 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
COMMENTARY |
Department of Pediatrics
Pediatric Clinical Research Office
Doernbecher Children’s Hospital
Oregon Health and Science University
Portland, Oregon 97239-3098
Abbreviations: HRC, heart rate characteristics • WBC, white blood cell • SIRS, systemic inflammatory response syndrome • ROC, receiver operating characteristic
There have been many reports of different heart rate–variability metrics and their association with sepsis in neonates, children, and adults.1–8 Common to all has been a lack of specificity between subjects, although sensitivity to dynamic changes within subjects has been very high.
In this issue of Pediatrics, Griffin et al9 report on a specific group of measures (or metrics) of heart rate variability, termed the heart rate characteristics (HRC) index, and its association with blood-culture–proven sepsis in neonates.
The basis for their research is the poor diagnostic sensitivity and specificity of the current clinical and laboratory evaluation of an infant for suspected sepsis. Current neonatal practice includes careful examination for sepsis-associated physical findings including temperature instability, tachycardia, tachypnea, apnea, prolonged capillary refill time, hypotension, and/or decreased urine output. Corroborative laboratory tests include significant elevation or depression of the white blood cell (WBC) count, the presence of immature neutrophils, metabolic acidosis (lowered pH and/or bicarbonate or elevated lactate), thrombocytopenia, coagulopathy, or evidence of renal or hepatic dysfunction. Of course, the "gold standard" is a positive culture of blood, cerebrospinal fluid, or urine, but there is an unacceptably long turn-around time for results (often 
24–48 hours) and a high false-negative rate. This was confirmed recently by the National Institute of Child Health and Human Development Neonatal Research Network, which found that many of the currently used physical signs and laboratory tests were nonspecific and did not predict the presence of sepsis accurately.10 Griffin et al9 point out that the poor diagnostic accuracy is likely attributable to the complexity and variability of the host response to infection, commonly referred to as the systemic inflammatory response syndrome (SIRS).11,12
The authors’ hypothesis was that the HRC index adds information to conventional laboratory tests in diagnosing neonatal sepsis. To test this hypothesis, they prospectively collected heart rate data in 678 consecutive infants who stayed >7 days in a single NICU over a 4-year period. They measured HRC and noted 149 episodes of sepsis with positive blood cultures, for which data were available for analysis in 137. They used multivariable logistic-regression modeling adjusted for repeated measures to analyze the relationship between the HRC index and laboratory tests (the ratio of immature to total WBC forms [I/T], WBC count, glucose, platelet count, HCO3, PCO2, and pH). It is unfortunate that the authors did not include any physical symptoms associated with SIRS or sepsis. They reported that the HRC index (available 92% of the time) was significantly associated with sepsis (receiver operating characteristic [ROC] curve area: 0.73; P < .001). The I/T ratio, WBC count, and platelet counts (available 4–8% of the time) and blood glucose and pH (available 28% and 38% of the time, respectively) were also associated with sepsis (ROC curve area: 0.75; P < .001). However, their main finding was that the HRC index and laboratory values added independent information to each other and increased the ROC curve area to 0.82 (P < .01) in a predictive model using all significant variables.
There are a number of important points to consider from this study. The clinical and laboratory definitions of the sepsis continuum (infection, sepsis, severe sepsis, septic shock, and organ dysfunction) need to be agreed on by those medical personnel caring for neonates. A recent pediatric consensus conference proposed such definitions for the pediatric population.12 However, the age groups in this article only included full-term newborns and did not discuss premature infants.12 Similar to the main goal of the pediatric sepsis definitions, which was to provide a uniform basis for clinicians and researchers to study and diagnose severe sepsis in children, working definitions for the sepsis continuum for all preterm and term neonates should be addressed in the near future.
The application of a model of heart rate variability specifically applied to the subclinical phase of neonatal sepsis, what the authors term the HRC index, is an interesting concept supported by the research team’s previous work.6,13–15 However, concern exists that the HRC index (consisting of the heart rate standard deviation, sample asymmetry, and sample entropy) will prove insensitive when applied to a larger, more heterogeneous population. For example, it is known that, compared with the normal healthy individual, metrics of heart rate variability are generally diminished with illness and injury.16 Therefore, differentiating between SIRS, sepsis, necrotizing enterocolitis, or an intraventricular hemorrhage in an individual neonate may prove difficult if not impossible. It is clear that additional multi-institutional, larger, prospective, randomized trials can address the question of the sensitivity and specificity of the HRC index.
The authors suggest that the HRC index may be used as a screening test for neonatal sepsis.9 It is unclear from this study whether the test may be used as a screening or diagnostic test, alone or in combination with other symptoms and tests (eg, laboratory tests and physical examination findings), and, most importantly, whether a small increase in the ROC (0.07) afforded by adding the HRC index to standard testing is clinically relevant in terms of improving the ability to screen for or diagnose neonatal sepsis. Whatever future studies are done, they should address this issue as well as incorporate physical examination findings into the multivariable logistic-regression model.
Nonetheless, it is clear that the HRC index, or some other heart rate–variability metric, alone or in combination with other clinical signs and laboratory tests, may potentially improve our ability to detect and diagnose neonatal SIRS and sepsis. Although biomarkers such as an elevated sedimentation rate, elevated C-reactive protein, base deficit, IL-6, procalcitonin, or some other as-yet-undiscovered marker may prove useful,12 the HRC index has the advantage of being available already (ie, through continuous electrocardiogram monitoring), noninvasive, and likely inexpensive. Griffin et al are to be commended for their focused research in this important area and encouraged to proceed with taking the next steps toward improving our ability to diagnosis neonatal sepsis.
 |
FOOTNOTES |
|---|
Address correspondence to Brahm Goldstein, MD, FCCM, Department of Pediatrics, Pediatric Clinical Research Office, Doernbecher Children’s Hospital, 707 SW Gaines St, Mail Code: CDRCP, Portland, OR 97239. E-mail: goldsteb{at}ohsu.edu
No conflict of interest declared.
|
REFERENCES |
|---|
 TOP REFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2005 American Academy of Pediatrics. All rights reserved. |
||
|
||
