship between RVEF and mortality, and to confirm potential beneficial therapeutic strategies.
1. Pulido JN, Afessa B, Masaki M, Yuasa T, Gillespie S, Herasevich V, Brown DR, Oh JK. Clinical spectrum, frequency, and significance of myocardial dysfunction in severe sepsis and septic shock. Mayo Clin Proc
87 (7):620–628, 2012.
2. Beesley SJ, Weber G, Sarge T, Nikravan S, Grissom CK, Lanspa MJ, Shahul S, Brown SM. Septic cardiomyopathy. Crit Care Med
46 (4):625–634, 2018.
3. Ehrman RR, Sullivan AN, Favot MJ, Sherwin RL, Reynolds CA, Abidov A, Levy PD. Pathophysiology, echocardiographic evaluation, biomarker findings, and prognostic implications of septic cardiomyopathy: a review of the literature. Crit Care
22 (1):112, 2018.
4. Sevilla Berrios RA, O’Horo JC, Velagapudi V, Pulido JN. Correlation of left ventricular systolic dysfunction determined by low ejection fraction and 30-day mortality in patients with severe sepsis and septic shock: a systematic review and meta-analysis. J Crit Care
29 (4):495–499, 2014.
5. Sanfilippo F, Corredor C, Arcadipane A, Landesberg G, Vieillard-Baron A, Cecconi M, Fletcher N. Tissue Doppler assessment of diastolic function and relationship with mortality in critically ill septic patients: a systematic review and meta-analysis. Br J Anaesth
119 (4):583–594, 2017.
6. Huang SJ, Nalos M, McLean AS. Is early ventricular dysfunction or dilatation associated with lower mortality rate in adult severe sepsis and septic shock? A meta-analysis. Crit Care
17 (3):R96, 2013.
7. Harmankaya A, Akilli H, Gul M, Akilli NB, Ergin M, Aribas A, Cander B. Assessment of right ventricular functions in patients with sepsis, severe sepsis and septic shock and its prognostic importance: a tissue Doppler study. J Crit Care
28 (6):1111, 2013.
8. Singh RK, Kumar S, Nadig S, Baronia AK, Poddar B, Azim A, Gurjar M. Right heart in septic shock: prospective observational study. J Intensive Care
9. Cirulis MM, Huston JH, Sardar P, Suksaranjit P, Wilson BD, Hatton ND, Liou TG, Ryan JJ. Right-to-left ventricular end diastolic diameter ratio in severe sepsis and septic shock. J Crit Care
10. Chan CM, Klinger JR. The right ventricle in sepsis. Clin Chest Med
29 (4):661–676, 2008. ix.
11. Dhainaut JF, Lanore JJ, de Gournay JM, Huyghebaert MF, Brunet F, Villemant D, Monsallier JF. Right ventricular dysfunction in patients with septic shock. Intensive Care Med
14: (Suppl. 2): 488–491, 1988.
12. Vincent JL, Reuse C, Frank N, Contempré B, Kahn RJ. Right ventricular dysfunction in septic shock: assessment by measurements of right ventricular ejection fraction using the thermodilution technique. Acta Anaesthesiol Scand
33 (1):34–38, 1989.
13. Vincent JL, Gris P, Coffernils M, Leon M, Pinsky M, Reuse C, Kahn RJ. Myocardial depression characterizes the fatal course of septic shock. Surgery
111 (6):660–667, 1992.
14. Liu D, Du B, Long Y, Zhao C, Hou B. Right ventricular function of patients with septic shock: clinical significance. Zhonghua Wai Ke Za Zhi
38 (7):488–492, 2000.
15. Levy MM, Fink MP, Marshall JC, Abraham E, Angus D, Cook D, Cohen J, Opal SM, Vincent JL, Ramsay G. International Sepsis Definitions Conference. 2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference. Intensive Care Med
29 (4):530–538, 2003.
16. Stichting NICE: Sepsis registration, version 2017. Available at: https://www.stichting-nice.nl/download/?Groep=Data%20Dictionary&Taal=Nederlands
(2018). Accessed February 16, 2019.
17. Vincent JL, Moreno R, Takala J, Willatts S, De Mendonça A, Bruining H, Reinhart CK, Suter PM, Thijs LG. The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure. On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med
22 (7):707–710, 1996.
18. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a severity of disease classification system. Crit Care Med
13 (10):818–829, 1985.
19. Bootsma IT, de Lange F, Koopmans M, Haenen J, Boonstra PW, Symersky T, Boerma EC. Right ventricular function after cardiac surgery is a strong independent predictor for long-term mortality. J Cardiothorac Vasc Anesth
31 (5):1656–1662, 2017.
20. Vallabhajosyula S, Kumar M, Pandompatam G, Sakhuja A, Kashyap R, Kashani K, Gajic O, Geske JB, Jentzer JC. Prognostic impact of isolated right ventricular dysfunction in sepsis and septic shock: an 8-year historical cohort study. Ann Intensive Care
7 (1):94, 2017.
21. Orde SR, Pulido JN, Masaki M, Gillespie S, Spoon JN, Kane GC, Oh JK. Outcome prediction in sepsis: speckle tracking echocardiography based assessment of myocardial function. Crit Care
18 (4):R149, 2014.
22. Landesberg G, Jaffe AS, Gilon D, Levin PD, Goodman S, Abu-Baih A, Beeri R, Weissman C, Sprung CL, Landesberg A. Troponin elevation in severe sepsis and septic shock: the role of left ventricular diastolic dysfunction and right ventricular dilatation∗. Crit Care Med
42 (4):790–800, 2014.
23. Rolando G, Espinoza EDV, Avid E, Welsh S, Pozo JD, Vazquez AR, Arzani Y, Masevicius FD, Dubin A. Prognostic value of ventricular diastolic dysfunction in patients with severe sepsis and septic shock. Rev Bras Ter Intensiva
27 (4):333–339, 2015.
24. Orde S, Slama M, Yastrebov K, Mclean A, Huang S. College of Intensive Care Medicine of Australia and New Zealand [CICM] Ultrasound Special Interest Group [USIG]. Subjective right ventricle assessment by echo qualified intensive care specialists: assessing agreement with objective measures. Crit Care
23 (1):70, 2019.
25. Haddad F, Hunt SA, Rosenthal DN, Murphy DJ. Right ventricular function in cardiovascular disease, part I: anatomy, physiology, aging, and functional assessment of the right ventricle. Circulation
117 (11):1436–1448, 2008.
26. Parker MM, McCarthy KE, Ognibene FP, Parrillo JE. Right ventricular dysfunction and dilatation, similar to left ventricular changes, characterize the cardiac depression of septic shock in humans. Chest
97 (1):126–131, 1990.
27. Redl G, Germann P, Plattner H, Hammerle A. Right ventricular function in early septic shock states. Intensive Care Med
19 (1):3–7, 1993.
28. Galiè N, Humbert M, Vachiery JL, Gibbs S, Lang I, Torbicki A, Simonneau G, Peacock A, Vonk Noordegraaf A, Beghetti M, et al. ESC Scientific Document Group. 2015 ESC/ERS Guidelines for the diagnosis and treatment of pulmonary hypertension: The Joint Task Force for the Diagnosis and Treatment of Pulmonary Hypertension of the European Society of Cardiology (ESC) and the European Respiratory Society (ERS): Endorsed by: Association for European Paediatric and Congenital Cardiology (AEPC), International Society for Heart and Lung Transplantation (ISHLT). Eur Heart J
37 (1):67–119, 2016.
29. Kimchi A, Ellrodt AG, Berman DS, Riedinger MS, Swan HJ, Murata GH. Right ventricular performance in septic shock: a combined radionuclide and hemodynamic study. J Am Coll Cardiol
4 (5):945–951, 1984.