Table 2 Summary of selected articles on septic cardiomyopathy

Left ventricle

Systolic

EF

Sevilla Berrios et al. (2014) [33]

Meta-analysis

585

To evaluate the significance of reduced LVEF in patients with severe sepsis and septic shock. Primary outcome was association between depressed LVEF and 30-day mortality

Depressed LVEF had a sensitivity of 52% (95% CI 29–73%) and specificity of 63% (95% CI 53–71%) for mortality and was therefore not a sensitive nor specific predictor of mortality

Huang et al. (2013) [32]

Meta-analysis

762

To evaluate the association of both reduced LVEF and increased LV dimensions with mortality in patients with severe sepsis and septic shock

No significant difference in LVEF and LV dimensions in survivors vs non-survivors

Jardin et al.

(1999) [24]

Single-center prospective cohort study

90

To evaluate changes in LV function, including LVEF and LV volumes, during volume resuscitation in patients with septic shock

LVEF was depressed in all patients. LV parameters were additionally unaffected by fluid loading

Parker et al. (1984) [1]

Single-center prospective cohort study

20

To evaluate cardiac function in septic shock

10/20 patients (50%) had depressed LVEF (< 0.40). Mean LVEF was lower among survivors (LVEF 0.32 ± 0.04) when compared to non-survivors. Mean ESV and EDV were increased in survivors

GLS

Boissier et al.

(2017) [17]

Single-center prospective cohort study/ICU

132

To evaluate the role of GLS, LVEF, and TDI in patients with septic shock. Primary outcome was the role of loading conditions on evaluation of cardiac contractility

GLS was impaired in a majority of the patients (> 70%); however, feasibility was limited (< 50%)

Chang et al.

(2015) [39]

Multi-center prospective cohort study/ICU

111

To evaluate LV function, as well as the prognostic value of GLS, in septic patients. Primary outcome was both ICU and hospital mortality

GLS is an independent prognostic indicator of ICU mortality. Patients with GLS ≥ − 13% had higher ICU mortality rates (HR 4.34; p < 0.001)

De Geer et al.

(2015) [43]

Single-center prospective cohort study/ICU

50

To evaluate GLS in patients with septic shock. Primary outcomes were mortality at 30 and 90 days

GLPS did not correlate between survivors and non-survivors and therefore could not be used to predict mortality

Innocenti et al.

(2016) [41]

Single-center prospective cohort study/ED observation unit

147

To evaluate LVEF and GLS in septic patients. Primary outcome was all-cause mortality at 7 days

LVEF is not an independent indicator of prognosis

Kalam et al. (2014) [37]

Meta-analysis

5721

To assess if GLS is a more accurate predictor of cardiovascular outcome compared to LVEF. Primary outcome was all-cause mortality. Secondary outcome was composite endpoint including cardiac death, malignant arrhythmia, and hospitalization

GLS is a better predictor of adverse outcomes (HR 0.50; p < 0.002) and mortality (HR 1.62; p = 0.009) than LVEF (HR 0.81; p = 0.572)

Ng et al.

(2016) [38]

Case–control study/ICU

62

To evaluate the role of GLS in the diagnosis of SMD. Primary outcome was to compare GLS values in patients with septic shock compared to patients with only sepsis

There was a significant difference in GLS values (− 14.5 vs –18.3%, p < 0.001) between patients with septic shock and sepsis. LVEF was not statistically significant between patients with septic shock and patients with sepsis

Orde et al.

(2014) [42]

Single-center prospective cohort study/ICU

60

To evaluate GLS in patients with severe sepsis or septic shock. Primary outcomes were mortality at 30 days and 6 months

No difference in mortality for LV GLS or GLS rate in survivors compared with non-survivors at 30 days or 6 months

Palmieri et al.

(2015) [40]

Single-center prospective cohort study/ED observation unit

115

To evaluate LV EF and peak GLS in patients with sepsis and septic shock. Primary outcome was death by any cause at 28 days from hospitalization

Abnormal GLS correlates significantly with mortality rate at 28 days. GLS values close to 0 demonstrated a higher mortality (HR 1.16%; p = 0.05).

Zaky et al.

(2016) [44]

Single-center prospective cohort study/ICU

54

To evaluate LVLS in patients with sepsis or septic shock. Primary outcomes were mechanical ventilation, ICU and hospital length of stay, and in-hospital mortality

Global LVLS was not associated with rates of mechanical ventilation, ICU or hospital length of stay, or in-hospital mortality

Systolic mitral annular velocity (S′)

Chang et al.

(2015) [39]

Multi-center prospective cohort study/ICU

111

To evaluate LV function, as well as the prognostic value of GLS, in septic patients. Primary outcome was both ICU and hospital mortality

There was no statistically significant difference in S′ between ICU non-survivors compared to survivors (11.0 ± 4.3 vs 11.4 ± 4.0; p < 0.66)

Weng et al. (2012) [49]

Single-center prospective cohort study/ICU

61

To evaluate the prognostic significance of several TDI variables, including systolic mitral annular velocity, S′, in patients with septic shock. Primary outcome was all-cause mortality

Non-survivors had a higher S′ when compared to survivors (11.0 vs 7.8 cm/s; p < 0.0001). Patients with S′ > 9 cm/s had a higher mortality rate (75 versus 17%; p < 0.0001). S′ > 9 cm/s had SN 75% and SP 86% to predict 90-day mortality

Weng et al. (2013) [50]

Single-center prospective cohort study/ICU

51

To evaluate LV longitudinal systolic dysfunction and LV intraventricular systolic asynchrony assessed by TDI in patients with septic shock and normal LVEF. Primary outcome was all-cause mortality at 28 days

Normal EF, LV longitudinal systolic dysfunction and LV systolic asynchrony assessed by TDI within 24 h of onset of septic shock were associated with improved mortality at 28 days

MAPSE

Zhang et al. (2017) [65]

Case-control study/ICU

45

To evaluate LVEF, MAPSE, Sa, and TAPSE in patients with septic shock. Primary outcome was sepsis

MAPSE values were significantly lower in septic patients when compared to non-septic patients (p ≤ 0.001)

MPI

Nizamuddin et al. (2017) [78]

Single-center prospective cohort study/ICU

47

To assess if changes in LV MPI were associated with higher 90-day mortality in patients with severe sepsis. Primary outcome was all-cause mortality

Decline in MPI over the initial 24-h study period was associated with higher mortality at 90 days (p = 0.04)

Diastolic

e’ and

E/e’

Brown et al.

(2012) [52]

Single-center prospective cohort study/ICU

78

To evaluate whether severity of diastolic dysfunction predicts mortality in patients with severe sepsis or septic shock. Primary outcome was mortality at 28 days

Grade I diastolic dysfunction was associated with increased mortality; grades II/III were not associated with increased mortality

Landesberg et al. (2012) [54]

Single-center prospective cohort study/ICU

262

To evaluate the association between diastolic dysfunction and mortality in severe sepsis and septic shock. Primary outcomes were in-hospital mortality and overall mortality at 6 months to 2 years

Decreased septal e’ or increased septal E/e’ were the strongest independent predictors of mortality (HR 0.76, p ≤ 0.001 and HR 1.08, p ≤ 0.001, respectively)

Rolando et al.

(2015) [57]

Single-center prospective cohort study/ICU

53

To evaluate the prognostic significance of myocardial dysfunction, including E/e’ ratio, in patients with severe sepsis and septic shock. Primary outcome was hospital mortality

E/e’ is an independent predictor of hospitality mortality (OR = 1.36; p = 0.02). An E/e’ > 11 had a sensitivity of 50% and specificity of 94% for predicting ICU mortality

Sanfilippo et al. (2017) [59]

Meta-analysis

1507

To evaluate the association of e’ and E/e’ with mortality in patients with severe sepsis or septic shock

A significant association was found between mortality and both a lower e’ (SC 0.33; 95% CI 0.05, 0.62; p = 0.02) and higher E/e’ (SC 0.33; 95% CI – 0.57, − 0.10; p = 0.006) in patients with severe sepsis and/or septic shock. There was high overall heterogeneity in both e’ and E/e’ analysis

Sturgess et al.

(2010) [56]

Single-center prospective cohort study/ICU

21

To evaluate the prognostic significance of TDI and cardiac biomarkers in septic shock. Primary outcome was hospital mortality

E/e’ is an independent predictor of hospital survival and is a better prognosticator than cardiac biomarkers. E/e’ was greater in non-survivors than survivors (15.32 ± 2.74 vs 9.05 ± 2.75, respectively; p = 0.0002)

Lanspa et al. (2016) [60]

Single-center prospective cohort study/ICU

167

To compare the feasibility and prognostic significance of a simplified definition of diastolic dysfunction (using e’ and E/e’) with 2009 ASE guidelines. Primary outcome was 28-day mortality

Simplified definition had better feasibility (87 vs 35%); similar clinical outcomes between groups suggesting limited utility of LAVI and DT in this setting

Right ventricle

Systolic

TAPSE

Gajanana et al.

(2015) [64]

Single-center prospective cohort study/ICU

120

To evaluate the prognostic value of TAPSE in patients with critical illness

A reduced TAPSE measurement (< 2.4 cm) was correlated with increased in-hospital mortality (χ(2) = 4.6, P = 0.03) and a longer length of hospital stay

TAPSE

TDI

RV FAC

Vallabhajosyula et al. (2017) [67]

Single-center retrospective cohort study/ICU

388

To evaluate the prognostic significance RV dysfunction in patients with severe sepsis and septic shock. Primary outcome was 1-year survival

Isolated RV dysfunction is an independent predictor of 1-year survival (HR 1.6; p = 0.002). Combined RV/LV dysfunction was not an independent predictor of 1-year survival (HR 0.9; p = 0.52)