National Association of Neonatal Nurses logo.



All About That Beat: Heart Rate Characteristic Monitoring in Neonatal Sepsis


Neonatal Sepsis

Sepsis is a common illness treated by neonatal clinicians that remains a major cause of neonatal morbidity and mortality, particularly among very-low-birth-weight (VLBW) infants. The term sepsis includes either evidence of culture-positive infection or vital sign changes indicative of systemic inflammatory response syndrome (SIRS) (Sullivan & Fairchild, 2015). SIRS causes activation of inflammatory pathways, releasing cytokines as well as other inflammatory mediators that can cause tissue and cell damage. A diagnosis of SIRS requires the presence of two or more of the following: tachypnea or respiratory alkalosis; tachycardia; fever or hypothermia; and high or low white blood cell count or bandemia (Sullivan & Fairchild, 2015). However, there are limitations to identifying early onset sepsis in neonates, as changes in vital signs can be indicative of premature organ systems rather than sepsis or SIRS (Sullivan & Fairchild, 2015). Shock is not well defined in the neonatal population and is a relatively uncommon occurrence, but it can be a fatal consequence of sepsis. In neonates, shock is typically described as poor systemic perfusion, which is associated with metabolic acidosis, hypotension, and tachycardia. Common causes of shock in the Neonatal Intensive Care Unit (NICU) include septicemia and necrotizing enterocolitis (NEC), with approximately 5% of infants having late onset sepsis (Sullivan & Fairchild, 2015).

Improvements in technology and implementation of evidence-based practices, most notably the use of antenatal corticosteroids and postnatal surfactant, have decreased the incidence of mortality and morbidity. Despite this, mortality among VLBW infants remains high, approximately 10%, and is inversely associated with gestational age (Fairchild & Aschner, 2012). Late onset sepsis mortality ranges from less than 10% to greater than 30%, and gram-negative bacteria account for approximately 20% of late-onset bacteremia in VLBW infants and more commonly lead to shock or death compared to gram-positive bacteria (Stoll, Hansen, Fanaroff, et al., 2002; Kermorvant-Duchemin, Laborie, Rabilloud, et al., 2008). However, viruses such as enteroviruses and herpes simplex viruses also can cause septic shock and carry increased mortality (Sullivan and Fairchild, 2015). Delayed or insufficient treatment of sepsis can contribute to the progression of septic shock, which can cause temporary or permanent organ damage, neurological and developmental deficits, or even death. A prompt, accurate diagnosis of sepsis is critical for survival (Sullivan & Fairchild, 2015).

Evaluation of Sepsis in Neonates

Identifying sepsis in neonates can be challenging, and healthcare providers are constantly monitoring infants in order to quickly and accurately diagnose and treat infected infants. Physiological monitoring and the use of biomarkers, in addition to monitoring for changes in clinical status, are common methods of evaluating infants for sepsis and may possibly prevent the progression of sepsis to septic shock or death (Sullivan a& Fairchild, 2015).

The most basic form of physiological monitoring is conventional vital sign monitoring, and trends in vitals are typically more useful than isolated values in neonates as normal vital sign ranges are not specific to each individualized patient (Sullivan & Fairchild, 2015). While a valuable tool, one limit to conventional vital sign monitoring is that it does not have the ability to detect changes in heart rate variability compared to baseline values (Sullivan & Fairchild, 2015). Another limit to conventional vital sign monitoring is that it lacks the ability to detect changes in vital sign patterns associated with impending clinical deterioration (Fairchild & Aschner, 2012).

The use of biomarkers for identifying sepsis has been widely studied, with C-reactive protein (CRP) being the most widely used acute phase reactant monitored by NICU clinicians, but this has not yet translated into clinically useful data for ruling sepsis in or out (Fairchild & Aschner, 2012). Serial CRPs can provide more information than isolated values, and trending of CRP values is recommended. The CRP is a non-specific measure of inflammation or tissue injury and can be elevated in nonseptic infants, which is one major limitation of CRP values. Another limitation of biomarker screening is that the values are typically obtained once a patient is already displaying signs of sepsis (Fairchild & Aschner, 2012). However, the CRP does have a high negative predictive value and can be reassuring for clinicians when the values are low.

The early onset sepsis risk calculator (EOS SRC) is a widely used tool developed to determine risk of sepsis using both maternal and infant risk factors. Use of the EOS SRC has decreased antibiotic usage in NICUs; however, it is recommended for infants older than 34 weeks and does not address premature infants younger than 34 weeks who are at highest risk (Kuzniewicz, Walsh, Li, et al., 2016). Despite multiple methods of evaluating for sepsis in neonates, the methods discussed here have limitations and lack specificity. In order to accurately and effectively diagnose and treat infected infants within a timely manner, other methods of evaluating for sepsis need to be investigated further. Monitoring of heart rate characteristics has been suggested to decrease mortality in VLBW infants and identify infants at risk for developing sepsis within the following 24 hours, prior to clinical deterioration (Sullivan & Fairchild, 2015).

Heart Rate Variability and Sepsis

Heart rate is mostly controlled by the autonomic nervous system and can fluctuate due to immunologic as well as cardiovascular changes. Decreased heart rate variability (HRV) occurs when there is a paucity of normal accelerations and decelerations of heart rate (Fairchild & Aschner, 2012). Literature suggests that low HRV in preterm infants can occur in sepsis prior to the presence or recognition of any obvious changes in clinical signs. Transient decelerations in heart rate, in addition to low HRV, have been identified in the hours leading up to a sepsis diagnosis in infants (Sullivan & Fairchild, 2015).

In response to sympathetic or parasympathetic nerve firing, sinoatrial node pacemaker cells respond to the release of neurotransmitters in a clinically well state. Sympathetic nerve firing causes release of norepinephrine, producing a small increase in heart rate; parasympathetic (vagus nerve) activation causes acetylcholine release, leading to a transient deceleration or decreased heart rate (Sullivan & Fairchild, 2015). Physiologic needs affect accelerations and decelerations, reflecting the balance of the autonomic nervous system (Fairchild and Aschner, 2012).

HRV can occur in a variety of pathophysiologic conditions, including sepsis (Fairchild, & O'Shea, 2010). The systemic inflammatory response releases inflammatory cytokines, which have been found to be associated with a decreased HRV (Fairchild and Aschner, 2012). Pathogen toxins such as lipopolysaccharide from Escherichia coli, a common bacterium known to cause sepsis in neonates, can also decrease HRV through pro-inflammatory cytokines and has been shown in both prenatal and postnatal animal models (Fairchild & O'Shea 2010). Additionally, studies in mice models showed that administration of bacteria or candida leads to vagus nerve stimulation (parasympathetic nerve activation) very rapidly, ultimately resulting in transient, repetitive heart rate decelerations (Sullivan & Fairchild, 2015). Vagus nerve stimulation dampens inflammatory cytokine production via the cholinergic anti-inflammatory pathway by leukocytes, via nicotinic cholinergic receptors, and is critical to innate host defenses (Sullivan & Fairchild, 2015; Fairchild and Aschner, 2012). Endogenous activation of this host defense may be reflective of heart rate decelerations occurring in sepsis (Fairchild & Aschner, 2012).

A common abnormal heart rate characteristic familiar to neonatal and obstetric clinicians is acute uteroplacental insufficiency resulting in fetal asphyxia. Acidosis and hypoxia cause decreased beat-to-beat variability as well as repetitive transient decelerations and have also been described in cases of chorioamnionitis (Fairchild & Aschner, 2012).

Predictive Monitoring for Neonatal Sepsis

Predictive monitoring is becoming increasingly popular in NICUs and identifies high-risk infants at risk for life-threatening morbidities. Predictive monitoring detects early stage illness, improving time to treatment with the ultimate goal of reducing morbidity and mortality (Sullivan & Fairchild, 2015). The heart rate characteristics (HRC) index monitor was the first vital sign predictive monitoring device that was developed for identification of sepsis in neonates (Sullivan & Fairchild, 2015).

The HRC index, also known as the Heart Rate Observation (HeRO) score, was developed at the University of Virginia after observation of the association between decreased HRV, decelerations, and sepsis among premature infants. It was designed to detect the fold-increase in risk of infants developing culture-proven or clinical sepsis in the following 24 hours, by using a mathematical algorithm. The HRC includes three measures: heart rate variability (standard deviation of R-R intervals), presence of abnormal heart rate decelerations (sample asymmetry), and complexity (sample entropy) (Moorman, Carlo, Kattwinkel, et al., 2011).

Since 2003, the HeRO monitor (Medical Predictive Science Corporation) has had 510(k) clearance from the United States Food and Drug Administration for monitoring HRV and decelerations. The HeRO monitor does not require additional sensors or other hardware and extracts electrocardiographic data from unit monitors (Fairchild & Aschner, 2012). The HeRO monitor analyzes real time electrocardiogram data from standard NICU monitors and displays a continuous HRC index that is updated hourly. The HRC index reflects heart rate variability and transient decelerations in the previous 12 hours and displays the current HRC index in addition to the trend over the last five days as well as the last 30 minutes of heart rate (Moorman, Carlo, Kattwinkel, et al., 2011). A HeRO score of "1" is considered "normal" for healthy VLBW infants, and a score of "5" indicates a five-fold increased risk that the infant will be diagnosed with sepsis within the following 24 hours (Fairchild & Aschner, 2012). The University of Virginia studied the HeRO score in more than 300 VLBW infants with more than 100 episodes of sepsis. This study was externally validated at Wake Forest University with a similar sample size and was again suggested to be highly correlated with a diagnosis of sepsis (Moorman, Carlo, Kattwinkel, et al., 2011). In a subsequent analysis of over 1000 neonatal patients at the University of Virginia and Wake Forest, elevated HeRO scores were again shown to be associated with sepsis and mortality (Fairchild & Aschner, 2012). Subsequent studies have suggested that high HeRO scores correlate with mortality and other infections and provides value to NICU clinicians in the diagnosis of sepsis.

Randomized Clinical Trial of HeRO Monitoring

A randomized clinical trial was carried out in nine NICUs across the United States in order to determine the clinical utility of HeRO monitoring in neonatal patients. VLBW infants, stratified by birthweight of <1000 g or 1001–1500 g, were randomized to either having the HeRO score displayed and viewable by clinicians or to have the HeRO score measured but not displayed (Moorman, Carlo, Kattwinkel, et al., 2011). Infants were followed for 120 days after randomization occurred or until discharge or neonatal death. This study is the largest randomized clinical trial of VLBW infants published to date, with a total of 3003 patients being randomized and 2989 patients completing the study (Moorman, Carlo, Kattwinkel, et al., 2011). Mortality among the HeRO display group was 22% lower compared to the non-display group (10.2% vs. 8.1%, P = .04), and in a prespecified subgroup analysis of ELBW infants, the relative reduction rate in death was 26% (P = .02) in the HeRO display group (Moorman, Carlo, Kattwinkel, et al., 2011). The incidence of blood culture-proven sepsis was not different in the two study groups; however, the 30-day mortality after the first episode of sepsis was lower in the HeRO display group (10.0% versus 16.1%, P = .01) (Moorman, Carlo, Kattwinkel, et al., 2011).


HeRO scores alone have been suggested to have comparable or better predictive accuracy for sepsis compared with laboratory values such as white blood cell count, blood glucose, and immature to total neutrophil ratio (Fairchild & Aschner, 2012). An advantage to HeRO monitoring compared with laboratory testing is that HeRO scores continuously display the risk of sepsis, while laboratory values are only obtained when ordered by a clinician either routinely or when a change in status warrants additional evaluation. Future improvement in HeRO monitoring includes implementation of guidelines and protocols to determine what intervention is necessary when an elevated HeRO score is obtained.


  1. Fairchild, K., & Aschner, J. (2012). HeRO monitoring to reduce mortality in NICU patients. Research and Reports in Neonatology, 2.
  2. Fairchild, K., & O'Shea, T. (2010). Heart rate characteristics: Physiomarkers for detection of late-onset neonatal sepsis. Clinical Perinatology, 37.
  3. Kermorvant-Duchemin, E., Laborie, S., Rabilloud, M., Lapillonee, A., & Claris, O. (2008). Outcome and prognostic factors in neonates with septic shock. Pediatric Critical Care Medicine, 9.
  4. Kuzniewicz, M., Walsh, E., Li, S., Fischer, A., Escobar, G. (2016). Development and implementation of an early-onset sepsis calculator to guide antibiotic management in late preterm and term neonates. The Joint Commission Journal on Quality and Patient Safety, 42(5).
  5. Moorman, J., Carlo, W., Kattwinkel, J., Schelonka, R., Porcelli, P., Navarrete, C., Bancalari, E., Aschner, J., Whit Walker M., Perez, JA., Palmer, C., Stukenborg, G., Lake, D., & Michael O'Shea T. (2011). Mortality reduction by heart rate characteristic monitoring in very low birth weight neonates: a randomized trial. Journal of Pediatrics, 159(6).
  6. Stoll, B., Hansen, N., Fanaroff, A., Wright, L., Carlo, W., Ehrenkranz, R., et al. (2002). Late-onset sepsis in very low birth weight neonates: the experience of the NICHD Neonatal Research Network. Pediatrics, 110.
  7. Sullivan, B., & Fairchild, K. (2015). Predictive monitoring for sepsis and necrotizing enterocolitis to prevent shock. Seminars in Fetal and Neonatal Medicine, 20(4).

Please note: The information presented and opinions expressed herein are those of the authors and do not necessarily represent the views of the National Association of Neonatal Nurses. Any clinical or technical recommendations made by the authors must be weighed against the health care provider's own judgment and the accepted guidelines published on the subject.


Interested in learning more about neonatal sepsis? 

Check out our latest CNENow! module on sepsis

Our Sponsor

pediatrix mednax