Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Prognostic implications of ultra-short heart rate variability indices in hospitalized patients with infective endocarditis

  • Shay Perek,

    Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Validation, Visualization, Writing – original draft, Writing – review & editing

    Affiliations Department of Internal Medicine A, Rambam Health Care Campus, Haifa, Israel, Department of Emergency Medicine, Rambam Health Care Campus, Haifa, Israel, The Ruth and Bruce Rappaport Faculty of Medicine, The Technion–Israel Institute of Technology, Haifa, Israel

    ⨯
  • Udi Nussinovitch,

    Roles Conceptualization, Methodology, Resources, Writing – original draft, Writing – review & editing

    Affiliations Department of Cardiology, Wolfson Medical Center, Holon, Israel, Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel

    ⨯
  • Neta Sagi,

    Roles Conceptualization, Data curation, Investigation, Writing – original draft, Writing – review & editing

    Affiliation Department of Pediatrics A, Rambam Health Care Campus, Haifa, Israel

    ⨯
  • Yori Gidron,

    Roles Conceptualization, Methodology, Resources, Writing – original draft, Writing – review & editing

    Affiliation Department of Nursing, Faculty of Social Welfare and Health Sciences, University of Haifa, Haifa, Israel

    ⨯
  • Ayelet Raz-Pasteur

    Roles Conceptualization, Data curation, Formal analysis, Resources, Supervision, Validation, Writing – original draft, Writing – review & editing

    a_raz@rambam.health.gov.il

    Affiliations Department of Internal Medicine A, Rambam Health Care Campus, Haifa, Israel, The Ruth and Bruce Rappaport Faculty of Medicine, The Technion–Israel Institute of Technology, Haifa, Israel

    ⨯

Abstract

Background

Infective endocarditis (IE) is a disease that poses a serious health risk. It is important to identify high-risk patients early in the course of their treatment. In the current study, we evaluated the prognostic value of ultra-short heart-rate variability (HRV), an index of vagal nerve activity, in IE.

Methods

Retrospective analysis was performed on adult patients admitted to a tertiary hospital due to IE. A logistic regression (LR) was used to determine whether clinical, laboratory, and HRV parameters were predictive of specific clinical features (valve type, staphylococcal infection) or severe short-term complications (cardiac, metastatic infection, and death). The accuracy of the model was evaluated through the measurement of the area under the curve (AUC) of the receiver operating characteristic curve (ROC). An analysis of survival was conducted using Cox regression. A number of HRV indices were calculated, including the standard deviation of normal heart-beat intervals (SDNN) and the root mean square of successive differences (RMSSD).

Results

75 patients, aged 60.3(±18.6) years old, were examined. When compared with published age- and gender-adjusted HRV norms, SDNN and RMSSD were found to be relatively low in our cohort (75%-76% lower than the median; 33%-41% lower than the 2nd percentile). 26(34.6%) patients developed a metastatic infection, with RMSSD<7.03ms (adjusted odds ratio (aOR) 9.340, p = 0.002), incorporated in a multivariate LR model (AUC 0.833). Furthermore, 27(36.0%) patients were diagnosed with Staphylococcus IE, with SDNN<4.92ms (aOR 5.235, p = 0.004), a major component of the multivariate LR model (AUC 0.741). Multivariate Cox regression survival model, included RMSSD (HR 1.008, p = 0.012).

Conclusion

SDNN, and particularly RMSSD, derived from ultra-short ECG recordings, may provide prognostic information about patients presenting with IE.

Citation: Perek S, Nussinovitch U, Sagi N, Gidron Y, Raz-Pasteur A (2023) Prognostic implications of ultra-short heart rate variability indices in hospitalized patients with infective endocarditis. PLoS ONE 18(6): e0287607. https://doi.org/10.1371/journal.pone.0287607

Editor: Atnafu Mekonnen Tekleab, St Paul’s Hospital Millennium Medical College, ETHIOPIA

Received: January 31, 2023; Accepted: June 8, 2023; Published: June 23, 2023

Copyright: © 2023 Perek et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: The dataset generated and analyzed during the current study are available in the figshare repository, https://doi.org/10.6084/m9.figshare.21679742.v1.

Funding: The authors received no specific funding for this work.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Infective endocarditis (IE) is associated with high mortality rates and severe complications, including embolic events, valvular destruction and arrhythmias [1]. Therefore, it is crucial to stratify patients’ risks at an early stage of their treatment. IE’s prognosis depends on clinical factors, lab testing, imaging findings, or a combination of them [2]. Several hematological, chemical, and inflammatory markers have been suggested for the evaluation of prognosis in patients with IE, including white blood cell count levels, C-reactive protein (CRP) levels, procalcitonin levels, neutrophil-lymphocyte ratios [3–5], in addition to complete blood count indices and N-terminal pro-B-type natriuretic peptide levels [6–8]. More complex assessments of cytokines and cell-derived macrovesicles may also have potential value in diagnosis and prognostication of IE patients [9–11]. Several studies have demonstrated that echocardiographic findings are effective predictors of mortality and embolic events [12]. Additionally, echo-guided decision-making for early surgery in IE patients has been shown to be a cost-effective management strategy [13]. Cardiac computed tomography may be an adjunctive test to echocardiography in patients with IE who are undergoing surgery [14].

While the majority of these methods are unavailable at the time of patient arrival, an electrocardiogram (ECG) is a routinely performed, non-invasive test, which is usually performed during the course of an emergency department (ED) visit. Electrocardiographic changes are prevalent in IE and are indicative of invasive disease, thereby have the ability to predict high morbidity and mortality [15]. Conduction abnormalities can also suggest that the disease is extending into the peri-valvular region [16] or signal the presence of complicated aortic valve disease [17, 18].

However, neuro-cardiac regulatory processes could also play a role in IE. The dynamic modulation of heart rate (HR) is considered a surrogate of the interaction between the sympathetic and parasympathetic nervous systems [19], and is quantified by the fluctuation or variability in the time intervals between normal heartbeats (i.e.; heart rate variability [HRV]). HRV decreases under stress, either emotional or physical, increases in rest and is considered an important noninvasive marker. It could be used to assess the function of the autonomic nervous system (ANS) as well as to determine if a cardiac response to autonomic modulation of heart rate is appropriate [20]. An inverse relation between HRV and inflammatory surrogates such as CRP has been established with 24 hour HRV. Moreover, HRV and CRP were found to have a synergistic effect in the prediction of death and myocardial infarction in ambulatory subjects with no apparent heart disease [21]. An increased mortality rate was associated with a lower HRV indices, calculated from short-term recordings, in patients with sepsis [22]. Similar findings were demonstrated in inflammatory bowel disease patients [23]. Interactions between autonomic innervation and immune responses are complex and intriguing. The vagus reflexively inhibits inflammation by activating the hypothalamic-pituitary-adrenal axis resulting with cortisol secretion [24]. Additionally, the vagus inhibits inflammation by a vagal-to-sympathetic innervation of the spleen, where unique T-cells signal macrophages to stop synthesizing pro-inflammatory cytokines [25]. In contrast to its anti-inflammatory role, vagus nerve stimulation could also increase cytotoxic T-cells, as well as natural killer cells [26].

Ultra short HRV analysis, focuses on computation of HRV indices from recordings lasting less than 5 minutes [27]. Several studies have demonstrated strong correlations between certain ultra-short HRV indices and parameters derived from longer HRV recordings, especially for time-domain indices [28–34]. Additionally, several ultra-short HRV parameters have been shown to be prognostic in patients admitted with various cardiac conditions. For instance, in patients with ST segment elevation myocardial infarction, reduced 10-second HRV has been found to be an independent predictor of all-cause mortality [35]. Ultra short HRV indices were also found to poses prognostic potential in COVID-19 [36], as well as in myocarditis patients [37]. Yet, there is little information available about the HRV response in IE patients. Low 24-hour HRV has been proposed as a method for IE complication marker [38]. Reports on ultra-short electrocardiographic markers for ANS dysfunction in IE are lacking. Our study thus aimed to evaluate the prognostic implications of ultra-short HRV indices derived from admission ECGs in patients with IE.

Materials and methods

Study design and population

This retrospective analysis is based on Rambam health care campus (RHCC; Haifa, Israel) patient medical records. This study was performed in line with the principles of the Declaration of Helsinki. Approval was granted by the Ethics Committee of Rambam Health Care Campus on 7 February 2017, number 0603-16-RMB. Since all the data was retrospectively collected, the IRB waived the need for individual informed consent. The following ICD-9-CM diagnosis codes were screened for: 421.0 (Acute and subacute bacterial endocarditis), 421.1 (Acute and subacute infective endocarditis in diseases classified elsewhere), 421.9 (Acute endocarditis, unspecified), 424.90 (Endocarditis, valve unspecified, unspecified cause), 424.91 (Endocarditis in diseases classified elsewhere), 424.99 (Other endocarditis, valve unspecified), 036.42 (Meningococcal endocarditis), 098.84 (Gonococcal endocarditis), 112.81 (Candidal endocarditis), 115.04 (Infection by Histoplasma capsulatum, endocarditis), 115.14) Infection by Histoplasma duboisii, endocarditis), and 115.94 (Histoplasmosis, unspecified, endocarditis).

All patients 18 years and older, who were diagnosed with IE based on the modified Duke criteria [39], between January 2010 and June 2015, were considered. Exclusion criteria included: missing ECG or HRV record from their ED visit, ECGs with irregular heartbeats (any non-sinus rhythm including atrial fibrillation or flutter, premature beats) or low resolution.

Data collection

All ED visits and hospital discharge letters from the study period, were screened for a diagnosis of IE, utilizing the MDClone (Beer-Sheva, Israel) computer software. This study included only patients who had undergone a 10-second resting ECG in the ED. In addition, all medical records were reviewed to identify and exclude potentially ineligible patients. Patients’ medical history (including IE risk factors), presenting symptoms (including clinical components of the modified Duke criteria), ED vital signs as well as laboratory results (i.e. complete blood count, chemistry panel and blood cultures), were collected. All echocardiography reports were studied, and disease specific findings, including valve dysfunction or rupture, peri-valvular abscess, as well as cardiac dysfunction were documented. Furthermore, date of death and IE disease specific complications were noted.

ECG analysis and computation of ultra-short HRV indices

Participating patients underwent a 10-second resting ECG (LAN Green model, Norav Medical, Yokneam, Israel), while lying motionless in a supine position for at least 30 seconds. The ECG electrodes were placed in anatomical positions according to standard procedure. An ECG viewing program was used to visualize the resting ECG files (Resting ECG version 5.62, Norav Medical), PR interval duration and QRS interval duration were automatically measured, and QT interval duration was calculated based on the Bazzett equation. Later, ECG files were analyzed with a custom version of the HRV analysis software able to import and analyze 10-second long recordings (HRV version 5.62, Norav Medical). These programs allowed for the automatic computation of HRV parameters. In addition, ECGs were manually checked and recordings with disturbances (e.g.; excessive noise, sudden baseline instability or low resolution), were excluded. ECGs which contained excessive premature ventricular or supraventricular activity (e.g.; atrial fibrillation, atrial flutter, atrial tachycardia) as well as advanced atrioventricular conduction abnormalities (e.g.; second degree atrioventricular blocks, complete heart block or other high degree AV conduction abnormalities) were also excluded. HRV linear time-domain variables (e.g.; standard deviation of RR intervals [SDNN] and root mean square successive differences [RMSSD]) were the focus of our study in light of their highest association with long-term recordings [33].

Endpoints and evaluated clinical features

The collected clinical features included age, gender, predispositions such as the presence of a prostatic valve, systolic and diastolic blood pressure (BP), resting heart rate, O2 saturation, core body temperature, white blood cells (WBC) count, circulatory levels of Neutrophils, Lymphocytes, Platelets, Hemoglobin and serum creatinine levels. The resting short-term resting heart rate has been shown to provide a crude estimate of the HRV parameters over a 5-minute period, although the effectiveness varies depending on the parameter measured [31].

Study outcomes included IE complications—cardiac complications (e.g.; new heart failure with reduced ejection fraction [HFrEF; based on echocardiographic findings], arrhythmia, abscess); metastatic infection (e.g.; emboli, abscess and mycotic aneurysm); and the need for valvular surgical intervention. As staphylococcus infection is considered a more severe condition [40], the eventual isolation of this pathogen was also examined as a study endpoint. In addition, survival analysis during the follow-up period was carried out.

Statistical analysis

The study database was analyzed with R software (version 4.0.3, The R Foundation for Statistical Computing, Vienna, Austria).

Descriptive statistics—continuous variables are presented as means with Standard Deviation (SD) or medians with interquartile range (IQR) and categorical variables are presented as percentages. Comparisons between groups were performed with Wilcoxon Rank Sum for continuous variables and Fisher’s exact test for categorical variables.

Ultra-short HRV indices for each patient in the study cohort were compared with published normal value ranges, corrected for age and gender [41].

Correlations between variables and Boolean outcomes were examined with logistic regressions (LR) and presented as odds ratio (OR) with 95% confidence intervals (95%CI) and p-values. LR was carried out after verification of compliance with relevant assumptions (e.g.; sufficiently large sample size, linearity of independent variables and log-odds, lack of strongly influential outliers and absence of multicollinearity). Variables found to have statistical significance (p-value<0.050) or trend (p-value<0.090) in univariate analyses, were introduced into a multivariate LR model, in a backward stepwise fashion. Multivariate model accuracy is presented with receiver operating characteristic (ROC) curves, including the area under the curve (AUC), with 95%CIs based on the bootstrapping method, as well as by the Hosmer and Lemeshow goodness-of-fit (HLGOF) and overall model p-value. HRV parameters were assessed as both continuous and Boolean variables (e.g. smaller than values corresponding to quartile 1 [Q1; 25th percentile]). Survival analysis was carried out with univariate and multivariate Cox regression and presented as hazard ratios (HR) with 95%Cis and p-values.

Results

Study population

In the initial evaluation, 229 patients were included based on ICD codes that corresponded to IE diagnoses. Upon additional review of patients’ medical records, 45 patients were excluded as they did not fulfill the modified Duke criteria for IE diagnosis. An additional 75 patients did not have an ED ECG on the day of their admission to the hospital. 34 other patients had either a technically flawed ECG or an irregular rhythm. Thus, the final study sample included 75 patients, aged 60.3 (±18.6) years old. The majority (51, 68.0%) were men and 38 (50.7%) of the patients suffered from native valve IE.

Patient clinical, laboratory and electrocardiographic characteristics in patients with infected native and prosthetic valve, are presented in Table 1. Patients with prosthetic valve IE were significantly older, had lower heart rate, lower hemoglobin levels, longer PR interval and QTc. Notably, time domain HRV indices did not differ between native and prosthetic valve IE patients.

thumbnail
Download:
Table 1. Patient clinical, laboratory and electrocardiographic characteristics (native valve IE vs. prosthetic valve IE).

https://doi.org/10.1371/journal.pone.0287607.t001

HRV indices compared to published norms

HRV parameters were relatively low in the study cohort. Compared with published normal 10-second HRV values, corrected for age and gender [41], SDNN (median 7.72[ms], IQR 4.92[ms] -21.42[ms]) was found to be lower than median values in 57 (76.0%) patients, with 31 (41.3%) patients having SDNN values lower than the 2nd percentile of their age and gender corrected range. As for RMSSD (median 11.8[ms], IQR 7.03[ms]– 28.18[ms]), 56 (74.6%) patients and 25 (33.3%) patients had values lower than age and gender corrected median and 2nd percentile, respectively.

HRV indices and study outcomes

SDNN and RMSSD were found to be statistically significant lower in patients who developed HFrEF or metastatic infections. SDNN was also reduced (statistical trend) in patients ultimately diagnosed with Staphylococcus IE (Table 2). While analysis of HRV indices with regard to the development of a new arrhythmia yielded statistically insignificant results, stratification based on the type of arrhythmia (tachyarrhythmia vs. bradyarrhythmia) was also carried out. RMSSD in the tachyarrhythmia sub-group (median 65.0[ms], IQR 16.9[ms]-120.3[ms]) was high compared to RMSSD in the bradyarrhythmia sub-group (median 12.0[ms], IQR 5.6[ms]-16.7[ms]). Due to this heterogeneity, additional correlations were not conducted for this outcome.

thumbnail
Download:
Table 2. HRV indices with relation to study outcomes.

https://doi.org/10.1371/journal.pone.0287607.t002

Heart failure with reduced ejection fraction

Six-teen patients (21.3%) were diagnosed with HFrEF, Univariate LR correlations are presented in Table 3. Patients who had developed HFrEF, were found to have higher heart rates on ED arrival as well as HRV in the lower quartile (SDNN<4.92ms, OR 4.36, p-value 0.014; RMSSD<7.03ms, OR 3.04, p-value 0.062). However, for predicting HFrEF, the multivariate LR model did not include any HRV parameter and was comprised of only initial heart rate on ED arrival (OR 1.05 [95%CI 1.01–1.08], p-value 0.002; ROC AUC 0.736).

thumbnail
Download:
Table 3. Development of new HFrEF–univariate logistic regression.

https://doi.org/10.1371/journal.pone.0287607.t003

Metastatic infection

Twenty-six (34.6%) patients developed a metastatic infection during their index admission (23 metastatic emboli, 2 metastatic abscess and 1 –mycotic aneurysm). Patients who had been diagnosed with a metastatic infection, were found to be younger, with native valve IE, higher heart rate and body temperature on ED arrival and lower initial platelet levels. In addition, these patients had HRV in the lower quartile (SDNN<4.92ms, OR 3.75, p-value 0.016; RMSSD<7.03ms, OR 5.14, p-value 0.003), as described in Table 4. Multivariate LR model included 3 parameters: native valve IE (aOR 8.21 [95%CI 2.41–35.77], p-value 0.001), ED body temperature upon arrival (aOR 2.56 [95%CI 1.17–6.46], p-value 0.026) and RMSSD<7.03ms (aOR 9.34 [95% CI 2.46–44.67], p-value 0.002). Multivariate LR model ROC curve is presented in Fig 1. AUC was 0.83 (95%CI 0.74–0.92), with HLGOF p-value of 0.487 and overall model p-value <0.0001. Notably, RMSSD in the lower quartile had the strongest aOR in this multivariate classification model.

thumbnail
Download:
Fig 1. Infective endocarditis metastatic infection–multivariate logistic regression prediction model receiver operating characteristic curve, with 95% confidence interval.

https://doi.org/10.1371/journal.pone.0287607.g001

thumbnail
Download:
Table 4. Metastatic infection–univariate logistic regression.

https://doi.org/10.1371/journal.pone.0287607.t004

Valve surgery

Twenty-one (28.0%) IE patients required valvular surgery. HRV indices were not found to be correlated with the need for valve surgery, with only patient age found to have a statistically significant correlation (OR 0.96 (95%CI 0.93–0.99); p-value 0.017).

Staphylococcus IE

Twenty-seven (36.0%) patients were diagnosed with Staphylococcus IE. SDNN<4.92ms was found to have statistically significant univariate correlation with Staphylococcus IE (OR 4.68; p-value 0.006), as presented in Table 5. Multivariate LR model included 2 parameters: ED creatinine (aOR 2.47 [95%CI 1.08–6.29], p-value 0.039) and SDNN<4.92ms (aOR 5.23 [95%CI 1.70–17.49], p-value 0.004). Multivariate LR model ROC curve is presented in Fig 2. AUC was 0.741 (95%CI 0.622–0.860), with HLGOF p-value of 0.585 and overall model p-value 0.001. Noticeably, SDNN in the lower quartile had the strongest aOR in the multivariate classification model.

thumbnail
Download:
Fig 2. Infective endocarditis Staphylococcus infection–multivariate logistic regression prediction model receiver operating characteristic curve, with 95% confidence interval.

https://doi.org/10.1371/journal.pone.0287607.g002

thumbnail
Download:
Table 5. Staphylococcus IE–univariate logistic regression.

https://doi.org/10.1371/journal.pone.0287607.t005

Survival analysis

Forty-four patients (58.6%) died within the study follow-up period (median follow-up 5.06 years, IQR 0.20–8.84 years). SDNN (HR 1.01 [95%CI 1.00–1.02], p-value 0.010) and RMSSD (HR 1.00 [1.00–1.01], p-value 0.012) were both found to have univariate statistically significant Cox regression relations (Table 6).

thumbnail
Download:
Table 6. Survival analysis–univariate cox regression.

https://doi.org/10.1371/journal.pone.0287607.t006

Multivariate Cox regression survival model, included 3 parameters: ED hemoglobin (HR 0.79 [95%CI 0.66–0.95], p-value 0.014), ED creatinine (HR 2.13 [95%CI 1.38–3.28], p-value 0.0005), RMSSD (HR 1.00 [95%CI 1.00–1.01], p-value 0.012). An increase in HRV indices was found to be associated with reduced survival over time, in both univariate and multivariate analyses.

Discussion

In this retrospective study, we have demonstrated, for the first time, that the use of ultra-short time domain HRV indices, collected upon arrival to the ED, can predict morbidity and mortality in IE patients. HRV indices were found to be lower in our cohort, when compared with published age and gender adjusted norms. 75%-76% of patients had HRV values lower than median norms and 33%-41% were found to have HRV levels lower than the normal 2nd percentile. The SDNN and the RMSSD are the primary time-domain measure used to estimate the vagally mediated changes reflected in HRV [42]. In our study, we have demonstrated independent associations between RMSSD, SDNN and IE complications as well as with survival. Of note, HRV indices did not differ significantly between patients with native versus prosthetic valve IE.

IE patients who had developed systemic emboli, were found to have lower RMSSD values upon their ED presentation. Low RMSSD has been established as a negative prognostic index in several patient populations. In a group of adult bone marrow transplant patients, the investigators observed a significant reduction in RMSSD prior to the clinical diagnosis and treatment of sepsis [43]. Similar findings regarding the prognostic role of RMSSD, have been demonstrated in oncologic [44, 45] and post myocardial infarction patients [46].

On the other hand, our study findings have demonstrated that an increase in RMSSD, as a continuous variable, is correlated with mortality over time. However, the magnitude of this association was weak. Several studies have shown correlations between high RMSSD and patient negative outcome, in other infectious diseases. An increased RMSDD, was documented in patients who developed septic shock within a few hours of presentation at the ED [47]. RMSDD was also found to be significantly increased among COVID-19 patients compared to healthy controls matched for age, gender and comorbidities [48]. Finally, in patients suffering from bacterial pneumonia, increased RMSSD obtained form 10-second ECGs has been associated with decreased survival [49].

In an analysis of HRV in patients undergoing peritoneal dialysis, higher RMSSD was associated with increased mortality [50]. Furthermore, in a large cohort of patients with chronic kidney disease, both very low and very high RMSDD were associated with increased risk for all-cause mortality [51]. These findings are in line with our observation that high RMSSD may predict a higher risk of death in patients with IE. The common factor in all these studies, may be that patients with high vagal activity at arrival to the hospital, could already have an infectious or inflammatory condition, to which the cholinergic anti-inflammatory vagus may respond. It is possible that, in the context of an infectious disease, sub-clinically elevated inflammation could cause an increase in vagal activity, indexed by high HRV, to activate the cholinergic anti-inflammatory reflex, and such a compensatory increase in HRV in such contexts, may then predict poor prognosis in infectious diseases such as IE. However, given our relatively small sample size and the small effect size of this HRV-survival relation, this requires replication in larger studies, together with testing this proposed compensation mechanism.

Low SDNN values were correlated with Staphylococcal IE, in both univariate and multivariate LR classification. Staphylococcal infection has been identified as a risk factor for embolic events in IE [52]. In our study, we have demonstrated the association between low HRV indices and both embolic events and Staphylococcal IE, indicating these known reciprocal relations may be identified by HRV parameters.

Similar to our findings, low SDNN has also been correlated with infectious disease complications. In children diagnosed with viral myocarditis, lower SDNN predicted the development of ventricular arrhythmia [53].

Echocardiography has a significant role in the risk stratification of endocarditis patients, by providing information correlated with negative outcomes, such as a large vegetation, para-valvular infection, signs of increased left-cavities filling pressures, pulmonary hypertension and low left ventricular ejection fraction [54]. At admission, early assessment of prognosis is critical and has the ability to identify high risk patients, who may need closer monitoring, require more aggressive treatment (e.g. early surgery) and who may benefit from more frequent follow-up assessments. Serial echocardiographic examinations allow close follow up of endocarditis patients during antibiotic therapy and after disease resolution. Echocardiogram frequency and type depend on the clinical presentation, the involved pathogen and initial echocardiographic findings on admission [55].

In addition to echocardiography, PET-CT is another tool which has recently been established for the diagnosis and follow up of IE, and can assist the early detection of its complications such as systemic embolism [56]. Following our research, we suggest considering more frequent echocardiography studies, as well as early PET-CT for endocarditis patients, with HRV indices associated with high risk of uncontrolled infection, systemic embolism, heart failure and increased mortality, regardless of their clinical condition.

The results of our study indicate a simple and readily available prognostic factor, namely HRV, which can be simply derived from brief 10-seconds ECGs or even from fingertip photo-plethysmography sensors. Nonetheless, this study has limitations, which are related to its retrospective cohort design and also to its possible selection bias. Several patients were excluded from our analysis due to incorrect IE diagnosis and missing or inadequate ECGs. By only including patients with IE diagnosis based on the fulfilment of the Duke criteria with a high-quality ECG tracing of regular cardiac rhythm, we created a smaller yet more accurate IE patient cohort with valid ultra-short HRV indices. By focusing on patients enrolled in a tertiary referral medical center for 5.5 years and applying various statistical tests, which included adjustment for confounders, we have attempted to limit the consequences of a potential selection bias. In spite of this, we cannot predict if a larger study and a longer follow-up period would provide different results, and as such, future research should be conducted to address this question.

In addition, we included ultra-short ECGs for measuring HRV. These could not enable us to derive reliably the frequency-domain parameters of HRV, such as high frequency and low frequency HRV, which could potentially have additional prognostic value in IE. Furthermore, the exact length of time the patients laid down in a supine position prior to undergoing ECG is unknown. Nonetheless, even short stabilization periods prior to HRV assessment may be acceptable, especially in younger, healthier patients [57] and in static conditions [58] (similar to the setting in our study). Finally, due to the relatively small sample, a stratification to native and prosthetic valve IE did not yield significant results. Future prospective studies should address these issues, in order to validate our findings. Following our research, we suggest to consider more frequent radiographic studies (e.g.; PET-CT, echocardiogram) for endocarditis patients, with HRV indices associated with high risk of uncontrolled infection, systemic embolism and increased mortality, regardless of their clinical condition.

Conclusions

Time domain ultra-short HRV indices based on ECG, carried out upon ED arrival, have the potential to allow for early risk stratification, in IE patients. Our findings should be validated, prior to implementation in patient management strategies.

References

  1. 1. Habib G, Lancellotti P, Antunes MJ, Bongiorni MG, Casalta J-P, Del Zotti F, et al. 2015 ESC Guidelines for the management of infective endocarditis: The Task Force for the Management of Infective Endocarditis of the European Society of Cardiology (ESC). Endorsed by: European Association for Cardio-Thoracic Surgery (EACTS), the European. European heart journal. 2015;36(44):3075–3128. pmid:26320109
  2. 2. Wei X-B, Su Z-D, Liu Y-H, Wang Y, Huang J-L, Yu D-Q, et al. Age, creatinine and ejection fraction (ACEF) score: a simple risk-stratified method for infective endocarditis. QJM: monthly journal of the Association of Physicians. 2019;112(12):900–906. pmid:31359051
  3. 3. Tascini C, Aimo A, Arzilli C, Sbrana F, Ripoli A, Ghiadoni L, et al. Procalcitonin, white blood cell count and C-reactive protein as predictors of S. aureus infection and mortality in infective endocarditis. International journal of cardiology. 2020;301:190–194. pmid:31405585
  4. 4. Mohanan S, Gopalan Nair R, Vellani H, C G S, George B, M N K. Baseline C-reactive protein levels and prognosis in patients with infective endocarditis: A prospective cohort study. Indian heart journal. 2018;70 Suppl 3(Suppl 3):S43–S49. pmid:30595302
  5. 5. Chen Y, Ye L-J, Wu Y, Shen B-Z, Zhang F, Qu Q, et al. Neutrophil-Lymphocyte Ratio in Predicting Infective Endocarditis: A Case-Control Retrospective Study. Mediators of inflammation. 2020;2020:8586418. pmid:33354163
  6. 6. Ferrera C, Vilacosta I, Fernández C, López J, Sarriá C, Olmos C, et al. Usefulness of thrombocytopenia at admission as a prognostic marker in native valve left-sided infective endocarditis. The American journal of cardiology. 2015;115(7):950–955. pmid:25708863
  7. 7. Tok D, Canpolat U, Tok D, Turak O, İşleyen A, Öksüz F, et al. Association of mean platelet volume level with in-hospital major adverse events in infective endocarditis. Wiener klinische Wochenschrift. 2015;127(5–6):197–202. pmid:25777146
  8. 8. Kahveci G, Bayrak F, Mutlu B, Bitigen A, Karaahmet T, Sonmez K, et al. Prognostic value of N-terminal pro-B-type natriuretic peptide in patients with active infective endocarditis. The American journal of cardiology. 2007;99(10):1429–1433. pmid:17493474
  9. 9. Araújo IR, Ferrari TCA, Teixeira-Carvalho A, Campi-Azevedo AC, Rodrigues LV, Guimarães Júnior MH, et al. Cytokine Signature in Infective Endocarditis. PloS one. 2015;10(7):e0133631. pmid:26225421
  10. 10. Ris T, Teixeira-Carvalho A, Coelho RMP, Brandao-de-Resende C, Gomes MS, Amaral LR, et al. Inflammatory biomarkers in infective endocarditis: machine learning to predict mortality. Clinical and experimental immunology. 2019;196(3):374–382. pmid:30697694
  11. 11. Guimarães Júnior MH, Ferrari TCA, Teixeira-Carvalho A, Moreira M de L, de Souza Santos LJ, Costa-Silva MF, et al. Cell-derived microvesicles in infective endocarditis: Role in diagnosis and potential for risk stratification at hospital admission. The Journal of infection. 2019;79(2):101–107. pmid:31207323
  12. 12. Thuny F, Di Salvo G, Belliard O, Avierinos J-F, Pergola V, Rosenberg V, et al. Risk of embolism and death in infective endocarditis: prognostic value of echocardiography: a prospective multicenter study. Circulation. 2005;112(1):69–75. pmid:15983252
  13. 13. Liao L, Kong DF, Samad Z, Pappas PA, Jollis JG, Lin SS, et al. Echocardiographic risk stratification for early surgery with endocarditis: a cost-effectiveness analysis. Heart (British Cardiac Society). 2008;94(5):e18. pmid:17575328
  14. 14. Wang TKM, Bin Saeedan M, Chan N, Obuchowski NA, Shrestha N, Xu B, et al. Complementary Diagnostic and Prognostic Contributions of Cardiac Computed Tomography for Infective Endocarditis Surgery. Circulation. Cardiovascular imaging. 2020;13(9):e011126. pmid:32900226
  15. 15. Meine TJ, Nettles RE, Anderson DJ, Cabell CH, Corey GR, Sexton DJ, et al. Cardiac conduction abnormalities in endocarditis defined by the Duke criteria. American heart journal. 2001;142(2):280–285. pmid:11479467
  16. 16. Miyamoto MI, Hutter AMJ, Blum JH, Torchiana DF. Cardiac conduction abnormalities preceding transoesophageal echocardiographic evidence of perivalvar extension of infection in a case of Salmonella prosthetic valve endocarditis. Heart (British Cardiac Society). 1997;78(4):416–418. pmid:9404262
  17. 17. Berk WA. Electrocardiographic findings in infective endocarditis. The Journal of emergency medicine. 1988;6(2):129–132. pmid:3385174
  18. 18. Kohli U, Obuobi S, Addetia K, Ota T, Nayak HM. PR interval prolongation is significantly associated with aortic root abscess: An age- and gender-matched study. Annals of noninvasive electrocardiology: the official journal of the International Society for Holter and Noninvasive Electrocardiology, Inc. 2021;26(4):e12849. pmid:33939235
  19. 19. Pecchia L, Castaldo R, Montesinos L, Melillo P. Are ultra-short heart rate variability features good surrogates of short-term ones? State-of-the-art review and recommendations. Healthcare technology letters. 2018;5(3):94–100. pmid:29922478
  20. 20. Dekker JM, Crow RS, Folsom AR, Hannan PJ, Liao D, Swenne CA, et al. Low heart rate variability in a 2-minute rhythm strip predicts risk of coronary heart disease and mortality from several causes: the ARIC Study. Atherosclerosis Risk In Communities. Circulation. 2000;102(11):1239–1244. pmid:10982537
  21. 21. Sajadieh A, Nielsen OW, Rasmussen V, Hein HO, Hansen JF. C-reactive protein, heart rate variability and prognosis in community subjects with no apparent heart disease. Journal of internal medicine. 2006;260(4):377–387. pmid:16961675
  22. 22. de Castilho FM, Ribeiro ALP, Nobre V, Barros G, de Sousa MR. Heart rate variability as predictor of mortality in sepsis: A systematic review. PloS one. 2018;13(9):e0203487. pmid:30204803
  23. 23. Kim K-N, Yao Y, Ju S-Y. Heart rate variability and inflammatory bowel disease in humans: A systematic review and meta-analysis. Medicine. 2020;99(48):e23430. pmid:33235125
  24. 24. Tracey KJ. Reflex control of immunity. Nature reviews. Immunology. 2009;9(6):418–428. pmid:19461672
  25. 25. Rosas-Ballina M, Olofsson PS, Ochani M, Valdés-Ferrer SI, Levine YA, Reardon C, et al. Acetylcholine-synthesizing T cells relay neural signals in a vagus nerve circuit. Science (New York, N.Y.). 2011;334(6052):98–101. pmid:21921156
  26. 26. Mihaylova S, Schweighöfer H, Hackstein H, Rosengarten B. Effects of anti-inflammatory vagus nerve stimulation in endotoxemic rats on blood and spleen lymphocyte subsets. Inflammation research: official journal of the European Histamine Research Society… [et al.]. 2014;63(8):683–690. pmid:24802890
  27. 27. Malik M, Camm AJ, Bigger JT, Breithardt G, Cerutti S, Cohen RJ, et al. Heart rate variability. Standards of measurement, physiological interpretation, and clinical use. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. European Heart Journal. 1996:354–81. pmid:8737210
  28. 28. Shaffer F, Ginsberg JP. An Overview of Heart Rate Variability Metrics and Norms. Frontiers in public health. 2017;5:258. pmid:29034226
  29. 29. Baek HJ, Cho CH, Cho J, Woo JM. Reliability of ultra-short-term analysis as a surrogate of standard 5-min analysis of heart rate variability. Telemedicine and e-Health. 2015;21(5):404–414. pmid:25807067
  30. 30. Munoz ML, Van Roon A, Riese H, Thio C, Oostenbroek E, Westrik I, et al. Validity of (Ultra-)Short recordings for heart rate variability measurements. PLoS ONE. 2015;10(9). pmid:26414314
  31. 31. Nussinovitch U, Elishkevitz KP, Kaminer K, Nussinovitch M, Segev S, Volovitz B, et al. The efficiency of 10-second resting heart rate for the evaluation of short-term heart rate variability indices. PACE—Pacing and Clinical Electrophysiology. 2011;34(11):1498–1502. pmid:21797904
  32. 32. Politi K, Kaminer K, Nussinovitch U. Reliability of ultrashort electrocardiographic indices in hypertension: The quest for a clinically applicable prognostic marker. Journal of Investigative Medicine. 2020. pmid:31420365
  33. 33. Nussinovitch U. Reliability of ultra-short indices for autonomic dysfunction in dyslipidemia. Clinical physiology and functional imaging. 2020;40(6):423–433. pmid:32886849
  34. 34. Nussinovitch U, Cohen O, Kaminer K, Ilani J, Nussinovitch N. Evaluating reliability of ultra-short ECG indices of heart rate variability in diabetes mellitus patients. Journal of Diabetes and its Complications. 2012;26(5):450–453. pmid:22682758
  35. 35. Karp E, Shiyovich A, Zahger D, Gilutz H, Grosbard A, Katz A. Ultra-short-term heart rate variability for early risk stratification following acute st-elevation myocardial infarction. Cardiology. 2009;114(4):275–283. pmid:19690410
  36. 36. Mol MBA, Strous MTA, van Osch FHM, Vogelaar FJ, Barten DG, Farchi M, et al. Heart-rate-variability (HRV), predicts outcomes in COVID-19. PloS one. 2021;16(10):e0258841. pmid:34710127
  37. 37. Perek S, Nussinovitch U, Cohen R, Gidron Y, Raz-Pasteur A. Ultra Short Heart Rate Variability Predicts Clinical Outcomes in Patients with a Clinical Presentation Consistent with Myocarditis: A Derivation Cohort Analysis. Journal of clinical medicine. 2022;12(1). pmid:36614887
  38. 38. Feaktsistava N, Sujayeva V, Sudzhaeva S. The significance of cardiac rhythm variability value assessment in partients with infective endocarditis. E Journal of Cardiovascular Medicine. 2013;1(2):32–35.
  39. 39. Li JS, Sexton DJ, Mick N, Nettles R, Fowler VG, Ryan T, et al. Proposed Modifications to the Duke Criteria for the Diagnosis of Infective Endocarditis. Clin Infect Dis. 2000;April(30(4)):633–638. pmid:10770721
  40. 40. Liesenborghs L, Meyers S, Lox M, Criel M, Claes J, Peetermans M, et al. Staphylococcus aureus endocarditis: distinct mechanisms of bacterial adhesion to damaged and inflamed heart valves. European heart journal. 2019;40(39):3248–3259. pmid:30945735
  41. 41. van den Berg ME, Rijnbeek PR, Niemeijer MN, Hofman A, Herpen G van Bots ML, et al. Normal values of corrected heart-rate variability in 10-second electrocardiograms for all ages. Frontiers in Physiology. 2018;9(4):424. pmid:29755366
  42. 42. Shaffer F, McCraty R, Zerr CL. A healthy heart is not a metronome: an integrative review of the heart’s anatomy and heart rate variability. Frontiers in Psychology. 2014;5(September):1–19. pmid:25324790
  43. 43. Ahmad S, Tejuja A, Newman KD, Zarychanski R, Seely AJ. Clinical review: a review and analysis of heart rate variability and the diagnosis and prognosis of infection. Critical care (London, England). 2009;13(6):232. pmid:20017889
  44. 44. Hu S, Lou J, Zhang Y, Chen P. Low heart rate variability relates to the progression of gastric cancer. World Journal of Surgical Oncology. 2018;16(1):1–5. pmid:29514707
  45. 45. Gidron Y, De Couck M, Schallier D, De Greve J, Van Laethem JL, Maréchal R. The relationship between a new biomarker of vagal neuroimmunomodulation and survival in two fatal cancers. Journal of Immunology Research. 2018;2018. pmid:29854838
  46. 46. Hayano J, Ueda N, Kisohara M, Yuda E, Carney RM, Blumenthal JA. Survival Predictors of Heart Rate Variability After Myocardial Infarction With and Without Low Left Ventricular Ejection Fraction. Frontiers in neuroscience. 2021;15:610955. pmid:33633535
  47. 47. Wee BYH, Lee JH, Mok YH, Chong S-L. A narrative review of heart rate and variability in sepsis. Annals of translational medicine. 2020;8(12):768. pmid:32647693
  48. 48. Kaliyaperumal D, Rk K, Alagesan M, Ramalingam S. Characterization of cardiac autonomic function in COVID-19 using heart rate variability: a hospital based preliminary observational study. Journal of basic and clinical physiology and pharmacology. 2021;32(3):247–253. pmid:33705614
  49. 49. Izhaki N, Perek S, Agbaria M, Raz-Pasteur A. Ultrashort Heart Rate Variability for Early Risk Stratification in Pneumonia Patients: Preliminary Analysis. The Israel Medical Association journal: IMAJ. 2022;24(11):741–746.
  50. 50. Pei J, Tang W, Li L-X, Su C-Y, Wang T. Heart rate variability predicts mortality in peritoneal dialysis patients. Renal failure. 2015;37(7):1132–1137. pmid:26123265
  51. 51. Drawz PE, Babineau DC, Brecklin C, He J, Kallem RR, Soliman EZ, et al. Heart rate variability is a predictor of mortality in chronic kidney disease: a report from the CRIC Study. American journal of nephrology. 2013;38(6):517–528. pmid:24356377
  52. 52. Habib G. Management of infective endocarditis. Heart (British Cardiac Society). 2006;92(1):124–130. pmid:16365367
  53. 53. Ling N, Li CL, Wang ZZ, Zhang HN, Xu H, An XJ. Heart rate variability in children with myocarditis presenting with ventricular arrhythmias. European Review for Medical and Pharmacological Sciences. 2018;22(4):1102–1105. pmid:29509262
  54. 54. Thuny F, Grisoli D, Collart F, Habib G, Raoult D. Management of infective endocarditis: challenges and perspectives. Lancet (London, England). 2012;379(9819):965–975. pmid:22317840
  55. 55. Habib G, Badano L, Tribouilloy C, Vilacosta I, Zamorano JL, Galderisi M, et al. Recommendations for the practice of echocardiography in infective endocarditis. European journal of echocardiography: the journal of the Working Group on Echocardiography of the European Society of Cardiology. 2010;11(2):202–219. pmid:20223755
  56. 56. Pelletier-Galarneau M, Abikhzer G, Harel F, Dilsizian V. Detection of Native and Prosthetic Valve Endocarditis: Incremental Attributes of Functional FDG PET/CT over Morphologic Imaging. Current cardiology reports. 2020;22(9):93. pmid:32647931
  57. 57. Krejčí J, Botek M MA. Stabilization period before capturing an ultra- short vagal index can be shortened to 60 s in endurance athletes and to 90 s in university students. PloS one. 2018;Oct 8(13(10)):e0205115. pmid:30296274
  58. 58. Kim JW, Seok HS, Shin H. Is Ultra-Short-Term Heart Rate Variability Valid in Non-static Conditions? Frontiers in physiology. 2021;12(3):596060. pmid:33859568