British Journal of Anaesthesia 1998; 81: 147–151
Pharmacological vasodilatation improves efficiency of rewarming from
hypothermic cardiopulmonary bypass
C. D. DEAKIN, G. W. PETLEY AND D. SMITH
Pharmacologically-induced vasodilatation has been
Summary proposed as a method to limit postoperative
An afterdrop in core temperature after hypother- hypothermia6 by improving rewarming of constricted
mic cardiopulmonary bypass (CPB) is related to vascular beds. In this previous study, pharmacologi-
inadequate peripheral rewarming. We proposed cal vasodilatation was combined with increased
that pharmacological vasodilatation during pump flow, and therefore increased heat delivery to
rewarming on bypass would improve peripheral ths patient, so that the contribution of each factor to
rewarming and reduce the degree of afterdrop. the subsequent reduction in postoperative hypother-
Fifty-nine of 120 patients were randomized to mia was unclear. This study was designed to deter-
receive a sodium nitroprusside (SNP) infusion mine whether pharmacological vasodilatation alone
during the rewarming stage of hypothermic during the rewarming phase of hypothermic CPB
CPB. Mean systemic vascular resistance (SVR) would improve the efficiency of rewarming and
during the rewarming phase of CPB was reduce postoperative hypothermia.
1129 dyne s91 cm95 in the control group and
768 dyne s91 m95 in the SNP group (P 0.001).
Patients receiving SNP rewarmed to 37.0 C Patients and methods
faster (299 min vs 376 min; P:0.003) and were After obtaining local Ethics Committee approval,
extubated earlier (490 min vs 621 min; P:0.001). 120 sequential patients undergoing elective first-time
Patients receiving SNP had a warmer mean coronary artery bypass grafting or valve replacement,
peripheral temperature (MPT) (32.9 C vs 32.4 C; or both, were entered into the study after written
P:0.05) on termination of CPB. Postoperative informed consent. Patients with peripheral vascular
core temperature fell less in the SNP group disease, diabetes mellitus or those subsequently
(35.6 C vs 35.2 C; P:0.01) as did MPT (31.8 C vs returning to theatre for operation were excluded.
31.2 C; P:0.004). SNP-induced vasodilatation Using a binary random number generator, patients
during rewarming from hypothermic CPB were randomly assigned to a control group or to
improves peripheral rewarming, reduces the receive sodium nitroprusside (SNP). SNP was
degree of postoperative core and peripheral administered via a syringe driver (Graseby 3400)
hypothermia and reduces time to extubation. (Br. from the start of rewarming on CPB to the termina-
J. Anaesth. 1998; 81: 147–151) tion of CPB at a rate adjusted manually to give a
mean arterial pressure (MAP) of 40 mm Hg. MAP
Keywords: surgery, cardiopulmonary bypass; hypothermia;
during rewarming in the control group was not
adjusted other than to maintain MAP below 80 mm
Hg using 1-mg increments of phentolamine as is our
During hypothermic cardiopulmonary bypass
usual practice. Patients receiving SNP did not receive
(CPB), core temperature is usually reduced to below
inotropic or vasoconstrictor agents concurrently.
30 C to reduce tissue oxygen consumption and
Peri-operative management was otherwise identical
afford some degree of cerebral and myocardial pro-
in both groups.
tection. Despite apparently adequate rewarming as
Patients received lorazepam 2 mg orally 2 h before
judged by a normal core temperature on termination
operation followed by morphine 5–10 mg i.m. and
of CPB, moderate postoperative hypothermia is
scopolamine 0.2–0.4 mg i.m. 1 h before surgery.
common.1 2 Morbidity from postoperative hypother-
Radial artery and peripheral venous cannulas were
mia includes increased time to extubation, increased
inserted under local anaesthesia. Anaesthesia was
myocardial work and oxygen consumption, impaired
induced using midazolam 0.02–0.05 mg kg91 i.v., fen-
coagulation,3 4 increased risk of wound infection and
tanyl 10–15 g kg91 i.v. and pancuronium 0.1 mg kg91
prolonged hospital stay.5 Postoperative hypothermia
i.v. The trachea was intubated and the lungs mechani-
cannot be explained by heat loss to the environment
cally ventilated using oxygen in nitrous oxide,
alone and is thought to result from a failure of heat
adjusted to maintain normocapnia. A heat and mois-
transfer to vasoconstricted peripheral tissues during
ture exchanger was placed in the breathing circuit. A
rewarming on CPB, with a subsequent shift of heat
from a warm core to a cold periphery.4 6 Increasing
blood flow to vascular beds by increasing pump flow C. D. DEAKIN, MA, MRCP, FRCA; D. SMITH MD, FRCA, (Department of
during rewarming on bypass reduces the degree of Anaesthetics); G. W. PETLEY, BSC, PHD, MINSTP, CPHYS, (Medical
Physics and Bioengineering); Southampton General Hospital,
postoperative hypothermia,7 but the resultant pres- Tremona Road, Southampton SO16 6YD. Accepted for publica-
sure in the arterial cannula of the extracorporeal tion: March 18, 1998.
circuit may cause mechanical damage to blood. Correspondence to C. D. D.
148 British Journal of Anaesthesia
central venous catheter was then inserted. Anaesthesia minute was summated to give an instantaneous TEB.
was maintained using propofol 80–150 mg h91 and TEB is thus zero at commencement of bypass,
isoflurane to 1.2% as necessary. Midazolam 4 mg and becomes negative as the patient is cooled, and
pancuronium 4 mg were added to the pump prime returns to a positive value as the patient is rewarmed
fluid. before termination of bypass.
Core temperature was measured using a calibrated Operating theatre and intensive care environments
tympanic membrane thermocouple temperature were maintained at a stable state as allowed by the air
probe (Mallinckrodt, Northampton, UK) placed conditioning. Operating theatre temperature was
before induction of anaesthesia after visually con- maintained between 20–24 C and relative humidity
firming with an otoscope that the external auditory between 38–50%. Temperature in the intensive care
meatus was free from wax. Peripheral temperature unit was maintained between 22–25 C. No heating/
was recorded using four calibrated thermistors cooling mattress was used during the operation. After
(Yellow Springs 400) placed on the chest wall, upper bypass, all i.v. crystalloid, colloid and blood products
arm, thigh and calf before induction of anaesthesia. were administered through a blood warmer heated to
Mean skin temperature (MST) was calculated using 37.8 C until core temperature reached 37 C.
a weighted equation as described by Ramanathan8: Patients were covered by a standard hospital woollen
blanket and cotton sheet. No active postoperative
MST=0.3 (chest wall+upper arm)+0.2 (leg+thigh)
warming was used.
All temperatures were recorded to a temperature log- Adequacy of rewarming was assessed by recording
ger9 at 1-min intervals from induction of anaesthesia the coldest postoperative core temperature, following
until postoperative core temperature had returned to termination of CPB (fig. 1A) and the time for the
37.0 C. postoperative core temperature to return to 37 C
The CPB unit incorporated a Dideco D703 (Sorin, following afterdrop (fig. 1B). Time rewarming on
Italy) hollow fibre membrane oxygenator with inte- bypass was the time from the start of rewarming to
gral heat exchanger primed with lactated Ringer’s the termination of CPB (fig. 1C). Time to extubation
solution. During hypothermic CPB, core tempera- was the time from the end of bypass to extubation of
tures were reduced to 28–30 C. Pump flow during the patient. The doctor deciding when to extubate
cooling and rewarming was adjusted according to a the patient was blinded to the group to which the
standardized local procedure. Flow was indexed at patient was allocated. Statistical analysis was per-
2.4 litre min91 m92 above 30 C, 1.8 litre min91 m92 formed with Microsoft Excel 5.0 using unpaired
between 28–30 C, and 1.2 litre min91 m92 below two-way t tests. Results are given as mean values with
28 C, except during cooling when it was reduced to confidence intervals (CI) for the mean difference
approximately 1.2 litre min91 m92. Rewarming aimed between the control and SNP group. Significance
to restore the patients core temperature to 37 C was taken as P 0.05.
before termination of CPB.
Heat extracted and returned to patients whilst on
CPB was recorded using a previously described
energy balance machine.10 The apparatus comprised One hundred and twenty patients were studied, of
an IBM compatible personal computer running ded- whom 59 were randomly assigned to receive SNP.
icated software which interactively collects data from There was no significant difference between control
a specifically designed energy flux measurement sys-
tem. The calculation of energy balance is based upon
the Pick principle using pump flow and arterial and
venous temperature in the perfusion circuit.
Net thermal energy balance (TEB) is calculated from
E=sp ∫0 (Ta9Tv) Q(t) δt
where: E:net thermal energy balance during CPB
(kJ), Q:pump flow at time t (litre min91), s:specific
heat capacity of blood (3.84 kJ kg91 C), p:density of
perfusate (approximately 1.0 kg litre91), a:arterial
infusion temperature ( C), Tv:venous infusion tem-
perature ( C), t:time.
The temperatures of arterial and venous blood
were recorded at the inflow and outflow to the heat
exchanger/oxygenator unit using previously cali-
brated thermistor probes (Electromedics Inc,
Englewood CO 80112 USA, model No 4700
M344700) specifically adapted to fit the gold insu-
lated temperature probe ports on the inflow to the
heat exchanger (Tv) and the outflow from the oxy- Figure 1 Core temperature changes during hypothermic
genator (Ta). Rotation of the pump was detected by cardiopulmonary bypass. 1 : Induction; 2 : start of active cooling
an infrared reflective sensor placed over the pump on cardiopulmonary bypass; 3 : stop active cooling; 4 : start
rewarming; 5 : separation from cardiopulmonary bypass.
head cover. As a marked limb passed beneath the A : Coldest post-operative core temperature following termination
detector, the sensor registered one rotation equiva- of CPB; B : time for the post-operative core temperature to return
lent to 46-ml volume. The energy flux measured each to 37.0 C following the afterdrop; C : time rewarming on bypass.
Rewarming from cardiopulmonary bypass 149
and SNP groups in patient characteristics or environ- thermal energy balance between control (805 kJ) and
mental conditions. Patient characteristics are given in SNP groups (764 kJ) (95%CI -59.9 kJ to 141.7 kJ;
table 1. Ventilation was controlled manually for two P:0.42), a power of 0.05 was insufficient to conclude
patients in the control group for a prolonged period that this was a true negative result.
because of cardiovascular instability in one patient
and postoperative bleeding in the other. Time to
extubation of these patients was therefore excluded Discussion
from the data. The power of the study was calculated Our results showed that pharmacological vasodilata-
from the time to rewarm to 37 C in both groups. tion with SNP during the rewarming phase of CPB
Standardized difference was 0.564 giving a power of rewarmed the patient more efficiently as shown by
0.87 using the nomogram described by Altman.11 warmer peripheral temperatures on termination of
Mean arterial pressure during the rewarming phase bypass, suggesting that heat transfer to peripheral
of CPB was 53.5 mm Hg in the control group and vascular beds had been increased through increased
40.1 mm Hg in the SNP group (95%CI 10.8 to blood flow. The subsequent postoperative hypother-
15.9 mm Hg; P 0.001). This equates to a mean sys- mia was less in patients receiving SNP, shown by
temic vascular resistance (SVR) of 1129 dyne s91 cm95 greater postoperative minimum core and peripheral
and 768 dyne s91 cm95 respectively. There was no sig- temperatures and shorter time to rewarm to 37 C.
nificant difference in the amount of perioperative i.v. Clinically, improved rewarming resulted in earlier
fluid administered to the two groups. extubation.
Time to rewarm to 37 C was 376 min in the con- Postoperative hypothermia is commonly seen in
trol group compared with 299 min in the SNP group patients after hypothermic CPB and is greater than
(95%CI 28 to 129 min; P:0.003). Time to extuba- that observed in patients undergoing thoracic opera-
tion was 621 min in the control group compared with tions not requiring the use of CPB.12 Hypothermia
490 min in the SNP group (95%CI 70 to 232 min; during CPB causes intense vasoconstriction and
P:0.001). greatly reduces blood flow through peripheral vascu-
Patients receiving SNP had a warmer MST on lar beds such as muscle and fat. During the rewarm-
termination of bypass (32.9 C) compared with the ing phase of CPB, the periphery is slow to warm
control group (32.4 C) (95%CI 0.06 to 0.94 C; because of the reduced blood supply to these vascular
P:0.05). There was no difference in core tempera- beds and the periphery remains relatively hypother-
ture on termination of bypass between the control mic when CPB is discontinued, despite attainment of
group (38 C) and SNP group (38.1 C) (95%CI - a normal core temperature.13 Once bypass is discon-
0.41 to 0.18 C; P:0.52). tinued, the core temperature decreases as heat is
Postoperative core temperature decreased further transferred to the cold periphery. Evidence for this
in the control group (35.2 C) than the SNP group theory is growing.4 6 Pharmacological vasodilatation
(35.6 C) (95%CI 0.1 to 0.6 C; P:0.01). Minimum may force the periphery to vasodilate during rewarm-
postoperative peripheral temperature was warmer in ing, increase blood flow and heat transfer to the cold
the group receiving SNP (31.8 C) compared with peripheral tissues, thereby reducing the subsequent
control (31.2 C) (95%CI 0.14 to 0.90 C; decrease in core temperature.
P:0.004). Once postoperative core temperature had The use of vasodilatation during rewarming from
returned to 37.0 C, there was no difference in the hypothermia on CPB was first reported by Noback
core-peripheral temperature difference between the and Tinker.6 Twenty control patients were compared
control (3.65 C) and SNP groups (3.59 C) (95%CI with eight patients receiving SNP during rewarming
90.3 to 0.5; P:0.72). on bypass. SNP was administered at a rate to main-
There was no difference in time spent rewarming tain MAP 70 mm Hg as CPB pump flows were
on bypass between control group (30.6 min) or SNP increased. Increased CPB pump flow alone during
group (31.8 min) (95%CI -2.8 to 5.2 min; P:0.55). rewarming increases perfusion of muscle beds and
Bypass pump flows during rewarming were 8.4% reduces the degree of postoperative hypothermia.7
higher in the SNP group (control 3.92 compared with Thus, although the study of Noback and Tinker
SNP 4.25 litre min91; 95%CI 0.17 to 0.48 litre min91; reduced the degree of postoperative hypothermia,6
P:0.001). Although there was no difference in mean the relative contribution of increased heat delivery
Table 1 Patient characteristics and environmental conditions.
95% CI of mean difference
Control group SNP group between groups P
Theatre temp. (°C) 22.2 22.3 90.20 to 0.37 0.64
ITU temp. (°C) 23.7 23.7 90.32 to 0.41 0.98
Age (yr) 65.7 64.0 92.40 to 5.81 0.41
Weight (kg) 71.9 77.0 90.2 to 10.4 0.06
Height (cm) 169.0 170.3 92.22 to 4.91 0.46
Body mass index (kg m92) 25.1 26.4 90.01 to 0.14 0.09
Body surface area (m2) 1.82 1.88 90.22 to 2.81 0.08
Mean SNP dose ( g kg91 min91) 1.40 <1.15 (SD)
Rate of post-op. propofol
infusion (ml h91) 13.2 13.2 90.97 to 1.14 0.87
Cardioplegia volume (ml) 1227 1179 958 to 154 0.37
150 British Journal of Anaesthesia
and pharmacological vasodilatation is difficult to postoperative rewarming have been unable to control
distinguish. ambient temperature more closely than our study6
Aps, Hutter and Williams14 reviewed the postoper- or do not report environmental conditions.20–22
ative progress of 143 cardiac surgery patients in Environmental temperature did not show any signifi-
whom specific attempts to ensure adequate rewarm- cant correlation with time to rewarm to 37 C or
ing on bypass had been made. Patients received infu- coldest postoperative temperature.
sions of either SNP or glyceryl trinitrate (GTN) at a By increasing peripheral blood flow, SNP adminis-
maximum rate that still allowed for adequate arterial tered during rewarming on bypass appears to
perfusion pressures during rewarming on bypass improve the efficiency of peripheral rewarming and
(details not given). Rewarming was further opti- reduce the degree of postoperative hypothermia.
mized by the use of patient insulation, use of a warm-
ing mattress and heating of inspiratory gases to
40 C. Although this technique minimized the degree
of postoperative hypothermia, the relative contribu- This work was funded in part by the Edith Powell and Geoffrey
tions of each technique cannot be distinguished. Holt Research Awards from the British Medical Association
(1995). We thank Mr S. A. Livesey, Mr V. T. Tsang and Mr J. L.
Both GTN and SNP have been used as peripheral Monro for allowing us to study their patients, and the nursing staff
vasodilators in these earlier studies. Recent work sug- in the Cardiothoracic Intensive Care Unit for their cooperation
gests that both GTN and SNP have similar arterio- with the study.
venous vasodilatory profiles.15 SNP however was
chosen in preference to GTN for this study because
its greater potency and shorter half-life make it more References
suitable for acute haemodynamic manipulation. SNP 1. Deakin C, Sewell A, Clewlow F, Pierce J. A comparison of
dilates large arteries,16 small arteries and arterioles,15 core temperature with thermal energy balance in predicting
adequacy of rewarming from hypothermic cardiopulmonary
increases limb blood flow through vasodilatation of bypass. British Journal of Anaesthesia 1996; 76: A41.
skeletal muscle vasculature17 and increases skin perfu- 2. Huang F, Wang M, Huang H. Differences in temperature
sion.18 SNP is not known to possess any thermogenic changes between pediatric and adult patients after cardiopul-
properties. monary bypass. Journal of Cardiothoracic and Vascular
By using the same rewarming plan for control and Anesthesia 1993: 7: 66–68.
3. Baker J, Baker A, Mazer C. Determinants of postoperative
SNP groups, we hoped to achieve the same bypass hypothermia after normothermic cardiopulmonary bypass.
flow rates during rewarming in both groups. Journal of Cardiothoracic and Vascular Anesthesia 1995; 9:
Surprisingly, flow was 8.4% higher (P:0.001) in the 154–157.
SNP group. Although it may be expected that 4. Pujol A, Fusciardi J, Ingrand P, Baudouin D, Le Guen A,
Menu P. Afterdrop after hypothermic cardiopulmonary
increased flow would be associated with increased bypass: The value of tympanic membrane temperature moni-
heat delivery (TEB) during rewarming to the SNP toring. Journal of Cardiothoracic and Vascular Anesthesia 1996;
group, the power of 0.05 for these data is insufficient 3: 336–341.
to distinguish any difference between the two groups. 5. Kurz A, Sessler D, Lenhardt R. Perioperative normotherrhia
Although the relative contribution of increased CPB to reduce the incidence of surgical-wound infection and
shorten hospitalization. Study of Wound Infection and
flow and vasodilatation to heat transfer cannot be Temperature Group. New England Journal of Medicine 1996;
distinguished directly, the small difference in CPB 334: 1209–1215.
flow between our two groups and the differential 6. Noback C, Tinker J. Hypothermia after cardiopulmonary
changes in core–peripheral temperature in the SNP bypass in man: amelioration by sodium nitroprusside induced
vasodilation during rewarming. Anaesthesiology 1980; 53:
group leads us to believe that SNP was at least par- 277–280.
tially responsible for the observed differences. 7. Stanley T, Jackson J. The influence of blood flow and arterial
In addition, the core temperature at the end of blood pressure during cardiopulmonary bypass on deltoid
bypass was the same in both groups. We therefore muscle gas tensions and body temperature after bypass.
believe that the observed differences between the Canadian Anaesthetists Society Journal 1979; 26: 277–281.
8. Ramanathan N. A new weighting system for mean skin tem-
control and SNP groups are attributable, at least in perature of the human body. Journal of Applied Physiology
part, to the effects of pharmacological vasodilatation. 1964; 19: 531–533.
The larger differences in peripheral temperature 9. Petley G, Clitheroe S, Clewlow F, Deakin C, Chauhan A.
between the control and SNP groups observed by Development and application of a general purpose ambulatory
monitor. Medical Engineering and Physics 1998; 20: 33–39.
Noback and Tinker6 may be caused by higher flows 10. Sansome A. Apparatus for the real-time calculation of thermal
enabling a higher rate of SNP infusion and a greater energy flux during cardiopulmonary bypass. British Journal of
degree of peripheral vasodilatation. Hospital Medicine 1989; 42: 494.
We avoided the use of warming blankets during 11. Altman D. Statistics and ethics in medical research: How large
rewarming from bypass and forced warm air convec- a sample? British Medical Journal 1980; 281: 1336–1338.
12. Carli F, Aber V. Thermogenesis after major elective surgical
tion blankets postoperatively because vasodilatation procedures. British Journal of Surgery 1987; 74: 1041–1045.
combined with active surface warming increases heat 13. Benson J, Patla V, Bonser R, Hutton P. Use of i.m. tempera-
transfer to the patient19 and may mask the effects of ture probes during cardiopulmonary bypass in humans.
thermal energy redistribution in the body caused by British Journal of Anaesthesia 1995; 75: 483–485.
14. Aps C, Hutter J, Williams B. Anaesthetic management and
the SNP infusion. Noback and Tinker used heating postoperative care of cardiac surgical patients in a general
mattresses during the rewarming phase of CPB in 16 recovery ward. Anaesthesia 1986; 41: 533–537.
of the 20 patients in the control group and 6 of the 8 15. MacAllister R, Calver A, Riezebos Jo, Collier J, Vallance P.
patients in the SNP group.6 They failed to show any Relative potency and arteriovenous selectivity of nitrovaso-
difference between the two groups but this may be dilators on human blood vessels: An insight into the target-
ing of nitric oxide delivery. Journal of Pharmacology and
because of the small numbers in the study. Experimental Therapeutics 1995; 273: 150–160.
Operating room and intensive care unit temperature 16. Komori K, Mawatari K, Itoh H, Sugimachi K. Impaired
varied by up to 4.0 C. Other studies investigating vasodilation of peripheral arteries in response to acetylcholine
Rewarming from cardiopulmonary bypass 151
in human beings with abdominal aortic aneurysm. Journal of 20. Jenkins I, Karliczek G, Geus F, Brenken U. Postbypass
Vascular Surgery 1994; 20: 803–807. hypothermia and its relationship to the energy balance of car-
17. Steinberg H, Brechtel G, Johnson A, Fineberg N, Baron A. diopulmonary bypass. Journal of Cardiothoracic and Vascular
Insulin-mediated skeletal muscle vasodilation is nitric oxide Anesthesia 1991; 5: 135–138.
dependent. A novel action of insulin to increase nitric oxide 21. Jani K, Carli F, Bidstrup B, Royston D, Taylor K. Temperature
release. Journal of Clinical Investigation 1994; 94: 1172–1179. following cardiopulmonary bypass procedures; the effects of
18. Morris S, Shore A, Tooke J. Responses of the skin microcircu- active rewarming. Life Support Systems 1986; 4: 269–272.
lation to acetylcholine and sodium nitroprusside in patients 22. Davis F, Parimelazaghagan K, Harris E. Thermal energy
with NIDDM. Diabetologia 1995; 38: 1337–1344. balance during cardiopulmonary bypass with moderate
19. Janke E, Pilkington S, Smith D. Evaluation of two warming hypothermia in man. British Journal of Anaesthesia 1977; 49:
systems after cardiopulmonary bypass. British Journal of 1127–1132.
Anaesthesia 1996; 77: 268–270.