Cardiac surgery with the use of cardiopulmonary bypass (CPB) is associated with thermal pertubations during operation as well as during the early postoperative period [1,2]. Body temperature impairment following normothermic and hypothermic CPB is a direct consequence of alterations in tissue heat content [3,4]. One major reason is the inflammatory response to CPB and surgical trauma, which elevates the central hypothalamic temperature setpoint. This elevation in temperature setpoint (fever) occurs normally after surgery and is reached by clinically relevant thermoregulatory mechanisms , e.g. vasoconstriction of dermal vessels (to reduce body heat loss) and shivering (to increase body heat production) [6,7]. Vasoconstriction and reduction of dermal blood flow can be estimated by either a skin surface temperature gradient (forearm temperature higher than fingertip temperature) [8,9] or by the peripheral perfusion index (PI) derived from a pulse oximetry signal [10,11]. Vasoconstriction and shivering are unfavourable in cardiac surgical patients because both increase morbidity, delay extubation and prolong the length of intensive care unit (ICU) stay . Early extubation after cardiac surgery is a main step for rapid postoperative recovery and may reduce the need for conventional intensive care [13-15]. Extubation criteria include pulmonary, haemodynamic, neurological assessments and demand a minimum value of core temperature. In a previous study we demonstrated that in patients following normothermic CPB extubation becomes most likely in the transition period from peripheral vasoconstriction to vasodilatation . The aim of this study was to test the hypothesis that changes in core and skin surface temperature are related to extubation time in patients following normothermic and hypothermic CPB.
Forty-six consecutive patients who were scheduled to undergo elective cardiac surgery with the use of CPB were prospectively studied. Written informed consent was obtained in all patients. Patients with a left ventricular ejection fraction <50%, age >80 yr, body mass index (BMI) >30 kg m−2, chronic obstructive pulmonary disease, neurological dysfunction, significant arrhythmias, thyroid disease or Raynaud's syndrome were excluded.
The established institutional protocol for fast track cardiac anaesthesia was used: Patients were premedicated with midazolam 1 h before operation. Anaesthesia was induced with midazolam (2–5 mg), remifentanil (0.1–0.5 μg kg−1 min−1) and propofol (2 mg kg−1). Cisatracurium (0.2 mg kg−1) was given prior to endotracheal intubation for neuromuscular blockade. After intubation all patients were normoventilated with oxygen in air (fractional concentration of oxygen: 0.3–0.5). Anaesthesia was maintained with remifentanil (0.1–0.4 μg kg−1 min−1) and propofol (5–8 mg kg−1 h−1). Catheters were placed in a radial artery and an internal jugular vein. Patients were selected for normothermic CPB (nasopharyngeal temperature >35.5°C) or hypothermic CPB (lowest nasopharyngeal temperature 28–34°C) at the discretion of the cardiac surgeons.
After surgery, the intubated and sedated patients were immediately transferred to the ICU. There, remifentanil was administered for analgesia and propofol for sedation to achieve a Ramsay sedation score of 2–3. Shivering was treated with meperidine. Dobutamine was administered for inotropic support, hypotension was treated with norepinephrine, hypertension was treated with nitroglycerine and tachycardia was controlled with esmolol. Special attention was paid to guarantee adequate hydration. In the operating room and in the ICU, ambient temperature was maintained at 19°C and 22°C, respectively. Active warming was performed routinely in our institution until a core temperature of 38°C was reached (forced-air cutaneous cover blankets powered by Bair Hugger heater blower units, Augustine Medical, Eden Prairie, MN, USA).
Skin surface temperature was measured in the ICU with one thermocouple located on the inside of the proximal forearm. A second one was fixed to the distal phalanx of the middle finger of the same arm. These thermocouples were at a sufficient distance from any heat-producing equipment. Furthermore, no intravenous infusions were given on this arm. The skin surface temperature gradient was calculated as forearm temperature minus fingertip temperature. Skin surface temperature gradients >0°C are defined as vasoconstriction and skin surface temperature gradients ≤0°C are defined as vasodilatation [8,9]. Urinary bladder, forearm and fingertip temperatures were measured, and the PI was calculated at least every 30 min for 10 h postoperatively (starting from arrival in the ICU) . Every 30 min an anaesthesiologist, blinded to temperature measurements, determined the eligibility for extubation according to criteria adapted from Wong and colleagues : (1) patient responsive and co-operative, (2) spontaneous respiratory rate between 8 and 25 breaths min−1; FiO2 ≥ 0.5; PaCO2 ≤ 6 kPa; PaO2 ≥ 10.7 kPa, (3) haemodynamic stability: absence of uncontrolled arrhythmia; cardiac index > 2L min−1 m−2; mean arterial pressure ≥ 60 mmHg; epinephrine or norepinephrine ≤ 0.25 μg kg−1 min−1; lactate < 4 mmol L−1; pH > 7.30, (4) chest tube drainage < 100 mL h−1 within 2 h and (5) urinary bladder temperature >36.0°C .
Calculation of PI
The PI estimates the fingertip blood flow using a fingertip pulse oximetry (SPO2 ) telemetry monitoring system (Philips Medical Systems, Böblingen, Germany). To obtain a value that is independent from saturation changes, the PI is calculated with a patented method using both red and infrared pulsations: PI = k1 * (ACred DCred−1) + k2 * (ACinfrared DCinfrared−1) with k1 and k2 being wavelength-dependent factors; ACred: the alternating component of the signal based on the red light; DCred: the constant component of the signal based on the red light; ACinfrared: the alternating component of the signal based on the infrared light, DCinfrared: the constant component of the signal based on the infrared light. Fingertip blood flow assessed by the skin surface temperature gradient is known to correlate highly with the PI (a PI of ≥ 2 is defined as vasodilatation) [8,9].
All continuous variables are given as mean ± standard deviation and were compared using the t -test or Wilcoxon's non-parametric test, as appropriate. Nominal variables were compared using contingency tables in combination with a χ2-test. Comparisons between extubation time and time to skin surface temperature gradient =0 were made using the Spearman's rank correlation coefficient. Statistical significance was assumed if the null hypothesis could be rejected at the 0.05 probability level. Software: Statistical Analysis System, SAS® Version 9.3.1, Cary, NC, USA.
Six (1 normothermic and 5 hypothermic) of the 46 patients enrolled in the study were excluded due to postoperative complications: early extubation was not possible because of bleeding resulting in reoperation in 3 patients, haemodynamic instability with significant use of catecholamines (>0.25 μg kg1 min−1) in 2 patients and extended pneumothorax in 1 patient. Thus, 40 patients aged 62 ± 14 yr (range 27–79 yr) were included in the study. Normothermic CPB was conducted in 28 patients and hypothermic CPB in 12 patients. No differences were found between groups regarding age, gender, BMI, CPB time and overall time for surgery. In contrast, there were significant differences in type of operation, aortic clamp time and lowest nasopharyngeal temperature during operation between the two groups (Table 1).
On arrival in the ICU, the mean urinary bladder temperature was 36.8 ± 0.5°C in the normothermic group and 36.4 ± 0.3°C in the hypothermic group (P = 0.014). No patient arrived in the ICU with a urinary bladder temperature below 36°C. The skin surface temperature gradient indicated severe vasoconstriction in both groups (Table 1). No significant differences were found in PI at ICU arrival between the groups (Table 1). However, the change from vasoconstriction to vasodilatation of dermal vessels (skin surface temperature gradient =0°C) was significantly faster following normothermic CPB than after hypothermic CPB (Table 1, Fig. 1). The time to reach a PI ≥ 2 was almost identical to the time to skin surface temperature gradient =0°C in both groups (Table 1). The mean extubation time differed significantly between the two groups (Table 1). There was a statistically significant linear relation between time to temperature gradient =0°C and extubation time for all study patients, or patients in the two groups (Fig. 2). There was no significant difference in the time needed to reach the maximum of bladder temperature between the two groups (Table 1). The maximum value of urinary bladder temperature did not differ between the two groups (Table 1, Fig. 3).
Shivering (treated with meperidine) was observed in 25 patients (18 normothermic group, 7 hypothermic group). Dobutamine was necessary in 4 patients (3 normothermic group, 1 hypothermic group). Hypotension was treated with norepinephrine in 14 patients (10 normothermic group, 4 hypothermic group). Hypertension was treated with nitroglycerine in 3 patients (2 normothermic group, 1 hypothermic group). Tachycardia had to be controlled with esmolol in 1 normothermic patient. Extubation was successful in all patients. None of these patients had to be re-intubated or suffered from complications during the early postoperative period.
Our data demonstrate a significant association between postoperative temperature changes and the eligibility for extubation in patients following normothermic and hypothermic CPB.
All patients arrived in the ICU with a urinary bladder temperature greater than 36°C. Although the bladder temperature was significantly higher in the normothermic group than in the hypothermic group, the mean skin surface temperature gradients indicated similar degrees of severe vasoconstriction in the two groups. This implies that patients following normothermic and hypothermic CPB are subject to major alterations in body temperature during the early postoperative period . A major reason is the systemic inflammatory response after CPB and surgical trauma, leading to an up-regulation of the hypothalamic temperature setpoint [5,6]. Thermoregulatory mechanisms (vasoconstriction and shivering) lead to an increase in body temperature. With increasing body temperature peripheral vasoconstriction became less severe. We found that postoperative thermal normalization (skin surface temperature gradient ≤0 and PI ≥ 2) was reached earlier in patients operated with normothermic CPB as compared to patients operated with hypothermic CPB.
There is general consensus that vasodilatation leads to an internal redistribution of body heat with heat dissipating from the central thermal compartment to the peripheral tissues . The change from peripheral vasoconstriction to vasodilatation indicates that core temperature has reached its maximum. This thermal redistribution results in a decrease in core temperature and an increase of peripheral tissue temperature. In general, skin surface temperatures are considerably lower than core temperature. Skin surface temperatures, when adjusted with an appropriate offset, nonetheless, reflect core temperature reasonably well . It is generally accepted that extubation should not be performed during shivering or at a core temperature lower than 36.0°C [13,18,20,21].
We observed that this change from vasoconstriction to vasodilatation coincided with eligibility for extubation. Our predefined extubation criteria were not met prior to this thermoregulatory change. It is plausible that when the hypothalamic temperature setpoint was reached, heat production and oxygen consumption decreased. Thus, patients became eligible for extubation. In fact, there must have been a higher intraoperative heat loss in patients operated with hypothermic CPB leading to a slower return into thermal balance (Fig. 3). This obvious difference in speed of thermal normalization did not change the relationship between skin surface temperature gradient and extubation time in patients operated with hypothermic CPB.
The shift from vasoconstriction to vasodilatation was not accompanied by an immediate decrease in bladder temperature . It is reasonable to hypothesize that bladder temperature lagged behind the hypothalamic setpoint temperature . The peak postoperative bladder temperature (38.2°C) was reached on average 181 min after arrival in the ICU for normothermic patients and after 202 min for the hypothermic ones. In the study of Grocott and colleagues  the maximum bladder temperature was reached between 6 and 10 h after arrival in the ICU. This difference might be due to a different anaesthetic protocol where warming was discontinued at a temperature of 36.5°C. In accordance with Grocott and colleagues , we did not find a difference in peak bladder temperature between patients managed with normothermic or hypothermic CBP.
Several criteria have to be fulfilled for successful extubation after cardiac surgery. Neurologic, pulmonary and haemodynamic are the most important ones. Although we found a correlation between the onset of thermoregulatory vasodilatation and extubation time after cardiac surgery, this does not imply causality. Thus, the decision to extubate cannot be based on the PI or the skin surface temperature gradient alone. These measures cannot supplement common extubation criteria. The time to reach a skin surface temperature gradient of 0°C and the time to extubate differed widely in some patients who were operated with normothermic CPB (Fig. 2). As reported recently, the PI and the skin surface temperature gradient may provide information about the eligibility for extubation in patients operated with normothermic CPB . Today, the data available regarding the relationship between temperature changes and fast track extubation is limited . Nonetheless, early extubation has proven to be safe . Skin surface temperature gradient can be measured easily. Pulse oximetry is standard in postoperative care and the derived PI can provide information about vasomotor tone and blood flow. Further studies are necessary to establish whether the demonstrated correlation between temperature and extubation time improves the decision in selecting patients for early extubation.
Our study has several limitations. It may be criticized for the use of vasoactive drugs and the lack of information regarding volume status and cardiac output. The number of patients studied was relatively small, and the power of statistical analysis is affected by the non-randomized design and the unequal size of the two groups. Another limitation is the fact that there were relatively more aortic valve replacements in the hypothermic group. Nevertheless, the information provided would be of help in clinical practice and should encourage further investigations of temperature-guided extubation.
In conclusion, this study demonstrates that extubation time is related to the transition from peripheral vasoconstriction to vasodilatation in patients following cardiac surgery. This relationship is independent of temperature management (normothermic or hypothermic) during CPB. Thus, this thermoregulatory change in skin surface temperature offers additional information for timing of early extubation in cardiac surgical patients.
1. Leslie K, Sessler DI. The implications of hypothermia for early trachel extubation following cardiac surgery. J Cardiothorac Vasc Anesth
1998; 12: 30–34.
2. Sessler DI. Perioperative heat balance. Anesthesiology
2000; 92: 578–596.
3. Rajek A, Lenhardt R, Sessler DI et al
. Tissue heat content and distribution during and after cardiopulmonary bypass
at 31°C and 27°C. Anesthesiology
1998; 88: 1511–1518.
4. The Warm Heat Investigators. Randomized trial of normothermic vs. hypothermic coronary bypass surgery. Lancet
1994; 343: 559–563.
5. Butler J, Rocker GM, Westaby S. Inflammatory response to cardiopulmonary bypass
. Ann Thorac Surg
1993; 55: 552–559.
6. Boulant JA. Role of preoptic-anterior hypothalamus in thermoregulation and fever. Clin Infect Dis
2000; 31: S157–S161.
7. Boulant JA. Thermoregulation. In: Mackowiak PA, ed. Fever: Basic Mechanisms and Management.
New York: Lippincott-Raven, 1997: 35–58.
8. Rubinstein EH, Sessler DI. Skin-surface temperature gradients correlate with fingertip blood flow in humans. Anesthesiology
1990; 73: 541–545.
9. House JR, Tipton MJ. Using skin temperature gradients or skin heat flux measurements to determine thresholds of vasoconstriction and vasodilation. Eur J Appl Physiol
2002; 88: 141–145.
10. Ozaki M, Sessler DI, Lopez M, Walter K. Pulse oximeter-based flow index correlates well with fingertip volume plethysmography. Anesthesiology
1993; 79(3A): 542.
11. Lima AP, Beelen P, Bakker J. Use of a peripheral perfusion index derived from the pulse oximetry signal as a noninvasive indicator of perfusion. Crit Care Med
2002; 30: 1210–1213.
12. Mort TC, Rintel TD, Altman F. The effects of forced-air warming on post bypass central and skin temperatures and shivering activity. J Clin Anesth
1996; 8: 361–370.
13. Ovrum E, Tangen G, Schiott C, Dragsund S. Rapid recovery protocol applied to 5658 consecutive ‘on-pump’ coronary bypass patients. Ann Thorac Surg
2000; 70: 2008–2012.
14. Cheng DC. Pro: Early extubation after cardiac surgery decreases intensive care unit stay and cost. J Cardiothorac Vasc Anesth
1995; 9: 460–464.
15. Cheng DC, Karski J, Peniston C et al
. Morbidity outcome in early versus conventional tracheal extubation after coronary artery bypass grafting: a prospective randomized controlled trial. J Thorac Cardiovasc Surg
1996; 112: 755–764.
16. Pezawas T, Rajek A, Skolka M, Schneider B, Plöchl W. Perspectives for core and skin surface temperature guided extubation in cardiac surgical patients after normothermic CPB. Intens Care Med
2004; 30: 1676–1680.
17. Johnson RI, Fox MA, Grayson A, Jackson M, Fabri BM. Should we rely on nasopharyngeal temperature during cardiopulmonary bypass
2002; 17: 145–151.
18. Wong DT, Cheng DC, Kustra R et al
. Risk factors of delayed extubation, prolonged length of stay in the intensive care unit, and mortality in patients undergoing coronary artery bypass graft with fast-track cardiac anesthesia. Anesthesiology
1999; 91: 936–944.
19. Thong WY, Strickler AG, Li S et al
. Hyperthermia in the forty-eight hours after cardiopulmonary bypass
. Anesth Analg
2002; 95: 1489–1495.
20. Grocott HP, Mackensen GB, Grigore AM et al
. Postoperative Hyperthermia is associated with cognitive dysfunction after coronary artery bypass graft surgery. Stroke
2002; 33: 537–541.
21. De Witte J, Sessler DI. Perioperative shivering: physiology and pharmacology. Anesthesiology
2002; 96: 467–484.
22. Frank SM, Kluger MJ, Kunkel SL. Elevated thermostatic setpoint in postoperative patients. Anesthesiology
; 2000; 93: 1426–1431.
23. Reis J, Mota JC, Ponce P, Costa-Pereira A, Guerreiro M. Early extubation does not increase complication rates after coronary artery bypass graft surgery with cardiopulmonary bypass
. Eur J Cardiothorac Surg
2002; 21: 1026–1030.