 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
PEDIATRICS Vol. 106 No. 1 July 2000, pp. 118-122
,
, and
From * Children's Hospital of Pittsburgh, Department of
Pediatrics, Division of Pediatric Emergency Medicine; University of
Pittsburgh, Departments of Anesthesia/Critical Care Medicine and the
Safar Center for Resuscitation Research; § Mercy Hospital, Department
of Pediatrics; and Geriatric Research Educational and Clinical
Center, VA Pittsburgh Health System and the University of Pittsburgh,
Department of Neurology, Pittsburgh, Pennsylvania.
 |
ABSTRACT |
|---|
Top
Abstract
Methods
Results
Discussion
Conclusion
References
|
|---|
Objective. In experimental models of ischemic-anoxic brain injury, changes in body temperature after the insult have a profound influence on neurologic outcome. Specifically, hypothermia ameliorates whereas hyperthermia exacerbates neurologic injury. Accordingly, we sought to determine the temperature changes occurring in children after resuscitation from cardiac arrest.
Study Design. The clinical records of 13 children
resuscitated from cardiac arrest were analyzed. Patients were
identified through the emergency department and pediatric intensive
care unit arrest logs. Only patients surviving for 
12 hours after
resuscitation were considered for analysis. Charts were reviewed for
body temperatures, warming or cooling interventions, antipyretic and
antimicrobial administration, and evidence of infection.
Results. Seven patients had a minimum temperature (T min)
of 
35°C and 11 had a maximum temperature (T max) of 38.1°C.
Hypothermia often preceded hyperthermia. All 7 patients with T min
35°C were actively warmed with heating lamps and 5 of 7 responded
to warming with a rebound of body temperatures 38.1°C. None of the
6 patients with T min >35°C were actively warmed but all developed T
max 38.1°C. Six patients received antipyretics and 11 received
antibiotics. Fever was not associated with a positive culture in any
case.
Conclusion. Spontaneous hypothermia followed by hyperthermia is common after resuscitation from cardiac arrest. Temperature should be closely monitored after cardiac arrest and fever should be managed expectantly. Key words: cardiac arrest, temperature, fever, hypothermia.
There is a growing body of literature demonstrating a
neuroprotective effect of hypothermia in a variety of animal species and mechanisms of injury.1 These observations have
led to recent clinical trials of hypothermia in humans after traumatic brain injury2 and resuscitation from cardiac
arrest.3,4 In contrast to the neuroprotective effect of
hypothermia, hyperthermia has considerable potential to exacerbate
ischemic brain injury.5 Despite the emerging recognition
that hypothermia may be neuroprotective and hyperthermia is injurious,
it is a common clinical practice to actively warm critically ill
moderately hypothermic patients (particularly children) after cardiac
arrest. Accordingly, we sought to examine our experience with pediatric
patients resuscitated from cardiac arrest. Specifically, we reviewed
charts to determine the frequency of documented (potentially
beneficial) hypothermia and documented (potentially harmful)
hyperthermia and the clinical response to these temperature changes.
Patients were identified through the cardiac arrest logs of the
emergency department (ED) and pediatric intensive care unit (PICU).
Records were considered for review if both chest compressions and
ventilation were documented and the patient survived for The PICU arrest log from 1997 identified 14 potential patients.
The ED arrest logs identified an additional 11 potential patients. Twelve patients were excluded: 1 went immediately to the operating room, 1 had overwhelming sepsis, 1 was placed immediately on
extracorporeal membrane oxygenation, 3 had incomplete arrests
(respiratory arrest only or hypotension only), 1 had only axillary
temperatures recorded, and 5 had either inadequate documentation to
confirm cardiac arrest or records were unavailable for review. A total
of 13 patient records were entered into the dataset.
Table 1 includes a brief clinical
synopsis of the 13 patients. A minimum temperature (T min) of
TABLE 1
METHODS
Top
Abstract
Methods
Results
Discussion
Conclusion
References
12 hours
after resuscitation. The 12 hours survival time was chosen to eliminate
patients with hypothermia secondary to equilibration with environmental
temperature after prolonged periods of cardiac arrest (eg, infants with
brief return of spontaneous circulation). All charted oral and rectal
temperatures were entered into the data collection set. Because
axillary temperatures can underestimate core temperature and,
therefore, may be unreliable for documenting hypothermia (but are
reliable for documenting fever), only values 38.1°C were entered
into the dataset.6 Charts were also reviewed for
documentation of warming or cooling interventions, antipyretic and
antimicrobial administration, and evidence of infection. Patients with
interventions that would independently effect body temperature (eg,
transfer to the operating suite immediately after resuscitation or
institution of extracorporeal membrane oxygenation) and whose
charts contained inadequate documentation were excluded from analysis.
RESULTS
Top
Abstract
Methods
Results
Discussion
Conclusion
References
35°C was documented in 7 patients and a maximum temperature (T max)
of 38.1°C was documented in 11 patients. Temperature data for
individual patients are presented in Fig
1 and Fig
2. In general, hypothermia preceded
hyperthermia. All 7 patients with T min 35°C were actively warmed
with heating lamps. Remarkably, 5 of 7 responded to warming with a
rebound of body temperature 38.1°C. None of the 6 patients with T
min >35°C were actively warmed, however, all developed a maximum
temperature 38.1°C. Six patients had as needed (prn) orders for
antipyretics on admission to the hospital and 1 patient had a prn order
written after the development of fever. Nine patients had antibiotic
orders written on admission to the hospital and 2 had antibiotic orders
written after the development of fever. All specimens cultured for
bacteria were negative including 9 blood, 4 urine, 1 stool, 1 cerebrospinal fluid and 1 tracheal secretion specimens. In addition,
there were 3 respiratory secretion cultures negative for virus. Two
patients had fetid diarrhea consistent with mucosal sloughing. Chest
radiographs were obtained in all patients: 1 had right lower lobe
pneumonia, 2 had pulmonary edema, and 6 had evidence of atelectasis.
Clinical Synopses of the Patients
View larger version (17K):
[in a new window]
Fig. 1.
Temperatures of patients who did not have heating or cooling
interventions after cardiac arrest. Legend refers to case histories
listed in Table 1.
View larger version (21K):
[in a new window]
Fig. 2.
Temperatures of patients who were actively warmed or cooled after
cardiac arrest. Legend refers to case histories listed in Table 1. H
indicates heating lamps applied; C, cooling blanket applied.
| |
DISCUSSION |
|---|
|
Top
Abstract
Methods
Results
Discussion
Conclusion
References
|
|---|
We found that children resuscitated from cardiac arrest frequently
present with hypothermia and that this is often followed by the
development of fever. Hypothermia preceding hyperthermia has also been
documented in adults resuscitated from cardiac arrest.7 Spontaneous hypothermia has also been observed in rodent models of
asphyxial cardiac arrest and 4-vessel occlusion models of global ischemia.8,9 One explanation of the initial hypothermia is
that equilibration with the environmental temperature occurs during
circulatory arrest as discussed by Biggart and Bohn.10 However, because we selected patients surviving for at least 12 hours
after resuscitation, they were less likely to have long periods of
absent blood flow during which equilibration would occur. Furthermore,
5 of the patients experienced in-hospital arrest with relatively rapid
resuscitation and yet developed delayed episodes of hypothermia.
Therefore, it is likely that the hypothermia was secondary, at least in
part to either a generalized reduction of metabolic rate or a central
nervous system mediated effects
in addition to equilibration with
environmental temperature.
The cause of delayed fever after cardiac arrest is unclear. In a prospective study, 13 of 33 (39%) adults resuscitated from cardiac arrest were bacteremic.11 Twelve of the bacteremic patients had enteric pathogens and an associated fetid diarrhea suggesting translocation of bacteria across ischemic gut. However, only 2 of our patients had documentation of diarrhea and there were no positive blood cultures among the 9 patients in whom blood cultures were obtained. Another possible cause for fever is a response to atelectasis or aspiration pneumonia. A prospective study concluded that pulmonary pathology is a frequent cause of fever in adults with stroke.12 The high frequency of chest radiograph abnormalities in our study (9 of 13 patients) supports pulmonary pathology as a potential cause of fever in children resuscitated from cardiac arrest. Another potential cause is the use of warming lights. However, the observation that fever developed in all 6 patients in whom warming lights were not used strongly suggests that fever is not solely an artifact of warming interventions. For those patients in whom warming lights were used, it is possible that some would have developed fever in the absence of heating lights but the use of lights may have facilitated an earlier onset and higher amplitude of fever. It is likely that multiple mechanisms contributed to the development of fever in our cohort of children and that the relative contribution of any single mechanism varied between patients.
The benefit of intraischemic hypothermia on neurologic outcome has been well-recognized for decades. Initial observations of remarkably favorable outcomes after resuscitation from ice-water drownings (etc) were harnessed into medical use with the application of induced hypothermia for neurologic and cardiac surgery. Recently, experiments have demonstrated a beneficial effect of induced hypothermia after brain injury in a variety of animal models. The studies are encouraging for 2 reasons: first, the hypothermia used was mild and therefore unlikely to cause physiologic derangements, and second, the protection was markedly robust. The laboratory data are supported by limited clinical trials of induced hypothermia in humans after traumatic brain injury2 and resuscitation from cardiac arrest.3,4 The resurgence in interest in hypothermia as a neuroprotectant has also resulted in initiation of a multicenter clinical trial in asphyxiated newborns.13
Risks of hypothermia include myocardial depression, coagulopathy, and impaired immune function.14 In addition, a paradoxical increase in metabolic demands may occur if the patient is capable of shivering. The risk of untoward effects increases with the depth and duration of hypothermia. Clinical trials in humans have selected a target temperature of 33°C (for 12-24 hours).2-4 Among the 7 patients in our study who were actively warmed, 4 were warmed for a minimum temperature above 33°C.
It is not possible to determine from this study whether the observed changes in temperature influenced survival or neurologic outcome in the patients studied. However, an association between fever and worsened neurologic outcome has been described in humans after stroke5 and cardiac arrest.7 One explanation for the association is that fever may be a marker for severity of the initial insult. For example, longer arrest times are more likely to result in gut ischemia and translocation of bacteria that can induce fever. Similarly, patients with more severe insults might have increased risk of aspiration with the attendant development of fever. However, there are recent compelling animal data supporting an independent deleterious effect of fever on neurologic injury. Baena et al demonstrated worsened outcome in rats warmed to 39.6°C for 3 h at one day after forebrain ischemia compared with controls.15 Kim et al reported similar results with delayed hyperthermia versus normothermia in a rodent model of stroke.16 Finally, Coimbra et al administered an antipyretic to rats subjected to global brain ischemia. Rats given the antipyretic were less likely to develop fever during recovery and had improved neurologic outcome.17 If fever can exacerbate ischemic brain injury, our finding of fever in 11 of 13 patients suggests that this is a significant clinical problem. Moreover, because brain temperature can significantly exceed body temperature in injured brain, the magnitude and frequency of elevated brain temperature is likely underreported in this study.18-20 Furthermore, the potential for injudicious use of warming lights and the inconstant use of antipyretics or cooling blankets suggests that there is room for improvements in clinical care. It is noteworthy that spontaneous or postrewarming hyperthermia developed in the PICU in Pittsburgh despite having a heightened awareness of the potential deleterious consequences of hyperthermia.21
We recognize that the retrospective design of the study, the uncontrolled use of warming/cooling measures, and the various diagnostic methods used to investigate source of fever are limitations of this study. However, cardiac arrest in children is a rare event that prohibits a prospective study from being completed in a reasonable time interval in a single institution. Furthermore, we believe that although the data are imperfect the observations are reliable and have important potential implications for current clinical practice. Although prospectively collected data would increase our confidence in the frequency of events (eg, hypothermia/hyperthermia), our study is sufficient evidence that these events occur frequently enough to be of clinical significance.
| |
CONCLUSION |
|---|
|
Top
Abstract
Methods
Results
Discussion
Conclusion
References
|
|---|
In summary, spontaneous hypothermia followed by hyperthermia is common after resuscitation from cardiac arrest. Although it is premature to recommend routine induction of hypothermia, it appears prudent to avoid aggressive attempts to rewarm hemodynamically stable patients with spontaneously developed mild-moderate hypothermia after cardiac arrest. It is concerning that rewarming measures appear to facilitate the development of fever that could exacerbate brain injury. Finally, temperature should be closely monitored after cardiac arrest and fever should be managed expectantly.
| |
FOOTNOTES |
|---|
Received for publication Aug 9, 1999; accepted Nov 22, 1999.
Reprint requests to (R.W.H.) Division of Pediatric Emergency Medicine, Children's Hospital of Pittsburgh, 3705 Fifth Ave, Pittsburgh, PA 15213-2583. E-mail: hickeyr+@pitt.edu
| |
ABBREVIATIONS |
|---|
ED, emergency department; PICU, pediatric intensive care unit; prn, as needed.
| |
REFERENCES |
|---|
|
Top
Abstract
Methods
Results
Discussion
Conclusion
References
|
|---|
This article has been cited by other articles:
 |
|
 |
|
  The International Liaison Committee on Resuscitati The International Liaison Committee on Resuscitation (ILCOR) Consensus on Science With Treatment Recommendations for Pediatric and Neonatal Patients: Pediatric Basic and Advanced Life Support Pediatrics, May 1, 2006; 117(5): e955 - e977. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Part 6: Pediatric Basic and Advanced Life Support Circulation, November 29, 2005; 112(22_suppl): III-73 - III-90. [Full Text] [PDF]
|
|
|
|
 |
|
 H. Kayman Management of Fever: Making Evidence-Based Decisions Clinical Pediatrics, June 1, 2003; 42(5): 383 - 392. [PDF]
|
|
|
|
|
|
A. J. Gunn and P. D. Gluckman Should We Try to Prevent Hyperthermia After Cardiac Arrest? Pediatrics, July 1, 2000; 106(1): 132 - 133. [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2000 American Academy of Pediatrics. All rights reserved. |
||
|
||
