- Email: [email protected]
Measuring brain temperature while maintaining brain normothermia in patients with severe traumatic brain injury
Journal of Clinical Neuroscience 18 (2011) 1059–1063
Contents lists available at ScienceDirect
Journal of Clinical Neuroscience journal homepage: www.elsevier.com/locate/jocn
Clinical Study
Measuring brain temperature while maintaining brain normothermia in patients with severe traumatic brain injury Jinn-Rung Kuo a,b,c, Chong-Jeh Lo a,d, Che-Chuan Wang c, Chin-Li Lu e, Shu-Chin Lin f, Chiao-Fang Chen f,⇑ a
Institute of Clinical Medicine, School of Medicine, National Cheng-Kung University, Tainan, Taiwan Department of Biotechnology, Southern Taiwan University, Tainan, Taiwan c Department of Neurosurgery, Chi-Mei Medical Center, Tainan, Taiwan d Division of Trauma and Emergency Surgery, Chang Gung Memorial Hospital–Kaohsiung Medical Center, Kaohsiung, Taiwan e Institute of Medical Research, Chi-Mei Medical Center, Tainan, Taiwan f Advanced Practices Nursing, Department of Nursing, Chi-Mei Medical Center, 901 Chung Hwa Road, Yung Kang City, Tainan, Taiwan b
a r t i c l e
i n f o
Article history: Received 29 July 2010 Accepted 2 November 2010
Keywords: Brain normothermia Intracranial temperature Rectal temperature Superficial temporal artery temperature
a b s t r a c t The aim of this study was to evaluate the relationship between superficial temporal artery temperature (Tt), rectal temperature (Tr) and intracranial temperature (ICT) when attempting to keep the brain in a normothermic condition in patients with severe traumatic brain injury (TBI). We also compared the incidence of temperature gradient reversal in patients who survived (survivors) and patients who did not (non-survivors) and the difference in temperature gradient reversal between survivors and non-survivors. Tr is normally lower than and ICT and temperature gradient reversal, when Tr exceeds ICT, has been demonstrated to be an early sign of brain death. A total of 28 patients with severe TBI were enrolled retrospectively in our study between November 2008 and February 2010. Each patient’s Tt, Tr and ICT was recorded every hour for 4 days. Our results show that the frequency of brain hyperthermia in our participants (ICT > 38 °C) was 17.7%. Using a paired t-test and Bland–Altman plots, it was shown that a significant temperature difference existed between Tt, Tr and ICT (p < 0.001). The use of Spearman’s correlation method revealed that Tt, Tr and ICT were positively correlated (p < 0.001). Brain death occurred in five patients at a mean of 9.6 hours (range: 8–12 hours) after a temperature gradient reversal between Tt, Tr and ICT. Fisher’s exact test showed that there was a significant difference in the incidence of temperature gradient reversal between Tt, Tr and ICT in survivors and non-survivors (p < 0.001). We conclude that a significant temperature difference exists between Tt, Tr and ICT when maintaining brain normothermia. The daily practice of non-invasive Tt measurement may cause doctors to underestimate ICT; reversal of the ICT and Tt and/or Tr temperatures could be an early marker of a poor prognosis for patients with severe TBI. Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction Traumatic brain injury (TBI) affects up to 2% of the population per year and is a major cause of disability and death among young people.1 Hyperthermia is frequently seen in patients during the acute injury period following a TBI, and the incidence and duration of the fever are significantly associated with neuronal damage.2,3 Previous studies have shown that the incidence of hyperthermic rates is 16% within 24 hours post-admission, 31.7% between 24 hours and 48 hours post-admission, 42% at 72 hours post-admission, and 60% to 70% between 48 hours and 96 hours post-admission.4–6 The difference between ICT and Tr has been evaluated via invasive temperature monitoring, and brain temperature is 0.5 °C to 1 °C higher than rectal temperature.7–9 The measurement of Tt, a non-invasive procedure that uses technology commonly found in ⇑ Corresponding author. Tel.: +886 6 2812811; fax: +886 6 2828928. E-mail address: [email protected] (C.-F. Chen). 0967-5868/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.jocn.2010.11.014
intensive care units (ICU), offers potential benefits in critically ill patients.10 Until now, a relationship between Tr, Tt and ICT has not been well established in patients with TBI. Dissociation, or reversal, between ICT and systemic temperatures, defined as when Tr exceeds ICT, has been demonstrated to be an early sign of brain death.8,11 However, the incidences and differences in temperature gradient reversals in patients who survive TBI (survivors) and patients who do not (non-survivors) have not been well evaluated. The aim of this study is to determine the relationship between Tr, Tt and ICT when attempting to keep the brain in a normothermic condition in patients in the ICU with a TBI. We also compared the incidences and differences in temperature gradient reversals in survivors and non-survivors. 2. Material and methods A total of 28 patients diagnosed with a TBI in a medical center in southern Taiwan from November 2008 to February 2010 were
1060
J.-R. Kuo et al. / Journal of Clinical Neuroscience 18 (2011) 1059–1063
evaluated retrospectively and enrolled in our study. This study was approved by the ethics committee of our hospital. A complete history was obtained, and a neurological assessment performed on each patient. All patients underwent a CT scan and were evaluated based on the initial Glasgow Coma Scale (GCS) score in the emergency room. All patients underwent a craniectomy for hematoma removal. The ipsilateral superficial temporal artery was preserved when the craniectomy was performed. An intraparenchymal catheter coupled with a thermistor (110-4BT, Pressure–Temperature Monitoring Kit, Integra Camino, San Diego, CA, USA) was implanted 3-cm deep, frontally, in the parenchyma of the injured hemisphere to standardize monitoring of intracranial pressure (ICP) and ICT. The methods and calibration of the brain temperature sensors and the monitoring techniques used in this study have been reported previously.12 The temperature data are precise to ± 3.0 °C within the range of 20 °C to 40 °C (according to the manufacturer). Postoperatively, patients underwent cerebral perfusion pressure (CPP)-guided management to maintain a CPP of P 60 mmHg and an ICP of 6 25 mmHg. The mean arterial blood pressure (MAP) and CPP were monitored using standard pressure transducers. All patients were sedated using propofol (0.5–6 mg/kg per hour). We also used atracurium (0.3–1.2 mg/kg per hour) if the patient was shivering during the postoperative period. All temperature measurements were recorded in degrees Celsius (°C) every hour for 4 days. The room temperature in the ICU was set at 25 °C. Tr was monitored with a rectal temperature thermistor (Medi-Therm II Hyper–Hypothermia Machine; Gaymar Industries, Orchard Park, NY, USA), and Tt was obtained with a temporal scanner (TemporalScanner™ TAT-5000, EXERGEN, Boston, MA, USA). Tt measurements were performed according to the following ICU standard temperature measurement protocol. The patient’s hair was brushed aside, if necessary, and the probe was positioned on the center of the patient’s forehead. The measurement button was pressed, and the measurement was performed by scanning the region from the center midline and sliding the probe across the head to the region behind the earlobe. The button was then released, and the temperature was recorded. The duration of this procedure was about 5 s. The Tt was evaluated on the side of the head where the operation was performed (defined as the craniectomy side). Brain hyperthermia,13 defined as an ICT > 38 °C, was controlled with an ice pillow (if ICT > 37.5 °C), the oral administration of 500 mg of acetaminophen (if ICT > 38 °C), or with an ice blanket device (if acetaminophen failed) (Gaymar Medi-Therm II Hyper– Hypothermia Machine). The correlation between Tr, Tt and ICT was evaluated using Pearson’s correlation method. The temperature gradient of ICT versus (vs.) Tr, and ICT vs. Tt at each point was calculated and plotted as described by Bland and Altman.14 Data were expressed as the median (interquartile range, IQR) or mean ± standard deviation (SD). All data were analyzed using statistical software (Statistical Package for the Social Sciences for Windows, version 16; SPSS, Chicago, IL, USA). A p value of < 0.05 was considered statistically significant.
3. Results In this study, 28 patients with TBI were evaluated retrospectively (19 male and 9 female; age range, 16–80 years; median age of 44.5 years; age IQR of 28–77.5 years). Traffic accidents had occurred in 73.8% of patients. The preoperative median GCS score for our participants was 6 (IQR 3.7–7.5). The median injury severity score (ISS) was 25 (IQR 24–28). Five patients expired during the course of treatment. The mortality rate of our participants was
17.8%. Table 1 shows the detailed demographic and clinical characteristics of survivors and non-survivors. We used the Bland and Altman plotting analysis to evaluate the temperature differences in ICT minus Tr (Fig. 1A), and ICT minus Tt (Fig. 1B). Our analysis revealed a significant difference in temperature gradient between these two sites from postoperative day 1 to postoperative day 4 in survivors. Our paired t-test showed that the mean temperature difference between ICT and Tr (0.23 ± 0.45 °C), and ICT and Tt (0.64 ± 0.6 °C) was significantly different (p < 0.001). The ICT was the highest temperature recorded in survivors (37.6 ± 0.6 °C). In patients who survived, Pearson’s correlation methods showed that ICT was positively correlated (p < 0.001) with Tr (Fig. 2) and Tt (Fig. 3). Fig. 4 shows the daily mean temperature change from postoperative day 1 to postoperative day 4. In survivors, when the ICT was maintained at < 38 °C, the ICT was the highest temperature, followed by Tr and Tt. In total, the frequency of brain hyperthermia was 17.7% in our study. The daily frequency of brain hyperthermia from postoperative day 1 to postoperative day 4 was 21.9%, 18.8%, 15.1%, and 13.2%, respectively. In five patients, brain death occurred at 8 hours, 8 hours, 10 hours, 10 hours, and 12 hours after a temperature gradient reversal in ICT–Tr and ICT–Tt, caused by an uncontrolled intractable increase in ICP and brain swelling (Fig. 5). In the non-survivor shown in Fig. 5, the reverse of the temperature gradient in ICT–Tr and in ICT–Tt started at 65 hours after surgery (Fig. 5A). The reversal of ICP and CPP (Fig. 5B) occurred at the same time as the reversal of ICT and Tr, when the ICT was 34.2 °C, Tr was 35.3 °C, Tt was 35.9 °C, ICP was 58 mmHg, and CPP was 9 mmHg. Fig. 6 shows the incidence of ICT–Tr gradient reversal in survivors and non-survivors from postoperative day 1 to day 4. The mean reversal temperature gradient was 3.0 ± 0.62 °C in non-survivors and 0.2 ± 0.16 °C in survivors (Fig. 7).
4. Discussion Brain hyperthermia is frequently seen in patients following TBI. The causes of hyperthermia may result from post-traumatic cerebral inflammation, direct hypothalamic damage or a secondary infection resulting in fever.3 In our study, the incidence rate of brain hyperthermia was 17%, which is lower than most previous
Table 1 Demographic and clinical characteristics of 28 patients with severe traumatic brain injury
Age, years Sex Male Female Initial GCS score Preoperative GCS score Injury severity score Brain CT midline shift (mm) Mechanism Traffic accident Fall Pupil reaction Both fixed One fixed Both reactive
Survivors (n = 23)
Non-survivors (n = 5)
p-value
39 (24–60)
67 (45.5–77)
1.000
17 6 9 (6–13) 6 (4–8) 25 (18–29) 6.4 (3.6–12.3)
2 3 4 (3–10) 3 (3–6.5) 25 (25–35) 17.3 (12.4–20.1)
0.290
21 2
3 2
0.135
1 8 14
5 0 0
<0.001
0.905 0.730 0.730 0.016
Data represent median (IQR) or count. Continuous variables were compared by the Mann–Whitney U-test, and categorical data by Fisher’s exact test. GCS = Glasgow Coma Scale score, IQR = interquartile range.
J.-R. Kuo et al. / Journal of Clinical Neuroscience 18 (2011) 1059–1063
1061
Fig. 3. A graph of the relationship between mean intracranial temperature (ICT) and mean superficial temporal artery temperature (Tt) on the craniectomy side in survivors using Pearson’s correlation method showing a significant positive correlation between ICT and Tt (p < 0.001).
Fig. 1. Bland–Altman plots comparing the difference between temperature measurements (A) showing a comparison of intracranial (ICT) and rectal temperature (Tr), and (B) a comparison of ICT and superficial temporal artery temperature (Tt).
Fig. 4. A graph from postoperative day 1 to day 4 of the time course of mean intracranial temperature (ICT, open circle), superficial temporal artery temperature (Tt, open square) and rectal temperature (Tr, open triangle) in survivors showing that the ICT was the highest temperature in surviving patients, followed by Tr and Tt.
Fig. 2. A graph of the relationship between mean intracranial temperature (ICT) (on the craniectomy side in survivors) and mean rectal temperature (Tr) using Pearson’s correlation method showing a significant positive correlation between ICT and Tr (p < 0.001).
reports (16–85%).4–6,15 This discrepancy may be due to early intervention in our study to maintain brain temperatures at 38 °C or less. In general, following an acute neurological injury, the human brain is at a higher temperature than other body temperatures, such as Tr.16 Rossi et al. have demonstrated that ICT is significantly higher than core temperature, especially during a fever after a brain injury.15 However, Childs et al. indicated that brain temperature could not be predicted from Tr at all times because brain temperatures are not always higher than core temperatures.12 Fig. 1A shows that in our study, when the brain temperature was controlled at less than 38 °C, the invasive Tr measurement may cause doctors to underestimate ICT (mean gradient of 0.23 °C), despite a strong positive correlation between these two parameters in survivors (Fig. 2). Therefore, our results support Childs’ study by indicating that brain temperature cannot always be predicted from Tr in survivors.12 We also want to emphasize the importance of monitoring brain temperature during the acute post-TBI period
1062
J.-R. Kuo et al. / Journal of Clinical Neuroscience 18 (2011) 1059–1063
Fig. 7. Bar graph showing differences in intracranial temperature (ICT) and rectal temperature (Tr) in survivors (open bar) and non-survivors (striped bar) of traumatic brain injury when the ICT is kept below 38 °C showing greater differences in non-survivors.
Fig. 5. A graph from 0 hours to 96 hours postoperatively of changes in a nonsurvivors in (A) intracranial temperature (ICT, solid circles), rectal temperature (Tr, solid squares), and superficial temporal artery temperature (Tt, solid triangles); and (B) intracranial pressure (ICP, solid circles) and cerebral perfusion pressure (CPP, solid squares) showing a temperature gradient reversal in ICT–Tr and ICT–Tt starting at 65 hours after surgery at an ICT of 34.2 °C, Tr of 35.3 °C, Tt of 35.9 °C, ICP of 58 mmHg, and a CPP of 9 mmHg. (This figure is available in colour at www.sciencedirect.com)
Fig. 6. Bar graph of a comparison from postoperative days 1 to 4 of the percentages of patients whose rectal temperature measurements (Tr) exceeded intracranial temperature (ICT) showing higher percentages in non-survivors of traumatic brain injury (TBI; filled bar) compared to survivors; open bar).
because Tr may cause doctors to underestimate ICT when the brain is vulnerable to secondary injury. Measurement of Tt is a common practice in the ICU because it is non-invasive, does not require contact with mucous membranes, and is not significantly affected by thermoregulatory changes.10
In our study, Fig. 1B shows that when we maintained a brain temperature of less than 38 °C, a significant temperature gradient existed between the ICT and Tt (mean gradient of 0.64 °C), despite a strong positive correlation between these two parameters in survivors (Fig. 3). These results suggest that the daily Tt measurement may cause doctors to underestimate ICT in survivors. Our results also remind us that if a highly accurate ICT measurement is required, neither Tr nor temporal artery thermometers are sufficiently accurate to replace ICT monitoring. Under normal conditions, brain temperature is dependent on heat production during brain parenchymal metabolism17 and regional blood flow and perfusion.8,11 Fountas et al. demonstrated that the ICT in 11 patients became lower than their systemic temperature (median time of 4.43 hours) prior to any changes in ICP and CPP. Fountas et al. concluded that dissociation between the ICT and Tr is an early sign of brain death.11 In our study, five patients developed brain death at 8 hours, 8 hours, 10 hours, 10 hours, and 12 hours after temperature gradient reversal in ICT–Tr and ICT–Tt (Fig. 5). The reversal phenomenon (decreased ICT relative to the systemic temperature) is also observed between ICT and Tt. As Fig. 5 shows, the value of CPP was less than 40 mmHg when the reversal phenomenon occurred postoperatively. We consider that reduced ICT might be due to a concomitant decrease in regional cerebral blood flow and a diminished cerebral metabolism. Fig. 6 shows the incidence of the reversal phenomenon in survivors and non-survivors. The Fisher-exact test revealed a significant difference between survivors and non-survivors (p < 0.01). We also found that the reversal temperature gradient (mean ± SD) in ICT–Tr was 3.0 °C ± 0.62 °C in non-survivors and 0.2 °C ± 0.16 °C in survivors. These results confirm a previous report showing that the reversal of ICT and Tr, and/or the ICT and Tr gradient, may be an early marker of a poor prognosis in patients with a severe head injury.11 However, the small number of patients in our study limited our ability to draw conclusions about the threshold at which the reversal point was likely to affect patient outcome after a TBI. The exact temperature gradient that will lead to death needs to be clarified in future studies. Chio et al.18 successfully demonstrated that selective brain cooling (33–35 °C) induced by an infusion of 4 °C normal saline via the external jugular vein is associated with a small decrease in Trattenuated cell ischemia and damage in rats. Forte et al.19 also suggested that regional brain cooling (mean of 35.2 °C) was effective in controlling the ICP in patients who had previously undergone a decompressive craniectomy. However, based on our findings that reversal of the ICT and Tr could be an early marker of poor prognoses in patients after a TBI, our data do not support the widely held view that a low brain temperature is beneficial for patients with TBI. The optimal brain temperature and methods to monitor the brain condition need to be clarified if selective brain cooling is to be performed.
J.-R. Kuo et al. / Journal of Clinical Neuroscience 18 (2011) 1059–1063
The temperature gradient from the cortex to the central brain ranges up to 0.9 °C.20 In any study that evaluates brain temperature, the accuracy of brain temperature measurement and the position of the implanted probe must be considered. The measurement methods, calibration of the brain temperature sensors and the monitoring technique used in our study have been reported previously.16 In our study the ICP/ICT measurement procedure was standardized by use of an intraparenchymal catheter coupled with a 3-cm-deep, frontal thermistor inserted into the parenchyma of the injured hemisphere. Our study had several limitations. First, the study was conducted retrospectively and only a small number of patients were enrolled, limiting our ability to draw conclusions. Second, the results were mostly based on the brain at normothermia. The relationship between brain temperature and Tt in hyperthermia was not evaluated. Finally, the study lacked a control group for other factors that may have influenced the results. 5. Conclusions We conclude that a significant temperature gradient exists between ICT and Tr and between ICT and Tt. The daily practice of non-invasive Tt measurement may cause doctors to underestimate ICT. Reversal of the ICT and Tt and/or Tr could be an early marker of patients who have a poor prognosis. Acknowledgments The authors wish to express their gratitude to all of the participants from neurology, neurosurgery, emergency critical medicine, and the intensive care unit. The authors also thank Ms Lin WenChun for her help with the statistical analysis. References 1. Bullock MR, Chesnut R, Ghajar J, et al. Guidelines for the surgical management of traumatic brain injury. Neurosurgery 2006;58(Suppl. 3). S2–S6.
1063
2. Stocchetti N, Rossi S, Zanier ER, et al. Pyrexia in head injured patients admitted to intensive care. Intensive Care Med 2002;28:1555–62. 3. Thompson HJ, Tkacs NC, Saatman KE, et al. Hyperthermia following traumatic brain injury: a critical evaluation. Neurol Dis 2003;12:163–73. 4. Albrecht R, Wass CT, Lanier WL. Occurrence of potentially detrimental temperature alterations in hospitalized patients at risk for brain injury. Mayo Clin Proc 1998;73:629–35. 5. Kilpatrick M, Lowry D, Firlik A, et al. Hyperthermia in the neurosurgical intensive care unit. Neurosurgery 2000;47:850–5. 6. Schwarz S, Hafner K, Aschoff A, et al. Incidence and prognostic significance of fever following intracerebral hemorrhage. Neurology 2000;54:354–61. 7. Mcilvoy L. The impact of brain temperature and core temperature on intracranial pressure and cerebral perfusion pressure. J Neurosci Nurs 2007;39:324–31. 8. Rumana C, Gopinath S, Uzura M, et al. Brain temperature exceeds systemic temperature in head injured patients. Crit Care Med 1998;26:562–7. 9. Olson DM, Graffagnino C. Managing temperature in brain-injured patients. Nursing 2005;35. 32cc1–12. 10. Hebbar K, Fortenberry JD, Roger K, et al. Comparison of temporal artery thermometer to standard temperature measurements in pediatric intensive care unit patient. Pediatr Crit Care Med 2005;6:557–61. 11. Fountas KN, Kapsalaki EZ, Feltes CH, et al. Disassociation between intracranial and systemic temperature as an early sign of brain death. J Neurosurg Anesthesiol 2003;15:87–9. 12. Childs C, Vail A, Protheroe R, et al. Differences between brain and rectal temperature during routine critical care of patients with severe traumatic brain injury. Anesthesia 2005;60:759–65. 13. Spiotta AM, Heuer GG, Bloom S, et al. Brain hyperthermia after traumatic brain injury does not reduce brain oxygen. Neurosurgery 2008;62:864–72. 14. Bland J, Altman D. Statistics methods of measurement: why plotting difference against standard method is misleading. Lancet 1995;346:1085–7. 15. Rossi S, Zanier ER, Mauri I, et al. Brain temperature, body temperature, and intracranial pressure in acute cerebral damage. J Neurol Neurosurg Psychiatry 2001;71:448–54. 16. Mcilvoy L. Comparison of brain temperature to core temperature: a review of the literature. J Neurosci Nurs 2004;36:23–31. 17. Fitch W. Brain metabolism. In: Cottrell JE, Smith DS, editors. Anesthesia and neurosurgery. St. Louis: Mosby; 1994. p. 1–16. 18. Chio CC, Kuo JR, Hsiao SH, et al. Effect of brain cooling on brain ischemia and damage markers after fluid percussion brain injury in rats. Shock 2007;28:284–90. 19. Forte LV, Peluso CM, Prandini MN, et al. Regional cooling for reducing brain temperature and intracranial pressure. Arq Neuropsoquiat 2009;67:480–7. 20. Mellergard P. Intracerebral temperature in neurosurgical patients: intracerebral temperature gradient and relationships to consciousness level. Surg Neurol 1995;43:91–5.
Contents lists available at ScienceDirect
Journal of Clinical Neuroscience journal homepage: www.elsevier.com/locate/jocn
Clinical Study
Measuring brain temperature while maintaining brain normothermia in patients with severe traumatic brain injury Jinn-Rung Kuo a,b,c, Chong-Jeh Lo a,d, Che-Chuan Wang c, Chin-Li Lu e, Shu-Chin Lin f, Chiao-Fang Chen f,⇑ a
Institute of Clinical Medicine, School of Medicine, National Cheng-Kung University, Tainan, Taiwan Department of Biotechnology, Southern Taiwan University, Tainan, Taiwan c Department of Neurosurgery, Chi-Mei Medical Center, Tainan, Taiwan d Division of Trauma and Emergency Surgery, Chang Gung Memorial Hospital–Kaohsiung Medical Center, Kaohsiung, Taiwan e Institute of Medical Research, Chi-Mei Medical Center, Tainan, Taiwan f Advanced Practices Nursing, Department of Nursing, Chi-Mei Medical Center, 901 Chung Hwa Road, Yung Kang City, Tainan, Taiwan b
a r t i c l e
i n f o
Article history: Received 29 July 2010 Accepted 2 November 2010
Keywords: Brain normothermia Intracranial temperature Rectal temperature Superficial temporal artery temperature
a b s t r a c t The aim of this study was to evaluate the relationship between superficial temporal artery temperature (Tt), rectal temperature (Tr) and intracranial temperature (ICT) when attempting to keep the brain in a normothermic condition in patients with severe traumatic brain injury (TBI). We also compared the incidence of temperature gradient reversal in patients who survived (survivors) and patients who did not (non-survivors) and the difference in temperature gradient reversal between survivors and non-survivors. Tr is normally lower than and ICT and temperature gradient reversal, when Tr exceeds ICT, has been demonstrated to be an early sign of brain death. A total of 28 patients with severe TBI were enrolled retrospectively in our study between November 2008 and February 2010. Each patient’s Tt, Tr and ICT was recorded every hour for 4 days. Our results show that the frequency of brain hyperthermia in our participants (ICT > 38 °C) was 17.7%. Using a paired t-test and Bland–Altman plots, it was shown that a significant temperature difference existed between Tt, Tr and ICT (p < 0.001). The use of Spearman’s correlation method revealed that Tt, Tr and ICT were positively correlated (p < 0.001). Brain death occurred in five patients at a mean of 9.6 hours (range: 8–12 hours) after a temperature gradient reversal between Tt, Tr and ICT. Fisher’s exact test showed that there was a significant difference in the incidence of temperature gradient reversal between Tt, Tr and ICT in survivors and non-survivors (p < 0.001). We conclude that a significant temperature difference exists between Tt, Tr and ICT when maintaining brain normothermia. The daily practice of non-invasive Tt measurement may cause doctors to underestimate ICT; reversal of the ICT and Tt and/or Tr temperatures could be an early marker of a poor prognosis for patients with severe TBI. Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction Traumatic brain injury (TBI) affects up to 2% of the population per year and is a major cause of disability and death among young people.1 Hyperthermia is frequently seen in patients during the acute injury period following a TBI, and the incidence and duration of the fever are significantly associated with neuronal damage.2,3 Previous studies have shown that the incidence of hyperthermic rates is 16% within 24 hours post-admission, 31.7% between 24 hours and 48 hours post-admission, 42% at 72 hours post-admission, and 60% to 70% between 48 hours and 96 hours post-admission.4–6 The difference between ICT and Tr has been evaluated via invasive temperature monitoring, and brain temperature is 0.5 °C to 1 °C higher than rectal temperature.7–9 The measurement of Tt, a non-invasive procedure that uses technology commonly found in ⇑ Corresponding author. Tel.: +886 6 2812811; fax: +886 6 2828928. E-mail address: [email protected] (C.-F. Chen). 0967-5868/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.jocn.2010.11.014
intensive care units (ICU), offers potential benefits in critically ill patients.10 Until now, a relationship between Tr, Tt and ICT has not been well established in patients with TBI. Dissociation, or reversal, between ICT and systemic temperatures, defined as when Tr exceeds ICT, has been demonstrated to be an early sign of brain death.8,11 However, the incidences and differences in temperature gradient reversals in patients who survive TBI (survivors) and patients who do not (non-survivors) have not been well evaluated. The aim of this study is to determine the relationship between Tr, Tt and ICT when attempting to keep the brain in a normothermic condition in patients in the ICU with a TBI. We also compared the incidences and differences in temperature gradient reversals in survivors and non-survivors. 2. Material and methods A total of 28 patients diagnosed with a TBI in a medical center in southern Taiwan from November 2008 to February 2010 were
1060
J.-R. Kuo et al. / Journal of Clinical Neuroscience 18 (2011) 1059–1063
evaluated retrospectively and enrolled in our study. This study was approved by the ethics committee of our hospital. A complete history was obtained, and a neurological assessment performed on each patient. All patients underwent a CT scan and were evaluated based on the initial Glasgow Coma Scale (GCS) score in the emergency room. All patients underwent a craniectomy for hematoma removal. The ipsilateral superficial temporal artery was preserved when the craniectomy was performed. An intraparenchymal catheter coupled with a thermistor (110-4BT, Pressure–Temperature Monitoring Kit, Integra Camino, San Diego, CA, USA) was implanted 3-cm deep, frontally, in the parenchyma of the injured hemisphere to standardize monitoring of intracranial pressure (ICP) and ICT. The methods and calibration of the brain temperature sensors and the monitoring techniques used in this study have been reported previously.12 The temperature data are precise to ± 3.0 °C within the range of 20 °C to 40 °C (according to the manufacturer). Postoperatively, patients underwent cerebral perfusion pressure (CPP)-guided management to maintain a CPP of P 60 mmHg and an ICP of 6 25 mmHg. The mean arterial blood pressure (MAP) and CPP were monitored using standard pressure transducers. All patients were sedated using propofol (0.5–6 mg/kg per hour). We also used atracurium (0.3–1.2 mg/kg per hour) if the patient was shivering during the postoperative period. All temperature measurements were recorded in degrees Celsius (°C) every hour for 4 days. The room temperature in the ICU was set at 25 °C. Tr was monitored with a rectal temperature thermistor (Medi-Therm II Hyper–Hypothermia Machine; Gaymar Industries, Orchard Park, NY, USA), and Tt was obtained with a temporal scanner (TemporalScanner™ TAT-5000, EXERGEN, Boston, MA, USA). Tt measurements were performed according to the following ICU standard temperature measurement protocol. The patient’s hair was brushed aside, if necessary, and the probe was positioned on the center of the patient’s forehead. The measurement button was pressed, and the measurement was performed by scanning the region from the center midline and sliding the probe across the head to the region behind the earlobe. The button was then released, and the temperature was recorded. The duration of this procedure was about 5 s. The Tt was evaluated on the side of the head where the operation was performed (defined as the craniectomy side). Brain hyperthermia,13 defined as an ICT > 38 °C, was controlled with an ice pillow (if ICT > 37.5 °C), the oral administration of 500 mg of acetaminophen (if ICT > 38 °C), or with an ice blanket device (if acetaminophen failed) (Gaymar Medi-Therm II Hyper– Hypothermia Machine). The correlation between Tr, Tt and ICT was evaluated using Pearson’s correlation method. The temperature gradient of ICT versus (vs.) Tr, and ICT vs. Tt at each point was calculated and plotted as described by Bland and Altman.14 Data were expressed as the median (interquartile range, IQR) or mean ± standard deviation (SD). All data were analyzed using statistical software (Statistical Package for the Social Sciences for Windows, version 16; SPSS, Chicago, IL, USA). A p value of < 0.05 was considered statistically significant.
3. Results In this study, 28 patients with TBI were evaluated retrospectively (19 male and 9 female; age range, 16–80 years; median age of 44.5 years; age IQR of 28–77.5 years). Traffic accidents had occurred in 73.8% of patients. The preoperative median GCS score for our participants was 6 (IQR 3.7–7.5). The median injury severity score (ISS) was 25 (IQR 24–28). Five patients expired during the course of treatment. The mortality rate of our participants was
17.8%. Table 1 shows the detailed demographic and clinical characteristics of survivors and non-survivors. We used the Bland and Altman plotting analysis to evaluate the temperature differences in ICT minus Tr (Fig. 1A), and ICT minus Tt (Fig. 1B). Our analysis revealed a significant difference in temperature gradient between these two sites from postoperative day 1 to postoperative day 4 in survivors. Our paired t-test showed that the mean temperature difference between ICT and Tr (0.23 ± 0.45 °C), and ICT and Tt (0.64 ± 0.6 °C) was significantly different (p < 0.001). The ICT was the highest temperature recorded in survivors (37.6 ± 0.6 °C). In patients who survived, Pearson’s correlation methods showed that ICT was positively correlated (p < 0.001) with Tr (Fig. 2) and Tt (Fig. 3). Fig. 4 shows the daily mean temperature change from postoperative day 1 to postoperative day 4. In survivors, when the ICT was maintained at < 38 °C, the ICT was the highest temperature, followed by Tr and Tt. In total, the frequency of brain hyperthermia was 17.7% in our study. The daily frequency of brain hyperthermia from postoperative day 1 to postoperative day 4 was 21.9%, 18.8%, 15.1%, and 13.2%, respectively. In five patients, brain death occurred at 8 hours, 8 hours, 10 hours, 10 hours, and 12 hours after a temperature gradient reversal in ICT–Tr and ICT–Tt, caused by an uncontrolled intractable increase in ICP and brain swelling (Fig. 5). In the non-survivor shown in Fig. 5, the reverse of the temperature gradient in ICT–Tr and in ICT–Tt started at 65 hours after surgery (Fig. 5A). The reversal of ICP and CPP (Fig. 5B) occurred at the same time as the reversal of ICT and Tr, when the ICT was 34.2 °C, Tr was 35.3 °C, Tt was 35.9 °C, ICP was 58 mmHg, and CPP was 9 mmHg. Fig. 6 shows the incidence of ICT–Tr gradient reversal in survivors and non-survivors from postoperative day 1 to day 4. The mean reversal temperature gradient was 3.0 ± 0.62 °C in non-survivors and 0.2 ± 0.16 °C in survivors (Fig. 7).
4. Discussion Brain hyperthermia is frequently seen in patients following TBI. The causes of hyperthermia may result from post-traumatic cerebral inflammation, direct hypothalamic damage or a secondary infection resulting in fever.3 In our study, the incidence rate of brain hyperthermia was 17%, which is lower than most previous
Table 1 Demographic and clinical characteristics of 28 patients with severe traumatic brain injury
Age, years Sex Male Female Initial GCS score Preoperative GCS score Injury severity score Brain CT midline shift (mm) Mechanism Traffic accident Fall Pupil reaction Both fixed One fixed Both reactive
Survivors (n = 23)
Non-survivors (n = 5)
p-value
39 (24–60)
67 (45.5–77)
1.000
17 6 9 (6–13) 6 (4–8) 25 (18–29) 6.4 (3.6–12.3)
2 3 4 (3–10) 3 (3–6.5) 25 (25–35) 17.3 (12.4–20.1)
0.290
21 2
3 2
0.135
1 8 14
5 0 0
<0.001
0.905 0.730 0.730 0.016
Data represent median (IQR) or count. Continuous variables were compared by the Mann–Whitney U-test, and categorical data by Fisher’s exact test. GCS = Glasgow Coma Scale score, IQR = interquartile range.
J.-R. Kuo et al. / Journal of Clinical Neuroscience 18 (2011) 1059–1063
1061
Fig. 3. A graph of the relationship between mean intracranial temperature (ICT) and mean superficial temporal artery temperature (Tt) on the craniectomy side in survivors using Pearson’s correlation method showing a significant positive correlation between ICT and Tt (p < 0.001).
Fig. 1. Bland–Altman plots comparing the difference between temperature measurements (A) showing a comparison of intracranial (ICT) and rectal temperature (Tr), and (B) a comparison of ICT and superficial temporal artery temperature (Tt).
Fig. 4. A graph from postoperative day 1 to day 4 of the time course of mean intracranial temperature (ICT, open circle), superficial temporal artery temperature (Tt, open square) and rectal temperature (Tr, open triangle) in survivors showing that the ICT was the highest temperature in surviving patients, followed by Tr and Tt.
Fig. 2. A graph of the relationship between mean intracranial temperature (ICT) (on the craniectomy side in survivors) and mean rectal temperature (Tr) using Pearson’s correlation method showing a significant positive correlation between ICT and Tr (p < 0.001).
reports (16–85%).4–6,15 This discrepancy may be due to early intervention in our study to maintain brain temperatures at 38 °C or less. In general, following an acute neurological injury, the human brain is at a higher temperature than other body temperatures, such as Tr.16 Rossi et al. have demonstrated that ICT is significantly higher than core temperature, especially during a fever after a brain injury.15 However, Childs et al. indicated that brain temperature could not be predicted from Tr at all times because brain temperatures are not always higher than core temperatures.12 Fig. 1A shows that in our study, when the brain temperature was controlled at less than 38 °C, the invasive Tr measurement may cause doctors to underestimate ICT (mean gradient of 0.23 °C), despite a strong positive correlation between these two parameters in survivors (Fig. 2). Therefore, our results support Childs’ study by indicating that brain temperature cannot always be predicted from Tr in survivors.12 We also want to emphasize the importance of monitoring brain temperature during the acute post-TBI period
1062
J.-R. Kuo et al. / Journal of Clinical Neuroscience 18 (2011) 1059–1063
Fig. 7. Bar graph showing differences in intracranial temperature (ICT) and rectal temperature (Tr) in survivors (open bar) and non-survivors (striped bar) of traumatic brain injury when the ICT is kept below 38 °C showing greater differences in non-survivors.
Fig. 5. A graph from 0 hours to 96 hours postoperatively of changes in a nonsurvivors in (A) intracranial temperature (ICT, solid circles), rectal temperature (Tr, solid squares), and superficial temporal artery temperature (Tt, solid triangles); and (B) intracranial pressure (ICP, solid circles) and cerebral perfusion pressure (CPP, solid squares) showing a temperature gradient reversal in ICT–Tr and ICT–Tt starting at 65 hours after surgery at an ICT of 34.2 °C, Tr of 35.3 °C, Tt of 35.9 °C, ICP of 58 mmHg, and a CPP of 9 mmHg. (This figure is available in colour at www.sciencedirect.com)
Fig. 6. Bar graph of a comparison from postoperative days 1 to 4 of the percentages of patients whose rectal temperature measurements (Tr) exceeded intracranial temperature (ICT) showing higher percentages in non-survivors of traumatic brain injury (TBI; filled bar) compared to survivors; open bar).
because Tr may cause doctors to underestimate ICT when the brain is vulnerable to secondary injury. Measurement of Tt is a common practice in the ICU because it is non-invasive, does not require contact with mucous membranes, and is not significantly affected by thermoregulatory changes.10
In our study, Fig. 1B shows that when we maintained a brain temperature of less than 38 °C, a significant temperature gradient existed between the ICT and Tt (mean gradient of 0.64 °C), despite a strong positive correlation between these two parameters in survivors (Fig. 3). These results suggest that the daily Tt measurement may cause doctors to underestimate ICT in survivors. Our results also remind us that if a highly accurate ICT measurement is required, neither Tr nor temporal artery thermometers are sufficiently accurate to replace ICT monitoring. Under normal conditions, brain temperature is dependent on heat production during brain parenchymal metabolism17 and regional blood flow and perfusion.8,11 Fountas et al. demonstrated that the ICT in 11 patients became lower than their systemic temperature (median time of 4.43 hours) prior to any changes in ICP and CPP. Fountas et al. concluded that dissociation between the ICT and Tr is an early sign of brain death.11 In our study, five patients developed brain death at 8 hours, 8 hours, 10 hours, 10 hours, and 12 hours after temperature gradient reversal in ICT–Tr and ICT–Tt (Fig. 5). The reversal phenomenon (decreased ICT relative to the systemic temperature) is also observed between ICT and Tt. As Fig. 5 shows, the value of CPP was less than 40 mmHg when the reversal phenomenon occurred postoperatively. We consider that reduced ICT might be due to a concomitant decrease in regional cerebral blood flow and a diminished cerebral metabolism. Fig. 6 shows the incidence of the reversal phenomenon in survivors and non-survivors. The Fisher-exact test revealed a significant difference between survivors and non-survivors (p < 0.01). We also found that the reversal temperature gradient (mean ± SD) in ICT–Tr was 3.0 °C ± 0.62 °C in non-survivors and 0.2 °C ± 0.16 °C in survivors. These results confirm a previous report showing that the reversal of ICT and Tr, and/or the ICT and Tr gradient, may be an early marker of a poor prognosis in patients with a severe head injury.11 However, the small number of patients in our study limited our ability to draw conclusions about the threshold at which the reversal point was likely to affect patient outcome after a TBI. The exact temperature gradient that will lead to death needs to be clarified in future studies. Chio et al.18 successfully demonstrated that selective brain cooling (33–35 °C) induced by an infusion of 4 °C normal saline via the external jugular vein is associated with a small decrease in Trattenuated cell ischemia and damage in rats. Forte et al.19 also suggested that regional brain cooling (mean of 35.2 °C) was effective in controlling the ICP in patients who had previously undergone a decompressive craniectomy. However, based on our findings that reversal of the ICT and Tr could be an early marker of poor prognoses in patients after a TBI, our data do not support the widely held view that a low brain temperature is beneficial for patients with TBI. The optimal brain temperature and methods to monitor the brain condition need to be clarified if selective brain cooling is to be performed.
J.-R. Kuo et al. / Journal of Clinical Neuroscience 18 (2011) 1059–1063
The temperature gradient from the cortex to the central brain ranges up to 0.9 °C.20 In any study that evaluates brain temperature, the accuracy of brain temperature measurement and the position of the implanted probe must be considered. The measurement methods, calibration of the brain temperature sensors and the monitoring technique used in our study have been reported previously.16 In our study the ICP/ICT measurement procedure was standardized by use of an intraparenchymal catheter coupled with a 3-cm-deep, frontal thermistor inserted into the parenchyma of the injured hemisphere. Our study had several limitations. First, the study was conducted retrospectively and only a small number of patients were enrolled, limiting our ability to draw conclusions. Second, the results were mostly based on the brain at normothermia. The relationship between brain temperature and Tt in hyperthermia was not evaluated. Finally, the study lacked a control group for other factors that may have influenced the results. 5. Conclusions We conclude that a significant temperature gradient exists between ICT and Tr and between ICT and Tt. The daily practice of non-invasive Tt measurement may cause doctors to underestimate ICT. Reversal of the ICT and Tt and/or Tr could be an early marker of patients who have a poor prognosis. Acknowledgments The authors wish to express their gratitude to all of the participants from neurology, neurosurgery, emergency critical medicine, and the intensive care unit. The authors also thank Ms Lin WenChun for her help with the statistical analysis. References 1. Bullock MR, Chesnut R, Ghajar J, et al. Guidelines for the surgical management of traumatic brain injury. Neurosurgery 2006;58(Suppl. 3). S2–S6.
1063
2. Stocchetti N, Rossi S, Zanier ER, et al. Pyrexia in head injured patients admitted to intensive care. Intensive Care Med 2002;28:1555–62. 3. Thompson HJ, Tkacs NC, Saatman KE, et al. Hyperthermia following traumatic brain injury: a critical evaluation. Neurol Dis 2003;12:163–73. 4. Albrecht R, Wass CT, Lanier WL. Occurrence of potentially detrimental temperature alterations in hospitalized patients at risk for brain injury. Mayo Clin Proc 1998;73:629–35. 5. Kilpatrick M, Lowry D, Firlik A, et al. Hyperthermia in the neurosurgical intensive care unit. Neurosurgery 2000;47:850–5. 6. Schwarz S, Hafner K, Aschoff A, et al. Incidence and prognostic significance of fever following intracerebral hemorrhage. Neurology 2000;54:354–61. 7. Mcilvoy L. The impact of brain temperature and core temperature on intracranial pressure and cerebral perfusion pressure. J Neurosci Nurs 2007;39:324–31. 8. Rumana C, Gopinath S, Uzura M, et al. Brain temperature exceeds systemic temperature in head injured patients. Crit Care Med 1998;26:562–7. 9. Olson DM, Graffagnino C. Managing temperature in brain-injured patients. Nursing 2005;35. 32cc1–12. 10. Hebbar K, Fortenberry JD, Roger K, et al. Comparison of temporal artery thermometer to standard temperature measurements in pediatric intensive care unit patient. Pediatr Crit Care Med 2005;6:557–61. 11. Fountas KN, Kapsalaki EZ, Feltes CH, et al. Disassociation between intracranial and systemic temperature as an early sign of brain death. J Neurosurg Anesthesiol 2003;15:87–9. 12. Childs C, Vail A, Protheroe R, et al. Differences between brain and rectal temperature during routine critical care of patients with severe traumatic brain injury. Anesthesia 2005;60:759–65. 13. Spiotta AM, Heuer GG, Bloom S, et al. Brain hyperthermia after traumatic brain injury does not reduce brain oxygen. Neurosurgery 2008;62:864–72. 14. Bland J, Altman D. Statistics methods of measurement: why plotting difference against standard method is misleading. Lancet 1995;346:1085–7. 15. Rossi S, Zanier ER, Mauri I, et al. Brain temperature, body temperature, and intracranial pressure in acute cerebral damage. J Neurol Neurosurg Psychiatry 2001;71:448–54. 16. Mcilvoy L. Comparison of brain temperature to core temperature: a review of the literature. J Neurosci Nurs 2004;36:23–31. 17. Fitch W. Brain metabolism. In: Cottrell JE, Smith DS, editors. Anesthesia and neurosurgery. St. Louis: Mosby; 1994. p. 1–16. 18. Chio CC, Kuo JR, Hsiao SH, et al. Effect of brain cooling on brain ischemia and damage markers after fluid percussion brain injury in rats. Shock 2007;28:284–90. 19. Forte LV, Peluso CM, Prandini MN, et al. Regional cooling for reducing brain temperature and intracranial pressure. Arq Neuropsoquiat 2009;67:480–7. 20. Mellergard P. Intracerebral temperature in neurosurgical patients: intracerebral temperature gradient and relationships to consciousness level. Surg Neurol 1995;43:91–5.