 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
PEDIATRICS Vol. 107 No. 1 January 2001, pp. 97-104
From the Department of * Pediatrics (Neonatology Division),
Internal Medicine (Cardiovascular Division), and Molecular
Physiology and Biological Physics, and the Cardiovascular Research
Center, University of Virginia Health Sciences Center, Charlottesville,
Virginia.
 |
ABSTRACT |
|---|
Top
Abstract
Methods
Results
Discussion
References
|
|---|
Background and Objective. Abrupt clinical deterioration because of sepsis is a major cause of morbidity and mortality in neonates, and earlier diagnosis should improve therapy of this potentially catastrophic illness. In practice, clinical signs and laboratory data have not been perceived as sensitive or specific for early stages of sepsis. Because heart rate characteristics (HRC) are abnormal during fetal distress and neonatal illness, we hypothesized that abnormal HRC might precede the clinical diagnosis of neonatal sepsis, adding independent information to standard clinical parameters.
Methods. In the neonatal intensive care unit at the
University of Virginia, we prospectively studied infants admitted from
August 1995 to April 1999 who were at risk for developing sepsis.
Infants in the sepsis (culture-positive) and sepsis-like illness
(culture-negative) groups had an abrupt clinical deterioration that
raised clinical suspicion of infection and prompted physicians to
obtain blood cultures and start antibiotic therapy. Infants without
sepsis raised no clinical suspicion of illness and had no cultures
obtained. We measured novel characteristics
moments and
percentilesof normalized heart rate (HR) time series for 5 days
before and 3 days after sepsis, sepsis-like illness, or a random time
in controls. We also calculated the Score for Neonatal Acute Physiology
(SNAP) and the Neonatal Therapeutic Intervention Scoring System (NTISS) as clinical scores of the severity of illness.
Results. There were 46 episodes of culture-positive sepsis
in 40 patients and 27 episodes of culture-negative sepsis-like illness
in 23 patients. We analyzed 29 control periods in 26 patients. Infants with sepsis and sepsis-like illness had lower birth weights and gestational ages and higher SNAP and NTISS scores than did infants without sepsis. The most important new finding was that the infants in
the sepsis and sepsis-like illness groups had increasingly abnormal HRC
for up to 24 hours preceding their abrupt clinical deterioration. The
abnormal HRC were reduced baseline variability and short-lived
decelerations in HR. These abnormalities led to significant changes in
HRC measures, for example, the third moment (skewness: .59 ± .10 for sepsis and .51 ± .12 for sepsis-like illness,
compared with 
.10 ± .13 for control over the 6 hours before
abrupt deterioration). Culture-positive and culture-negative patients
had similar HRC and clinical scores, including a significant rise in
SNAP in the 24 hours before the event. Multivariable logistic regression analysis showed that HRC and clinical scores independently added information in distinguishing infants with sepsis and sepsis-like illness from control patients in the 24 hours before abrupt
deterioration.
Conclusions. Newborn infants who had abrupt clinical deterioration as a result of sepsis and sepsis-like illness had abnormal HRC and SNAP that worsened over 24 hours before the clinical suspicion of sepsis. A strategy for monitoring these parameters in infants at risk for sepsis and sepsis-like illness might lead to earlier diagnosis and more effective therapy.heart rate variability, neonatal sepsis, Score for Neonatal Acute Physiology, Neonatal Therapeutic Intervention Scoring System, newborn.
Abrupt clinical deterioration of premature low birth weight
infants in the neonatal intensive care unit (NICU) is a common and
potentially catastrophic event. Because sepsis is high on the list of
possible causes, these episodes prompt physicians to obtain a blood
culture and to start antibiotic therapy. Some episodes are the result
of culture-proven sepsis, while other episodes that are equally
suspicious for sepsis do not have an associated positive blood culture.
Some of the sepsis-like episodes with negative blood cultures may be
caused by true sepsis with a false-negative blood culture, while others
may be initiated by factors such as acute pulmonary or gastrointestinal
diseases.
Sepsis and sepsis-like illnesses may be initiated by cytokines and may
have a subclinical phase where early detection could be beneficial. For
example, Kuster and colleagues1 have recently found
elevated levels of circulating cytokines for up to 2 days before the
clinical diagnosis of neonatal sepsis. Early detection of sepsis
and sepsis-like illnesses in newborn infants, however, remains
difficult using clinical signs and currently available laboratory
tests. Clearly, improved early diagnosis of neonatal sepsis and
sepsis-like illness is an important goal, and new approaches are
required.
We devised a system to examine heart rate characteristics (HRC) for
infants in the NICU and used this to test the hypothesis that HRC are
altered early in the course of sepsis and sepsis-like illnesses.
Because heart rate (HR) patterns are altered during fetal distress and
severe neonatal illness,2-5 we hypothesized that changes
in HRC might occur during the early, subclinical stage of illness. In
particular, we hypothesized that early stages of sepsis might
lead to reduced HR variability and transient HR decelerations, and we
developed statistical measures to quantify these changes. Our
analytical approach was based on the observation that these abnormal HR
patterns lead to asymmetry of frequency histograms of RR
intervals, and we used the moments and percentiles of RR distributions
to report on this asymmetry.
Thus, this study was undertaken to evaluate HRC in infants at risk for
sepsis and sepsis-like illness. To relate these HRC abnormalities to
the level of clinical illness, we measured the Score for Neonatal Acute
Physiology (SNAP) and the Neonatal Therapeutic Intervention Scoring
System (NTISS). The purpose was to determine whether HRC abnormalities
preceded an abrupt clinical deterioration that prompted obtaining a
blood culture and initiation of antibiotics in infants with sepsis or a
sepsis-like illness.
Study Population
We monitored infants who had risk factors for acquiring
late-onset sepsis in the NICU at the University of Virginia from August 1995 to April 1999. These risk factors included low birth weight, prematurity, need for central venous access, and NICU stay >2 weeks.
We studied 3 groups of infants defined by the actions of the physicians
and the results of blood cultures. Infants who had an abrupt clinical
deterioration after 3 days of age that prompted physicians to obtain
blood cultures and to give antibiotics were in the sepsis (positive
blood culture) or sepsis-like illness (negative blood culture) group.
The infants without sepsis group raised no clinical suspicion of sepsis
and had no blood cultures obtained over a 10-day period. HRC monitoring
results were not visible to the treating physicians and did not
influence medical management.
HR Data Acquisition and Analysis
An analog electrocardiogram voltage signal from the bedside
monitor (Marquette, Milwaukee, WI) was digitized and filtered and then
evaluated for QRS complexes using specially-equipped personal computers
(National Instruments AT-DSP2200, Austin, TX). RR intervals were
compared with the previous 100 intervals and excluded if they differed
by >5 standard deviations (SDs). All datasets were visually inspected,
and records with obviously artifactual data were excluded. This
resulted in removal of 1% of all the datasets. HRC measures were
calculated from 4096-beat epochs of RR intervals.
HRC Analysis
We selected HRC measures that give a description of the symmetry
of the histogram of RR intervals. Moments are descriptive statistics
calculated from the individual differences of data points from the
mean. The first moment is itself the mean and the second moment is the
SD, the square root of the average of the squared individual
differences. The third moment or skewness reports on the symmetry of
the histogram. A symmetrical histogram has a skewness of 0, and a
histogram with a tail of values that are larger than the median has
positive skewness.6 We also calculated percentiles of the
data. The median is the 50th percentile data point (p50), meaning that
it resides at the midpoint of the data after sorting from smallest to
largest. In addition to the median, which we called the p50, we also
calculated the p10, p25, p75, and p90. Before these calculations, the
mean and SD were used to normalize the data, so that the mean and SD of
each 4096-beat record were 0 and 1, respectively. This normalization allowed direct comparison of HRC among all records. It is important to
note that these measures are based only on the distribution of the data
points and not the sequence in which they occur. These measures are not
changed by missing points, unlike frequency domain measures.7
Clinical Scores
The SNAP8 and NTISS9 scores were
obtained either prospectively or from chart review by trained research
assistants who were closely supervised (M.P.G.). Scores were calculated
for 24-hour epochs relative to the time that the suspicion of sepsis
was raised. The investigators were blinded to the results of the HRC
analysis at the time of scoring.
Strategy
To examine the time course of HRC early in sepsis, we analyzed
the 5 days before and 3 days after a reference time point. For this, we
used the time that the blood culture was obtained (sepsis and
sepsis-like illness groups) or a random time (infants without sepsis).
The event time for infants without sepsis was assigned randomly during
the sixth or seventh day of their 10-day course. We analyzed the data
in 6-hour epochs based on this reference point. We calculated HRC for
all the 4096-beat datasets and summarized each 6-hour epoch as the
median value of each measure for each patient. We excluded epochs with
<50% of the expected number of heartbeats. For comparison with
clinical illness severity scores in the regression analysis, we
analyzed HRC in 24-hour epochs and compared the results with clinical
scores obtained over the same period.
Statistical Analysis
The significance of differences in demographic characteristics
and HRC for isolated time points was examined using the Mann-Whitney U rank-sum test or analysis of variance (ANOVA;
SigmaStat, Jandel, San Rafael, CA). The significance
of differences between groups for the 24 hours before and after the
event were analyzed using ANOVA with a Tukey test for multiple
comparisons (SigmaStat, Jandel). Receiver operating characteristic
(ROC) analysis was performed using Microsoft Excel (Microsoft
Corportation, Redmond, WA) and the Analyze-it plug-in (Analyze-it
Software, Leeds, United Kingdom). We used multivariable logistic
regression analysis to examine the ability of HRC and clinical scores
to distinguish septic infants from infants without sepsis (S-Plus,
MathSoft, Seattle, WA).
Study Population
Table 1 shows the demographic
characteristics of the infants studied. There were 46 culture-proven
episodes of sepsis in 40 patients. There were 2 deaths associated with
Staphylococcus aureus and Enterococcus infection.
The most common organisms isolated were coagulase-negative
Staphylococcus (n = 20) and S
aureus (n = 15). There were 27 episodes of
culture-negative sepsis in 23 infants. In the control group, there were
29 control periods in 26 patients. In the culture-positive sepsis
group, the mean (SD) birth weight, gestational age, and
postconceptional age at event were 784 g (409), 26 weeks (2), and
31 weeks (4). In the culture-negative sepsis-like illness group, the
values were 756 g (258), 26 weeks (2), and 30 weeks (3). In the
infants without sepsis, the values were 976 g (265), 28 weeks (3),
and 33 weeks (3). For each parameter, the values for the control group
were significantly higher than for the sepsis and sepsis-like illness
groups (P < .001), but the sepsis group was not
significantly different from the sepsis-like illness group.
TABLE 1
METHODS
Top
Abstract
Methods
Results
Discussion
References
RESULTS
Top
Abstract
Methods
Results
Discussion
References
Population Characteristics
An Example of the HR Analysis
Figure 1A shows a time series of 4096 RR intervals from an infant 6 days before an episode of culture-positive sepsis and represents a normal HR pattern. Figure 1, B and C show abnormal RR interval time series from the same infant that were each obtained within 3 to 6 hours before sepsis was suspected and blood cultures were obtained. Both have a baseline of reduced variability and are punctuated by sharp upward deflections that represent short-lived episodes of HR decelerations. Although episodes of bradycardia in NICU patients are common and not necessarily significant,10 frequent episodes of apnea, which are often associated with HR decelerations, are often interpreted as reflecting early stages of sepsis.11 The HR in these records always exceeded 120 beats per minute, and these episodes would have failed to trigger HR alarms set at usual thresholds of 100 beats per minute. In fact, the mean RR intervals of these records are not very different at 323 (1A), 308 (1B), and 302 msec (1C). Distinguishing the data in Fig 1B from normal is nonetheless straightforward by calculating the SD, here 11 msec, compared with 32 msec for the normal. This measure, however, would fail to diagnose the abnormal time series in Fig 1C, where the episodes of subclinical HR decelerations are sufficient to elevate the SD to an apparently normal value of 26 msec.
|
To diagnose these abnormalities, we developed an approach based on the
frequency histograms of the RR intervals shown in the right-hand column
in Fig 1. The long RR intervals during the decelerations generated
asymmetry of the histogram (Fig 1, E and F). We quantified the symmetry
of histograms using the third moment or skewness, a descriptive
statistic that, like the SD, is based on the differences between
individual data points and the mean. The skewness is positive when
there is a longer tail of values extending toward longer RR
intervals.6 The skewness values of the 3 time series are
different: .12 for the data in 1A, indicating a near-symmetric
distribution, but 1.99 and 1.33 for the data in Fig 1, B and C,
indicating a large degree of asymmetry.
The abnormalities of the histograms can also be quantified by
considering the relationship of values in the distribution to the mean.
Accordingly, we determined the values of 5 percentile values of the
normalized datathe 10th, 25th, 50th (median), 75th, and 90th. These
parameters differed among the 3 datasets shown. For example, the p50
was more negative in the abnormal datasets. The values were .13 in the
normal record in Fig 1A but .21 and .33 in the abnormal records
shown in Fig 1, B and C. This change results from the preponderance of
values to the left of the center of the distribution, a consequence of the asymmetry of the histogram.
Summary HRC Data
Figure 2 shows the time course of
HRC measures for the 3 patient groups. Each data point is the mean of
the median values for 6 hours, and bars are standard error of the mean
(SEM). The vertical line marks the time of the abrupt clinical
deterioration for which blood cultures were obtained and antibiotics
were started. The label CRASH stands for Cultures, Resuscitation, and
Antibiotics Started Here. The data points at time 0 represent the
6-hour epoch before, but not including, the time of the deterioration.
Values of mean RR interval (Fig 2A) and SD (Fig 2B) did not
discriminate among the groups. The skewness (Fig 2C) and p50 values
(Fig 2D), in contrast, changed markedly in the epochs from 24 hours
before to 24 hours after diagnosis. The skewness was .59 ± .10 for the sepsis group and .51 ± .12 for the sepsis-like illness
group, compared with .10 ± .13 for control over the 6 hours
before clinical suspicion. The values for sepsis (culture-positive) and
sepsis-like illness (culture-negative) infants were not significantly
different, but both were different from the control values
(P < .001). The p50 was .0298 ± .014 for
sepsis and .0223 ± .015 for sepsis-like illness, compared with
+.0503 ± .016 for control over the 6 hours before clinical suspicion. Again, the values for sepsis and sepsis-like illness groups
were not significantly different, but both were different from the
control values (P < .001).
|
Clinical Scores
Figure 3 shows the SNAP and NTISS scores. The sepsis and sepsis-like illness groups had higher levels of both scores throughout, suggesting a greater degree of illness and intensity of interventions. The NTISS level declined over the 10-day period in the control group, consistent with declining requirement for therapies and gradual weaning from support systems. On day 0, the 24 hours before (but not including) the clinical suspicion of sepsis, SNAP rose further in the sepsis and sepsis-like illness groups (P = .01, ANOVA) and there was no decline in NTISS. Over the 24 hours before and after the CRASH, the values of SNAP and NTISS for sepsis (culture-positive) and sepsis-like illness (culture-negative) infants were not significantly different, but both were different from the control values (P < .001).
|
Multivariable Logistic Regression Analysis
We performed multivariable logistic regression analysis on the data from the 24 hours before the clinical suspicion of sepsis. HRC were represented by the median values of each of the 5 percentiles, and clinical data were represented by both the SNAP and NTISS scores for this period. We pooled the data from the sepsis and sepsis-like illness groups and tested whether a regression model could distinguish the pooled data from the control group. The finding was that the groups were highly significantly different (P < .0001; ROC area: .9). Although no single HR percentile measure made a significantly greater contribution than did the others to the discriminatory ability of the model, SNAP contributed significantly more than NTISS (P < .003, Wald z test). Both HRC and clinical scores contributed independently to the final model (P = .02). Results were similar when only sepsis or sepsis-like illness infants were analyzed, and for regression models using HR moments rather than percentiles.
Dynamic Changes in HRC and Illness Severity Scores Early in the Course of Sepsis
We used ROC analysis to quantify differences among the groups and to examine the time course of the differences. The area under the ROC plot is .5 when the groups are not different, and it is 1.0 when the groups are entirely distinct. Panels A and B of Fig 4 show the results for HR moments and percentiles. For this analysis, we show the results after pooling data from the sepsis and sepsis-like groups, but the results for either group individually were very similar. Although mean and SD did not discriminate controls from sepsis and sepsis-like illness infants, skewness showed dynamic differences beginning 12 to 24 hours before the diagnosis. For example, there was a significant difference between the values of skewness measured in the sepsis groups 3 days before sepsis, compared with values 6 to 12 hours before (P < .05). The difference then resolved over the 2 days after diagnosis and antibiotic therapy. Two of the HR percentiles, p10 and p50, showed a similar course. SNAP and NTISS also discriminated between the control group and the sepsis and sepsis-like illness groups, and SNAP showed an increasing difference for 1 to 2 days before the diagnosis. The increasing ROC area for NTISS is caused by the failure of NTISS to decline in the sepsis and sepsis-like illness infants shown in Fig 3B.
|
We also used multivariable logistic regression models to distinguish
the groups. As shown in Fig 4D, models that used HRC alone and clinical
scores (SNAP and NTISS together) alone effectively discriminated the
groups, with larger changes at least 1 day before the diagnosis. A
model using both HRC and clinical scores was superior, in keeping with
the finding above that both HRC and clinical data contributed
independently to the model. There were dynamic changes in the
regression model values. A comparison of regression model values for
day 3 with subsequent days showed that the increases on day 0 were
statistically significant for all models (P < .05).
| |
DISCUSSION |
|---|
|
Top
Abstract
Methods
Results
Discussion
References
|
|---|
We studied the time course of HRC and clinical illness severity scores in infants at risk for late-onset sepsis in the NICU of a tertiary care hospital. Our most important finding is that abrupt clinical deteriorations that prompted physicians to obtain blood cultures and start antibiotics were preceded for up to 24 hours by increasingly abnormal HRC of reduced baseline variability and subclinical, short-lived decelerations in HR, and by increasingly abnormal SNAP scores.
Abnormalities of HR including reduced beat-to-beat variability and episodes of bradycardia have long been recognized in fetal and neonatal distress.3-5 These abnormalities should lead to asymmetry of frequency histograms of RR intervals. In this study, we found that the third moment (skewness) and the 10th, 25th, 50th, 75th, and 90th percentile values of normalized RR distributions distinguished records of infants without sepsis from those in early stages of sepsis or sepsis-like illness. Thus, these novel measures serve to quantify well-established markers of early fetal and neonatal distress, and they may add to clinical observations by detecting subclinical changes in HRC. In addition, the new measures have the advantage of reliability in datasets with missed beats, unlike conventional frequency domain measures of HR time series.7,12
Early diagnosis of neonatal sepsis is difficult,13 because the clinical signs are neither uniform nor specific. One limitation of our study is that we did not standardize the criteria for obtaining blood cultures and giving antibiotics, but rather observed the practices of the physicians. The threshold for obtaining blood cultures and giving antibiotics differs among physicians, and our study groups may have been heterogenous for level of clinical illness as a result. The rise in SNAP for the 24 hours before the CRASH, however, confirms that there was worsening of physiologic variables in the sepsis groups, as expected for infants with clinical deterioration.
Not all patients with clinical signs of sepsis have positive blood cultures. The current hypothesis is that the clinical syndrome is brought about by the host response to insults, such as bacterial infection. The major host response is release of cytokines, small circulating peptides that serve as mediators of the inflammatory response. The syndrome common to sepsis and sepsis-like illness has been named the systemic inflammatory response syndrome,14 and the pathogenesis suggested to be an imbalance between proinflammatory and antiinflammatory effects of cytokines.15
We do not know the mechanism of the HRC changes. In sepsis and
sepsis-like illness, circulating cytokines play a major role in
initiating and maintaining the inflammatory response, and cytokine levels correlate with the severity of illness.16-18
Cytokines have widespread effects on signal transduction processes and
may interfere with normal events of HR control by the sympathetic and
parasympathetic nervous systems. For example, the cytokines tumor
necrosis factor-
, interleukin (IL)-1
, and IL-6 increase HR but
blunt HR responses to -adrenergic agonists.19,20 In
addition, sepsis and sepsis-like illness are associated with alterations in -adrenergic receptor number and
distribution21,22 and with multiple steps of signal
transduction via -adrenergic receptors.23 The recent
demonstration that circulating levels of the cytokines IL-1 receptor
antagonist and IL-6 are elevated for 2 days before clinical evidence of
neonatal sepsis and sepsis-like illness1 is consistent
with our finding of changes in HRC and clinical scores 12 to 24 hours
before the clinical diagnosis. Interestingly, HR variability is reduced
by infusion of endotoxin in human volunteers.24 More
precise delineation of the mechanism of the findings awaits a better
understanding of neonatal HR control during illness and health.
Although the cause of illness is not known with certainty in the infants with negative blood cultures, we suspect that infection was the cause in several. Although the blood culture is believed to be the gold standard for establishing the diagnosis of sepsis caused by systemic bacterial infection, there are concerns regarding its reliability,25 especially if single samples of small volume are submitted26,27 as is often the practice in critically ill newborn infants. For example, as many as 60% of culture results may be falsely negative if only .5 mL of blood is obtained from infants with low colony count sepsis.28 In a study of 298 aerobic culture specimens, the mean blood volume submitted was .53 mL and 55% of samples contained <.5 mL.29 We might then expect 30% to 40% of all infants with sepsis to have negative blood cultures. This is in agreement with 2 studies in which premortem blood cultures failed to identify ~20% of infants with infection proven by postmortem cultures and autopsy.30,31
The findings are limited by differences in birth weight, gestational age, and postconceptional age in the infants without sepsis, compared with the other groups. Infants who developed sepsis and sepsis-like illness had lower birth weight (~200 g) and gestational age and postconceptional age (~2 weeks) and would be recognized clinically as having increased risk. They also had more abnormal HRC, SNAP, and NTISS at 4 days before the abrupt clinical deterioration, reflecting this baseline difference of increased risk. Our definition required at least 10 consecutive days during which the infant did not receive antibiotics and did not have symptoms suggestive of sepsis that prompted physicians to obtain a blood culture. This eliminated some of the high-risk very low birth weight and gestational age infants, because their increased probability of sepsis and sepsis-like illness frequently led physicians to obtain blood cultures and give antibiotics. Thus, we were unable to match precisely the characteristics of the groups with sepsis and sepsis-like illness.
Importantly, we found dynamic changes of SNAP and HRC for up to 24 hours before the abrupt clinical deterioration. We found, for example, that the SNAP score rose in the 1-day epoch before the abrupt clinical deterioration, suggesting that there is diagnostic information in conventional vital signs and laboratory tests. Because SNAP is calculated only on a 24-hour basis, however, we do not know whether the physiologic and laboratory abnormalities measured by SNAP appeared just before the clinical suspicion of sepsis and sepsis-like illness or whether they occurred gradually over the 24-hour period. We also found significant dynamic changes in novel HRC, such as skewness and p50, that may add to clinical observation in early detection of sepsis and sepsis-like illness. An advantage of HRC over clinical scores is their noninvasive and continuous nature.
Clinicians have found no clinical signs or laboratory test findings to be reliable for very early diagnosis of neonatal sepsis, and 10 to 20 infants are treated for each infant with a positive blood culture.32 Successful surveillance strategies leading to an earlier diagnosis of sepsis are urgently needed for very low birth weight infants to decrease mortality and morbidity.33 Our findings suggest that monitoring of HRC might be developed into a useful adjunct to conventional clinical observation and practice. The analysis presented here based on clinical characteristics and HRC is effective in distinguishing low-risk infants having a stable course from high-risk infants hours before the clinical diagnosis of sepsis. A more useful test would be to detect early, subclinical sepsis in high-risk infants whose baseline HRC are not normal. There was a significant difference between the values of skewness measured in the sepsis groups 3 days before sepsis, compared with values 6 to 12 hours before. This finding affirms the dynamic nature of the HRC changes and motivates further research on clinical implementation of HRC monitoring.
| |
ACKNOWLEDGMENTS |
|---|
This work was supported by the American Heart Association, Mid-Atlantic Consortium; Children's Medical Center Research Fund, University of Virginia; Virginia's Center for Innovative Technology; and Medical Automation Systems, Charlottesville, Virginia.
We thank T. Smoot, J. Nelson, K. Monahan, J. Richman, A. Dinn, H. Huq, F. Artrip, R. Vaughn, T. Tamburello, W. King, and S. Booth for integral contributions to the data acquisition and analysis; F. Harrell and E. Bissonette for regression analysis; and G. Beller, J. Kattwinkel, and B. Duling for support.
| |
FOOTNOTES |
|---|
Received for publication Dec 15, 1999; accepted Mar 31, 2000.
This work was presented in part at the meeting of the Pediatric Academic Societies; May 4, 1999; San Francisco, CA.
Reprint requests to (M.P.G.) Department of Pediatrics, Box 386, University of Virginia, Charlottesville, VA 22908. E-mail: mpg7 u{at}virginia.edu
| |
ABBREVIATIONS |
|---|
NICU, neonatal intensive care unit; HRC, heart rate characteristics; HR, heart rate; RR interval, interval between heartbeats; SNAP, Score for Neonatal Acute Physiology; NTISS, Neonatal Therapeutic Intervention Scoring System; SD, standard deviation; p50, 50th percentile data point; ANOVA, analysis of variance; ROC, receiver operating characteristic; SEM, standard error of the mean; CRASH, Cultures, Resuscitation, and Antibiotics Started Here; IL, interleukin.
| |
REFERENCES |
|---|
|
Top
Abstract
Methods
Results
Discussion
References
|
|---|
-adrenoceptors in the rat liver during different phases of sepsis.
Biochim Biophys Acta
1998;
1407:225-233 [Medline]-adrenoceptors and cAMP levels during the late stage of sepsis.
Biochim Biophys Acta, 1998;1404:377-384-adrenergic receptor-dependent and -independent stimulation of
adenylate cyclase is impaired during severe sepsis in humans.
Int
Care Med
1998;
24:1315-1322 [CrossRef][Medline]This article has been cited by other articles:
 |
|
 |
|
  S. Blad, A.-K. Welin, I. Kjellmer, K.G. Rosen, and C. Mallard ECG and Heart Rate Variability Changes in Preterm and Near-Term Fetal Lamb Following LPS Exposure Reproductive Sciences, July 1, 2008; 15(6): 572 - 583. [Abstract] [PDF]
|
|
|
|
 |
|
 J. P. Cimiotti, J. Haas, L. Saiman, and E. L. Larson Impact of Staffing on Bloodstream Infections in the Neonatal Intensive Care Unit Arch Pediatr Adolesc Med, August 1, 2006; 160(8): 832 - 836. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. P. Griffin, D. E. Lake, E. A. Bissonette, F. E. Harrell Jr, T. M. O'Shea, and J. R. Moorman Heart Rate Characteristics: Novel Physiomarkers to Predict Neonatal Infection and Death Pediatrics, November 1, 2005; 116(5): 1070 - 1074. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. P. Griffin, D. E. Lake, and J. R. Moorman Heart Rate Characteristics and Laboratory Tests in Neonatal Sepsis Pediatrics, April 1, 2005; 115(4): 937 - 941. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Goldstein Heart Rate Characteristics in Neonatal Sepsis: A Promising Test That Is Still Premature Pediatrics, April 1, 2005; 115(4): 1070 - 1072. [Full Text] [PDF]
|
|
|
|
 |
|
 D. Kaufman and K. D. Fairchild Clinical Microbiology of Bacterial and Fungal Sepsis in Very-Low-Birth-Weight Infants Clin. Microbiol. Rev., July 1, 2004; 17(3): 638 - 680. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Malik, C. P. S. Hui, R. A. Pennie, and H. Kirpalani Beyond the Complete Blood Cell Count and C-Reactive Protein: A Systematic Review of Modern Diagnostic Tests for Neonatal Sepsis Arch Pediatr Adolesc Med, June 1, 2003; 157(6): 511 - 516. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. E. Lake, J. S. Richman, M. P. Griffin, and J. R. Moorman Sample entropy analysis of neonatal heart rate variability Am J Physiol Regulatory Integrative Comp Physiol, September 1, 2002; 283(3): R789 - R797. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Turksen and T.-C. Troy Permeability barrier dysfunction in transgenic mice overexpressing claudin 6 Development, January 4, 2002; 129(7): 1775 - 1784. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2001 American Academy of Pediatrics. All rights reserved. |
||
|
||
