Coefficient of Variation of Coarsely Sampled Heart Rate is Associated With Early Vasopressor Independence in Severe Sepsis and Septic Shock

The Development and Validation of a Machine Learning Model to Predict Bacteremia and Fungemia in Hospitalized Patients Using Electronic Health Record Data

OBJECTIVES

Bacteremia and fungemia can cause life-threatening illness with high mortality rates, which increase with delays in antimicrobial therapy. The objective of this study is to develop machine learning models to predict blood culture results at the time of the blood culture order using routine data in the electronic health record.

DESIGN

Retrospective analysis of a large, multicenter inpatient data.

SETTING

Two academic tertiary medical centers between the years 2007 and 2018.

SUBJECTS

All hospitalized patients who received a blood culture during hospitalization.

INTERVENTIONS

The dataset was partitioned temporally into development and validation cohorts: the logistic regression and gradient boosting machine models were trained on the earliest 80% of hospital admissions and validated on the most recent 20%.

MEASUREMENTS AND MAIN RESULTS

There were 252,569 blood culture days-defined as nonoverlapping 24-hour periods in which one or more blood cultures were ordered. In the validation cohort, there were 50,514 blood culture days, with 3,762 cases of bacteremia (7.5%) and 370 cases of fungemia (0.7%). The gradient boosting machine model for bacteremia had significantly higher area under the receiver operating characteristic curve (0.78 [95% CI 0.77-0.78]) than the logistic regression model (0.73 [0.72-0.74]) (p < 0.001). The model identified a high-risk group with over 30 times the occurrence rate of bacteremia in the low-risk group (27.4% vs 0.9%; p < 0.001). Using the low-risk cut-off, the model identifies bacteremia with 98.7% sensitivity. The gradient boosting machine model for fungemia had high discrimination (area under the receiver operating characteristic curve 0.88 [95% CI 0.86-0.90]). The high-risk fungemia group had 252 fungemic cultures compared with one fungemic culture in the low-risk group (5.0% vs 0.02%; p < 0.001). Further, the high-risk group had a mortality rate 60 times higher than the low-risk group (28.2% vs 0.4%; p < 0.001).

CONCLUSIONS

Our novel models identified patients at low and high-risk for bacteremia and fungemia using routinely collected electronic health record data. Further research is needed to evaluate the cost-effectiveness and impact of model implementation in clinical practice.

β-bloquants dans la prise en charge du choc septique

Les adrénorécepteurs α et en particulier β sont les principales cibles de l’adrénaline et de la noradrénaline libérées par le système sympathique activé. Durant le choc septique, la dysautonomie est une stimulation prolongée à un haut niveau d’intensité du système nerveux sympathique à l’origine d’une altération de la contractilité, de la vasoréactivité et d’une immunodépression. Ainsi, l’administration précoce d’un traitement β-bloquant lors du choc septique pourrait pondérer les effets délétères de cette surstimulation sympathique. Néanmoins, si les preuves expérimentales sont en faveur de cette approche, l’accumulation des preuves cliniques reste encore insuffisante.

Proposed mechanisms of relative bradycardia

Relative bradycardia is the term used to describe the mechanism where there is dissociation between pulse and temperature. This finding is important to recognize since it may provide further insights into the potential underlying causes of disease. There is no known proposed mechanism to explain this phenomenon. We hypothesize that relative bradycardia is the central mechanism reflecting and influenced potentially by the direct pathogenic effect on the sinoatrial node as well as cross-talk between the autonomic nervous system and immune system. Cardiac pacemaker cells may act as a target for inflammatory cytokines leading to alteration in heart rate dynamics or their responsiveness to neurotransmitters during systemic inflammation. These factors account for the important role of how the host response to infectious and non-infectious causes influences the appearance of relative bradycardia. We propose several methods that may be useful to confirm the proposed theoretical framework to further enhance our understanding of this paradoxical phenomenon. This includes measuring, during the episode of relative bradycardia, proinflammatory and anti-inflammatory cytokines, monitoring heart rate variability (HRV), and assessing underlying comorbidities and outcomes in patients with the same disease.

Systolic blood pressure variability in patients with early severe sepsis or septic shock: a prospective cohort study

Background

Severe sepsis and septic shock are often lethal syndromes, in which the autonomic nervous system may fail to maintain adequate blood pressure. Heart rate variability has been associated with outcomes in sepsis. Whether systolic blood pressure (SBP) variability is associated with clinical outcomes in septic patients is unknown. The propose of this study is to determine whether variability in SBP correlates with vasopressor independence and mortality among septic patients.

Methods

We prospectively studied patients with severe sepsis or septic shock, admitted to an intensive care unit (ICU) with an arterial catheter. We analyzed SBP variability on the first 5-min window immediately following ICU admission. We performed principal component analysis of multidimensional complexity, and used the first principal component (PC₁) as input for Firth logistic regression, controlling for mean systolic pressure (SBP) in the primary analyses, and Acute Physiology and Chronic Health Evaluation (APACHE) II score or NEE dose in the ancillary analyses. Prespecified outcomes were vasopressor independence at 24 h (primary), and 28-day mortality (secondary).

Results

We studied 51 patients, 51% of whom achieved vasopressor independence at 24 h. Ten percent died at 28 days. PC₁ represented 26% of the variance in complexity measures. PC₁ was not associated with vasopressor independence on Firth logistic regression (OR 1.04; 95% CI: 0.93–1.16; p = 0.54), but was associated with 28-day mortality (OR 1.16, 95% CI: 1.01–1.35, p = 0.040).

Conclusions

Early SBP variability appears to be associated with 28-day mortality in patients with severe sepsis and septic shock.

Electronic supplementary material

The online version of this article (doi:10.1186/s12871-017-0377-4) contains supplementary material, which is available to authorized users.

Roles of PD-1, Tim-3 and CTLA-4 in immunoregulation in regulatory T cells among patients with sepsis

OBJECTIVE

This study aims to elucidate the roles of PD-1, Tim-3 and CTLA-4 in sepsis.

METHODS

Sepsis patients (n = 182) were selected as sepsis group and divided into three subgroups: mild sepsis group, severe sepsis group and septic shock group; 185 healthy volunteers were enrolled as control group. Flow cytometry and blood routine examination were performed for T lymphocytes and surface co-stimulatory molecules expressions. Pearson correlation test was applied for the correlation of co-stimulatory molecules expressions on T lymphocytes with critical illness in sepsis. Logistic regression analysis was conducted for risk factors in sepsis.

RESULTS

Heart rate and WBC in subgroups were higher than control group (P < 0.05). The differences in APACHE II, SAP II and SOFA score among subgroups were statistically significant (P < 0.05). Compared with control group, lymphocyte ratio and percentage of CD4(+) T cells reduced in subgroups (P < 0.05). The differences in expression levels of CD4(+)PD-1(+), CD8(+)PD-1(+), and CD8(+)CTLA-4(+) showed statistical significances (P < 0.05). Apparently, expression levels of CD4(+)TIM-3(+), CD8(+)TIM-3(+), CD4(+)PD-1(+), CD8(+)PD-1(+), and CD4(+)CTLA-4(+) were positively correlated with APACHE II score (all P < 0.05). Logistic regression analysis showed that heart rate and expression level of CD4(+)PD-1(+) might be risk factors while the percentage of CD4(+) T cells might be a protective factor for sepsis (P < 0.05).

CONCLUSION

PD-1 aggravates immune responses consistent with promotion of T cell exhaustion in sepsis. Expression level of CD4(+)PD-1(+) and heart rate are potential risk factors while percentage of CD4(+) T cells is a possible protective factor for sepsis.