Efficacy of forced-air warming and warmed intravenous fluid for prevention of hypothermia and shivering during caesarean delivery under spinal anaesthesia

A randomised controlled trial

Abstract

BACKGROUND 

Peri-operative hypothermia and shivering are frequent events in patients during caesarean delivery under spinal anaesthesia.

OBJECTIVE 

We assessed the efficacy of combined pre-anaesthetic forced-air warming in combination with warmed intravenous fluid infusion for preventing hypothermia and shivering during caesarean delivery under spinal anaesthesia.

DESIGN 

A randomised controlled study.

SETTING 

A tertiary care teaching hospital from July 2017 to April 2018.

PATIENTS 

A total of 50 pregnant women, American Society of Anaesthesiologists physical status 2, aged 20 to 45 years, scheduled for caesarean delivery under spinal anaesthesia.

INTERVENTION 

Patients were enrolled and randomised into two groups: an active warming group (n = 25), which received combined pre-anaesthetic whole body forced-air warming for 15 min and prewarmed intravenous fluids, and a control group, which received no active warming or warmed fluids (C group; n = 25). Spinal anaesthesia was induced with 10 mg bupivacaine containing fentanyl (10 μg).

MAIN OUTCOME MEASURES 

Tympanic membrane temperature and shivering severity were measured at baseline and every 10 min during surgery, and then every 10 min for 1 h postoperatively. Neonatal outcomes (tympanic membrane temperature at birth, umbilical venous blood pH, Apgar score) were also recorded.

RESULTS 

The incidences of peri-operative hypothermia (0 vs. 48%, P < 0.001) and shivering (22 vs. 52%, P = 0.031) were significantly lower in the active warming than in the C group. The maximum temperature change was also significantly lower in the active warming than in the C group. Maternal thermal comfort scores were higher in the active warming than in the C group. Neonatal parameters were not significantly different between the groups.

CONCLUSION 

The combination of pre-anaesthetic forced-air warming and warmed intravenous fluid infusions appears to be effective for preventing hypothermia and shivering during caesarean delivery under spinal anaesthesia.

TRIAL REGISTRATION 

This trial was registered with Clinical Trials.gov (identifier: NCT03256786).

Introduction

Peri-operative hypothermia or shivering are reported to occur in up to 60 and 85% of women, respectively, during caesarean delivery under spinal anaesthesia.^1,2 Hypothermia and shivering increase the risk of complications such as surgical site infection, myocardial ischemia and peri-operative coagulopathy.^3–5 In addition, maternal hypothermia may cause hypothermia in the newborn.^6–8 A recent meta-analysis found that active warming during caesarean delivery can decrease the incidence of hypothermia and shivering.⁸ However, single modality interventions have shown little efficacy in preventing hypothermia during caesarean delivery.^9–11 Multimodal intra-operative active warming, with the use of forced-air warming along with warmed intravenous (i.v.) fluid infusions, has also proved to be ineffective for prevention of hypothermia and shivering.¹²

Peri-operative hypothermia is caused by the combined effect of anaesthesia-induced impairment of the thermoregulatory mechanism, environmental conditions in the operating theatre and surgical factors.¹³ The most important cause is heat redistribution from the core to the periphery due to vasodilation after the induction of general or neuraxial anaesthesia.^14,15 The amount of heat redistribution depends on the temperature gradient between the peripheral and core compartments. Active warming of the skin before surgery (i.e. prewarming) can reduce this gradient by increasing the heat content of the peripheral thermal compartment.^16,17 However, use of a short period of prewarming alone has a little effect on peri-operative core temperature and does not modify the intra-operative temperature decline. Although heat redistribution is the most important cause of peri-operative hypothermia, surgical factors also contribute to systemic heat loss during the intra-operative period.¹³

The aim of this randomised controlled trial was to evaluate the efficacy of a combination of pre-anaesthetic forced-air warming and warmed i.v. fluid infusions for prevention of maternal hypothermia and shivering in women undergoing caesarean delivery under spinal anaesthesia.

Materials and methods

This prospective, single blind, randomised, controlled study was carried out from July 2017 to April 2018. The study was approved by the Ethics Committee of our hospital and was registered at ClinicalTrials.gov: NCT03256786. Written informed consent was obtained from all participants.

Fifty pregnant women scheduled for elective caesarean delivery under spinal anaesthesia were randomly assigned to one of two treatment groups: an active warming group that received combined active warming (i.e. pre-anaesthetic forced-air warming for 15 min and prewarmed i.v. fluid infusions; or a control (C) group, which received routine care (i.e. blankets only; no active warming, and fluids at room temperature). Patients were eligible for inclusion if they were aged 20–45 years and they were American Society of Anaesthesiologists physical status II. Exclusion criteria were pre-operative core temperature more than 37.2°C; severe endocrine, cardiovascular or respiratory disease; current treatment with anticoagulant therapy; twin pregnancy; or body weight less than 50 or more than 100 kg.

The random allocation sequence was created by one of the investigators using a computer-generated randomisation schedule. On arrival in the pre-operative holding room, another investigator, who was not involved in data collection, opened an opaque sealed envelope containing the patient's group assignment.

No premedication was given to any patient. At our institution, the temperatures in the pre-operative holding area, the operating room and the postanaesthesia care unit (PACU) are maintained between 22°C and 24°C. In the pre-operative holding area, baseline core temperature was measured using an infrared tympanic thermometer (ThermoScan IRT 1020; Braun, Kronberg, Germany). The device accuracy mean error was found to be ±0.2°C with patient temperatures from 35.8 to less than 37°C and ±0.3°C when patient temperatures were less than 35.8°C. The highest of three consecutive measurements in the same ear was recorded. After baseline temperature recording, the control group patients received usual care, that is they were covered with two layers of a warmed cotton blanket, but received no active warming. Active warming group patients received 15 min of active warming with a forced-air blanket (WarmTouch Full-Body Multi-Access Blanket; Covidien PLC, Mansfield, Massachusetts, USA), which was placed over the entire body and then covered with a cotton blanket. During the warming period, the thermostat of the forced-air warmer was set at 41°C. At the end of active warming, the forced-air warming blanket was removed and replaced with two layers of a prewarmed cotton blanket.

On arrival in the operating room, standard monitoring (electrocardiogram, pulse oximeter, noninvasive blood pressure) was applied (patient monitor M1205A; Philips, USA/MicroO2; Siemens, Germany). Spinal anaesthesia was induced by an investigator unaware of the group allocation. With the patient in the left lateral decubitus position, the puncture site was first infiltrated with lidocaine, and then dural puncture was performed at the L3–4 or L4–5 level with a 25-gauge Whitacre spinal needle. After confirmation of free-flowing, clear cerebrospinal fluid, 10 mg 0.5% hyperbaric bupivacaine containing fentanyl (10 μg) was injected into the subarachnoid space. The extent of sensory blockade was tested by pinprick (pin prick felt, but not painful), and anaesthesia was considered adequate if the sensory block was at the T6 dermatomal level or higher.

Active warming group patients received prewarmed i.v. fluids stored at 41°C in a warming cabinet (KRS-205; Karis, Gyeonggi-do, Korea) for at least 8 h, whereas control group patients received i.v. fluids stored at room temperature. The i.v. fluid was taken out of storage only just before infusion. Fluid administration was started by nurses who were not involved in the investigation. A co-load of 8 to 10 ml kg⁻¹ lactated Ringer solution was administered intravenously using a pressure bag, commencing at the time of the intrathecal injection. If SBP or mean blood pressure fell below 80% of baseline, or if SBP fell below 90 mmHg, or if mean blood pressure fell below 60 mmHg, phenylephrine 50 μg was injected intravenously. If the blood pressure continued to decrease, the same dose was repeated. If hypotension was accompanied by low heart rate, ephedrine 4 mg was injected intravenously. Oxygen was supplied at the rate of 5 l min⁻¹ through a facemask throughout the surgery. After the spinal anaesthetic, all patients were covered with two layers of a warmed cotton blanket from the neck to the nipple, and over the lower extremities, to protect against heat loss. Neither group received forced-air warming intra-operatively. After surgery, women in both groups received whole body forced-air warming in the PACU, with the device thermostat set to 41°C.

At birth, the infant's umbilical vein blood was sampled for pH measurement, and tympanic temperature was recorded. Apgar scores at 1 and 5 min were determined by the paediatrician. One investigator, blinded to the group allocation, evaluated the perioperative outcomes. For all participants, core temperature, vital signs and shivering scores were recorded on arrival in the pre-operative holding area (baseline), on arrival in the operating room, immediately after induction of spinal anaesthesia, every 10 min after induction of anaesthesia and every 10 min for 1 h postoperatively. Maternal hypothermia was defined as core body temperature less than 36°C. Shivering was rated according to the Bedside Shivering Assessment Scale (0 = no shivering; 1 = shivering localised to neck and thorax; 2 = shivering involving the upper extremities with or without shivering in the thorax; and 3 = shivering involving the entire body).¹⁸ Patients with a shivering score of 2 after induction of spinal anaesthesia had i.v. meperidine (25 mg) administered. Maternal thermal comfort scores were obtained at admission to the PACU using a verbal numerical scale, ranging from 0 = completely unsatisfied to 100 = completely satisfied. Maximum temperature change was defined as the difference between the baseline temperature and the lowest temperature observed during the study period. Total volume of intra-operative i.v. fluids, estimated blood loss and duration of operation were also recorded.

The primary outcome was the incidence of hypothermia in the peri-operative period (i.e. from arrival in the pre-operative holding area to 1 h after surgery). The secondary outcomes were the incidence of shivering in the peri-operative period, maximum maternal temperature change, thermal comfort score, neonatal temperature at birth, umbilical venous blood pH and Apgar scores at 1 and 5 min.

Statistical analyses

In a previous study, the difference between the incidences of peri-operative hypothermia between a warming group and a control group was reported as 27% (64 vs. 91%).¹² In our study, the sample size was calculated by assuming that the difference in hypothermia incidence would be more than 35% (active warming group vs. C group = 5 vs. 40%). We calculated that a total of 22 patients per group would be needed to demonstrate this difference with an alpha error of 0.05 and a power of 80%. We included 50 patients in each group to allow for possible protocol violations over the study period.

Data were summarised as mean (SD), or number (proportion). The Kolmogorov–Smirnov test was used to identify the variables with normal distribution. The independent-sample t-test was used to compare normally distributed variables and the Mann–Whitney U-test to compare nonnormally distributed variables. Categorical variables were analysed using the Pearson Chi-square or Fisher's exact tests, as appropriate. Statistical analyses were performed using SAS version 9.2 (SAS Inc., Cary, North Carolina, USA). Differences were considered statistically significant when P value was less than 0.05. Cohen's effect size d (continuous variable) or h (category variable) were also used to compare variables by R package, version 3.4.4.

Results

A total of 50 patients (25 patients in each group) were initially enrolled in the study. Two patients in the active warming group were excluded from the analyses because of conversion to general anaesthesia, leaving 23 women in the active warming group and 25 women in the C group for the final analyses (Fig. 1). There were no statistically significant differences between the groups in terms of patient characteristics and baseline data (Table 1).

Fig. 1: Consort flow diagram. AW group, active warming group; C group, control group.

Table 1: Patient characteristics and peri-operative data

Table 2 presents the maternal outcomes in the two groups. The incidence of peri-operative hypothermia was significantly lower in the active warming group than in the C group (0 vs. 48%, P < 0.001). The incidence of shivering in the peri-operative period was significantly lower in the active warming group than in the C group (22 vs. 52%, P = 0.031). The proportion of patients with a shivering score of 2 was not significantly different between the groups (4 vs. 12%, P = 0.338). No patient had severe shivering score (= 3) in either group. The proportion of patients receiving meperidine was not significantly different between the groups (4 vs. 12%, P = 0.338). The maximum change in maternal temperature was also significantly lower in the active warming group than in the C group (P = 0.002). Maternal thermal comfort scores were higher in the active warming than in the C group (P = 0.003). However, the incidence of nausea/vomiting did not differ significantly between the two groups.

Table 2: Maternal outcomes during the study period

Figure 2 illustrates the changes in core temperature during the peri-operative period in the two groups. In the active warming group, after the pre-anaesthetic whole body forced-air warming period, core temperature increased and was significantly higher than the baseline core temperature. Starting from immediately after the pre-anaesthetic forced-air warming up to the end of the study period, the active warming group exhibited a significantly higher mean core temperature than the control group. The difference in core temperature between the two groups increased further immediately after the prewarmed i.v. fluid co-load, with the maximal difference being observed at arrival in the PACU. In the PACU, after whole body forced-air warming, the core temperature increased in both groups.

Fig. 2: Peri-operative changes in core temperature of patients both active warming (AW group, •) and no warming (C group, ◊). Values are presented as the mean and the whiskers are SD. (a) is the pre-anaesthetic forced-air warming for 15 minutes, (b) is the period of warmed i.v. fluid infusion in the AW group. Baseline, arrival in pre-anaesthetic holding area; SA 0 and SA 10, 0 to 10 min after spinal anaesthesia; OP 10 to OP 40, during surgery; PACU 0 to PACU 60, during postanaesthetic care unit. ^* P < 0.05 - AW group vs. C group. ^† P < 0.05 vs. baseline temperature.

Umbilical venous blood pH, neonatal temperature at birth and Apgar score were not significantly different between the groups (Table 3).

Table 3: Neonatal outcomes

Discussion

The aim of this study was to investigate the efficacy of combined pre-anaesthetic forced-air warming and prewarmed i.v. fluid infusions for preventing hypothermia and shivering during caesarean delivery under spinal anaesthesia. The study findings suggest that this combined active warming method may be an effective strategy for the prevention of hypothermia and shivering.

Previous studies have investigated the efficacy of several different warming techniques and different durations of warming. However, there is still no consensus on the efficacy of active warming during caesarean delivery in improving maternal outcomes. Recently, during spinal anaesthesia in pregnant women undergoing caesarean delivery, Cobb et al studied the effect of intra-operative warmed i.v. fluid and lower body forced-air warming to prevent hypothermia and shivering. The mean core temperature in the warming group on arrival at the PACU was 35.9 (0.5) °C, which was 0.4°C higher than the mean core temperature in the control group; nevertheless, hypothermia occurred in 64% of patients in the warming group. This suggests that warming interventions applied only during the intra-operative period are not sufficient to prevent peri-operative hypothermia and shivering under spinal anaesthesia.

During caesarean delivery performed under spinal anaesthesia, heat loss occurs through several mechanisms. Neuraxial anaesthesia results in inhibition of central thermoregulatory control¹³ as well as peripheral vasodilatation in blocked dermatomes. As a consequence, there is redistribution of body heat from the core to the peripheral compartment in dermatomes affected by the spinal anaesthetic: this is the primary cause of hypothermia during the first hour after induction of general or neuraxial anaesthesia, and remains an important cause of hypothermia during the first 3 h.^14,19 Moreover, after a drop in the core temperature, hypothermia triggers vasoconstriction and shivering in unblocked regions. However, vasoconstriction above the level of the block is insufficient to prevent further hypothermia. Active prewarming prior to induction of anaesthesia increases the heat content of the peripheral compartment, reduces the core-to-periphery temperature gradient, and thus reduces or prevents redistribution hypothermia.²⁰ In our study, pre-anaesthetic whole body forced-air warming for 15 min increased core temperature significantly above baseline core temperature in the active warming group. However, as previous studies have shown, use of short prewarming alone has a little effect on peri-operative core temperature and does not modify the intra-operative temperature decline. Although heat redistribution is the most important cause of peri-operative hypothermia, surgical factors also contribute to systemic heat loss during the intra-operative period.¹³ Therefore, effective prevention of hypothermia requires intra-operative active warming in addition to prewarming. Infusion of warmed i.v. fluid acts through peripheral and central conduction, with rapid distribution of heat throughout the body.⁸ As pregnant women undergoing caesarean delivery receive relatively large amounts of fluid during surgery, warmed i.v. fluid may be particularly effective in these patients.

In our study, the incidence of shivering in the peri-operative period was significantly lower in the active warming group than in the C group. Shivering is a physiological mechanism to increase heat production. The mechanism for shivering initiation is not fully understood. The hypothalamus acts as the integrating centre for thermal inputs from peripheral and nonhypothalamic thermosensitive neurons. If hypothalamic temperature decreases as a result of heat loss, shivering could be triggered.^21–23 Therefore, maintaining normothermia can help prevent the thermoregulatory component of shivering. Meperidine is commonly used for controlling postanaesthetic shivering. The mechanism of meperidine has yet to be fully elucidated, but it has been found to be better for controlling shivering than pure u-opioid agonists such as fentanyl, or morphine. Thus, the antishivering effect of meperidine may be mediated in part by activation of k-opioid receptors, but not of u-opioid receptors. Meperidine also acts on the thermoregulatory centre directly.²⁴

We found no significant differences in neonatal outcomes between the two groups. These results are mostly concordant with previous studies. However, several previous studies have found that umbilical artery blood pH tends to be higher in the group receiving warming interventions. One explanation is that this may be due to decrease of maternal blood pH in patients with hypothermia-induced shivering; however, the exact mechanism has not yet been identified.¹⁶ In the present study, pregnant women in both groups had mean body temperature more than 36°C and, in addition, the baby was born prior to onset of maternal shivering in all cases; this may explain why no significant difference in foetal blood pH was seen between the groups.

This study has a limitation. We used infrared tympanic thermometers to measure core temperature. The reliability of these thermometers for measuring core temperature has been questioned.^25,26 Although tympanic contact thermistors and thermocouple ear thermometry may be more reliable, they are not widely used in clinical practice. Invasive core temperature monitors, such as oesophageal probes, are difficult to use in patients undergoing spinal anaesthesia. A previous study has demonstrated that infrared systems can provide accurate measurements of surface temperatures, and it is the measurement site rather than the device used that determines the accuracy.²⁷ In this study, to reduce measurement error, we recorded the highest of three consecutive measurements from the same ear. In addition, we also assume that any measurement bias was equally distributed between the two groups.

To conclude, routine warming of the patients during caesarean delivery is not widely practiced, although many patients experience peri-operative hypothermia. In the clinical setting, a brief 15-min pre-operative skin surface warming can effectively increase body heat content and reduce the risk of redistribution hypothermia associated with anaesthesia and, in combination with an intra-operative warming procedure, it can reduce the fall in core temperature in patients undergoing spinal anaesthesia. In the present study, the combination of pre-anaesthetic whole body forced-air warming and intra-operative prewarmed i.v. fluid infusion was shown to be effective for preventing hypothermia and shivering in patients undergoing caesarean delivery under spinal anaesthesia.

Acknowledgements relating to this article

Assistance with the study: we thank Hye Sun Lee, Biostatistician, Department of Research Affairs, Yonsei University College of Medicine, for performing the statistical analysis.

Financial support and sponsorship: this research was supported by Hallym University Research Fund, 2017 (HURF-2017-09).

Conflicts of interest: none.

Presentation: this study was presented as a poster presentation at the 95th Annual Scientific Meeting of the Korean Society of Anaesthesiologists, Seoul, Republic of Korea, 8 to 10 November 2018.