Incidence, causes, and effects of intraoperative hypothermia during elective oesophagectomy at a tertiary care centre- An observational study

Abstract

Background and Aims:

Patients undergoing oesophagectomies are at risk of intraoperative hypothermia. The study aims to determine the incidence of hypothermia during oesophagectomy surgeries.

Methods:

This observational bidirectional study evaluated 266 adult patients (175 retrospective and 91 prospective) undergoing elective oesophagectomy at a tertiary care centre between 2019 and 2022. Intraoperative core temperature was monitored hourly using a nasopharyngeal probe. Postoperative complications were recorded until discharge. Logistic regression analysis assessed the association between the incidence of hypothermia and perioperative variables. Those factors found significant in the simple logistic regression were then subjected to multiple logistic regression. A goodness-of-fit and sensitivity analysis was performed.

Results:

The incidence of intraoperative hypothermia (core temperature <36°C) was 90.6% [95% confidence interval (CI): 87, 94]. Nearly 60% of patients were found to be hypothermic before the surgery began. The incidence of clinically significant hypothermia (CSH) (core temperature <35°C) was 32.3% (95% CI: 30, 40). The perioperative factors significantly and independently associated with intraoperative hypothermia were preoperative significant weight loss [odds ratio (OR): 2.03, 95% CI: 1.17, 3.52, P = 0.012] and intraoperative requirement of vasopressors (OR: 2.76, 95% CI: 1.10, 6.93, P = 0.030). We found no association between intraoperative hypothermia and postoperative complications.

Conclusion:

The study recorded a high incidence of intraoperative hypothermia during oesophagectomy surgeries. Hypothermia began in the preoperative period for nearly two-thirds of patients. Factors associated with CSH included preoperative significant weight loss and the intraoperative need for vasopressors. No significant association was found with postoperative complications and discharge.

INTRODUCTION

Intraoperative hypothermia is defined as a drop in the core body temperature below 36°C; however, the clinical effects of hypothermia become pronounced when the core temperature falls below 35°C.[1,2,3,4] Intraoperative hypothermia results in delayed emergence from anaesthesia, arrhythmias, myocardial infarction, increased oxygen demand, shivering, surgical site infections, and extended Intensive Care Unit (ICU) stays. Factors such as extremes of age, malnutrition, sepsis, and endocrinopathies contribute to the risk of developing intraoperative hypothermia.[1,5] Anaesthetic agents, particularly regional anaesthesia techniques, induce hypothermia due to heat loss resulting from vasodilation.

Patients with carcinoma of the oesophagus are prone to develop intraoperative hypothermia. These patients experience dysphagia and significant weight loss. Factors such as hair loss due to neoadjuvant chemotherapy, prolonged surgery duration, the opening of two body cavities, and the use of capnothorax and/or pneumoperitoneum contribute to the development of intraoperative hypothermia. Studies have indicated that age, vasopressor doses, and the use of intraoperative opioids significantly influence hypothermia, as well as an increase in morbidity, mortality, and disease recurrence.[4] Yamasaki et al.[3] found that the incidence of intraoperative hypothermia (bladder temperature <35°C) in 121 patients undergoing elective oesophagectomy was 42%. The study identified intraoperative hypothermia as an independent risk factor for the development of early postoperative complications [odds ratio (OR): 2.57, 95% confidence interval (CI): 1.09, 6.08]. Zhang et al.[6] compared standard and aggressive temperature management methods in oesophagectomy surgeries and found a higher incidence of 2-day myocardial injury (31.4% vs 8.6%) and 3-day severe cardiac arrhythmia (28.6% vs 28.9%) in the standard temperature management group.

We hypothesised that various preoperative and intraoperative factors affect the incidence of intraoperative hypothermia during oesophagectomy and adversely affect postoperative outcomes. The primary objective of this study was to find out the incidence and severity of hypothermia during elective oesophagectomies in adult patients at a tertiary care cancer centre. The secondary objectives were to find the association between intraoperative hypothermia and various perioperative factors and their effects on postoperative complications following oesophagectomies.

METHODS

We conducted a bidirectional observational study at a tertiary cancer institute from November 2019 to June 2022, following the Institutional Ethics Committee (IEC) approval (vide approval number OIEC/3867/2021/0002 dated 8/11/2021) and registration at Clinical Trials Registry-India (vide registration number CTRI/2021/12/038675, dated 16/12/2021; accessible at http://ctri.nic.in/). The cohort comprised 266 adult patients with oesophageal carcinoma undergoing elective oesophagectomy surgery. Prospective data were collected from 91 patients after obtaining written informed consent (21 December 2021 to 7 June 2022). The IEC granted a waiver of consent for the 175 patients recruited retrospectively (19 November 2019 to 16 December 2021). Patients undergoing emergency or concomitant procedures or those with incomplete data were excluded. The study was executed in compliance with the principles of the Declaration of Helsinki and Good Clinical Practice guidelines.

Standard institutional practices were consistently followed for all patients. The patients were transferred from their respective wards to the holding area of the operating theatre 1 hour before surgery. All patients received standard preoperative instructions (2 h nil by mouth for clear liquids) and preparations (including antibiotic prophylaxis and thromboprophylaxis). In the operating theatre, standard monitors (pulse oximeter, electrocardiogram, end-tidal carbon dioxide monitor, and non-invasive blood pressure monitor) were used alongside an invasive arterial line. A mid-thoracic epidural catheter at the T7-T8 or T8-T9 level was placed for all surgeries performed via an open approach, while it was omitted for minimally invasive procedures. Anaesthesia was induced using standard intravenous induction agents (propofol 1.5–2 mg/kg or etomidate 0.15–0.2 mg/kg), followed by a non-depolarising neuromuscular blockade (vecuronium 0.1–0.15 mg/kg or atracurium 0.5–0.6 mg/kg). Controlled ventilation was maintained throughout the surgery, and lung isolation with a double-lumen tube or a bronchial blocker was facilitated during the thoracic part of the surgery.

A nasopharyngeal temperature probe (measured from the tip of the nostril to the tragus) was inserted immediately post-induction to monitor core temperature throughout the surgery and recorded in the case report form every hour. Ambient temperatures in the operating room were monitored every eight hours and maintained between 17 and 20°C. A forced air warming blanket set at 44°C and placed over the non-operative area, warm intravenous fluids administered via a fluid warming system at 41°C, a closed-circuit breathing system with a heat-moisture exchange filter, and appropriate surgical drapes were offered to all the patients in accordance with the institutional policy. Intravenous fluids, analgesics, and antiemetics were administered at the discretion of the operating theatre (OT) anaesthetist. A local anaesthetic infusion was utilised through the epidural catheter for all patients unless complicated by persistent intraoperative hypotension. Intravenous fluid boluses and vasopressor infusion were employed judiciously to maintain mean arterial blood pressure at the consultant anaesthesiologist’s discretion.

Upon completion of the surgery, the trachea was extubated if the patient remained haemodynamically stable, alert, and conscious following adequate neuromuscular recovery. The temperature probe was removed before transferring the patient to the ICU, where patients were monitored for 16 to 20 hours before being shifted to the ward. A research nurse conducted follow-up until discharge to observe postoperative complications. The surgical team documented all postoperative complications and Clavien–Dindo scores in the database, which were subsequently analysed.

The primary outcome was the incidence and severity of intraoperative hypothermia, measured via a nasopharyngeal probe. Hypothermia is generally defined as a core body temperature below 36°C at any point during surgery. However, based on clinical relevance, we defined clinically significant hypothermia (CSH) as a core temperature <35°C. Severity was further classified as mild (34°C–36°C), moderate (32°C–34°C), or severe (<32°C).[7] Missing data were defined as incomplete anaesthesia charting or hourly temperature monitoring not mentioned or not recorded. Preoperative significant weight loss was defined as a loss of more than 10% of pre-disease body weight in 6 months. Delayed extubation was defined as any unplanned mechanical ventilation resulting in the inability to extubate the trachea immediately after surgery. Severe preoperative dysphagia was defined as difficulty in swallowing solid and liquid food, and requiring Ryle’s tube feeding in the preoperative period. Postoperative pulmonary complications were defined as any pulmonary ailment that required an intervention for treatment (e.g., antibiotics, additional intercostal drain, assisted ventilation, etc.).

As a feasibility pilot project, data for 198 retrospective and 100 prospective patients were collected. After eliminating patients with missing data and those with inoperability, we analysed the data of 266 patients (175 retrospectively and 91 prospectively).

The analysis utilised Statistical Package for Social Sciences (SPSS) version 25 (IBM Corp., released in 2017; IBM SPSS Statistics for Windows, Version 25). Continuous variables [age, height, weight, body mass index (BMI), duration of surgery, quantity of intravenous fluids, postoperative hospital stay] were expressed as mean and standard deviation (SD), while categorical variables were represented as frequencies and percentages. Contingency tables were created for perioperative variables in relation to the occurrence of CSH. Categorical data (comorbidities, occurrence of postoperative complications) were compared using the Chi-square or Fisher’s exact test. For normally distributed continuous variables (age, BMI, height, weight), means were compared using the independent T-test. For variables that were not normally distributed (intraoperative blood loss, postoperative hospital stay), median with interquartile range (IQR) were compared using the Mann–Whitney U test. Variables with a P < 0.05 in univariate analysis were included in multivariate analysis. ORs with 95% CI were calculated. A goodness-of-fit and sensitivity analysis was performed to test the robustness of the model.

RESULTS

Figure 1 shows the recruitment of patients. The demographic details of the population are given in Table 1.

Figure 1.

Recruitment of patients

Table 1.

Demographics of the population

Parameter	Value
Age (Mean (SD)) yrs. (n=266)	54.7 (11.27)
Gender (Male: Female) (n=266)	59.4%: 40.6% (152:114)
BMI (Mean (SD)) Kg/m2 (n=266)	21.9 (5.65)
ASA status (I: II: III) (n=266)	45.9%: 48.1%: 4.1% (122:128:16)
Approach (open: laparoscopic: robotic) (n=266)	64.7%: 31.2%: 4.2% (172: 83: 11)
ECOG (0:1:2:3) (n=266)	17.3%: 76.7%: 4.9%: 1.1% (46: 204: 13: 3)
Types of oesophagectomy- TTE/THE/ILE/LTA	68.8%:2.3%:15.4%:13.5% (183:6:41:36)
Duration of surgery (min) (Median, Q1, Q3) (n=266)	460, 390, 540
Blood loss (mL) (Median, Q1, Q3) (n=266)	400, 300, 600
Postoperative hospital stay (days) (Median, Q1, Q3) (n=266)	11, 9, 15

Data expressed as mean (standard deviation) or percentages. SD=standard deviation, BMI=Body Mass Index, ASA=American Society of Anesthesiologists, ECOG=Eastern Cooperative Oncology Group, TTE=Transthoracic Oesophagectomy, THE=Transhiatal Oesophagectomy, ILE=Ivor Lewis Oesophagectomy, LTA=Left Thoracoabdominal Approach, Q1=lower quartile 25^th percentile, Q3=upper quartile 75^th percentile)

The incidence of hypothermia (a drop in core body temperature below 36°C at any time during surgery) was 90.6% (n = 241) (95% CI: 0.87, 0.94). Mild hypothermia was noted in 79.3% of the patients, moderate in 10.9%, and severe in 0.4% of the patients. CSH was noted in 32.3% of the patients (n = 86) (95% CI: 0.3, 0.4). Preoperative hypothermia was noted in 60.9% of the patients (n = 162 at T < 36°C), and 9.3% of patients (n = 24) had CSH before the start of the surgery. At the end of surgery, 49.2% of patients (n = 131) were hypothermic, while temperatures below 35°C were observed in 16.5% of patients (n = 44). CSH during surgery lasted for a median duration of 240 minutes, whereas the median duration of surgery was 460 minutes (range: 150–660 minutes, IQR: 390–540 minutes). There was no pattern noticed in the development of intraoperative hypothermia with the time spent in the OT [Figure 2].

Figure 2.

Temperature trends over time

Simple logistic regression identified the following variables which were significantly associated with CSH: a BMI <18.5 kg/m² (OR: 2.07, 95% CI: 1.17, 3.65, P = 0.012), the presence of severe preoperative dysphagia (OR: 2.12, 95% CI: 1.15, 3.93, P = 0.016), preoperative significant weight loss (OR: 2.11, 95% CI: 1.22, 3.63, P = 0.007), and requirement of vasopressor infusion during the surgery (OR: 2.96, 95% CI: 1.19, 7.33, P = 0.019) [Table 2]. Multiple logistic regression showed that preoperative significant weight loss (OR: 2.03, 95% CI: 1.17, 3.52, P = 0.012) and the requirement of intraoperative vasopressor infusion (OR: 2.76, 95% CI: 1.10, 6.93, P = 0.030) were independently associated with the occurrence of CSH [Table 3]. The discrimination metrics for multivariable modelling (P < 0.15) revealed the area under the curve (AUC) value as 0.59 (95% CI: 0.52, 0.67) [Figure 3]. Hosmer and Lemeshow Test was performed to assess the goodness-of-fit (P = 0.772). We further conducted a sensitivity analysis for retrospective data (AUC 0.55, 95% CI: 0.41, 0.68) wherein only preoperative weight loss was found to be significant (P = 0.021, OR: 2.14, 95% CI: 1.22, 4.07) [Figure 4].

Table 2.

Simple logistic regression analysis of the association between clinically significant hypothermia and perioperative factors

Perioperative Variables		Clinically significant hypothermia (temperature <35°C)		Odds ratio (95% CI)	P
		YES	NO
Age		56.85 (11.42)	53.66 (11.08)	1.02 (0.99, 1.04)	0.056
Surgical Approach	Open	56 (65.1%)	116 (64.4%)	Reference	-
	VATS + Laparoscopic	29 (33.7%)	54 (30.0%)	1.1 (0.6, 1.9)	0.705
	Robotic	1 (1.2%)	10 (5.6%)	0.2 (0.0, 1.7)	0.138
BMI	≤18.5kg/m2	31 (36%)	42 (23.3%)	2.07 (1.17, 3.65)	0.012
	≥18.5kg/m2	55 (64%)	138 (76.7%)	Reference
Preoperative significant weight loss	Yes	36 (41.9%)	44 (24.4%)	2.11 (1.22, 3.63)	0.007
	No	50 (58.1%)	136 (75.6%)	Reference
Severe dysphagia	Yes	25 (29.1%)	28 (15.6%)	2.12 (1.15, 3.93)	0.016
	No	61 (70.9%)	152 (84.4%)	Reference
Use of intraoperative vasopressor infusion	Yes	12 (14%)	9 (5%)	2.96 (1.19, 7.33)	0.019
	No	74 (86%)	171 (95%)	Reference
Arrhythmia during or after the surgery	Yes	3 (3.5%)	13 (7.2%)	2.15 (0.59, 7.76)	0.282
	No	83 (96.5%)	167 (92.8%)	Reference
Surgical Site Infection	Yes	2 (2.3%)	10 (5.6%)	2.47 (0.53, 11.53)	0.348
	No	84 (97.7%)	170 (94.4%)	Reference
Myocardial Ischaemia after the surgery	Yes	0 (0.0%)	3 (1.7%)	0.98 (0.96,1.01)	0.553
	No	86 (100.0%)	177 (98.3%)	Reference
Postoperative Pulmonary Complication	Yes	18 (20.9%)	36 (20.0%)	0.94 (0.50, 1.78)	0.860
	No	68 (79.1%)	144 (80.0%)	Reference
Postoperative Mechanical Ventilation	Yes	17 (19.8%)	33 (18.3%)	0.91 (0.47, 1.74)	0.779
	No	69 (80.2%)	147 (81.7%)	Reference

Data expressed as mean (standard deviation) or number of patients (percentages). BMI=Body Mass Index, VATS=Video-assisted thoracoscopic surgery

Table 3.

Multiple logistic regression for the significant association between the variable and the incidence of hypothermia

Variable	Categories	Odds Ratio	95% Confidence Interval	P
Preoperative weight loss	Yes	2.03	1.17-3.52	0.012
	No	Ref
Intraoperative use of vasopressors	Yes	2.76	1.10-6.93	0.030
	No	Ref

Figure 3.

ROC curve for goodness of fit. ROC = Receiver Operating Characteristics

Figure 4.

ROC curve for validation of the retrospective model based on prospective data. ROC = Receiver Operating Characteristics

In our study, 54.7% of patients experienced postoperative complications, with 32.41% (n = 47) of them being hypothermic. No individual complication was significantly associated with any grade of intraoperative hypothermia [Table 2]. Delayed extubation (OR: 0.29, 95% CI: 0.12, 0.67, P = 0.009) was significantly related to moderate hypothermia. In-hospital mortality (3.33%, P = 0.182) did not show significant differences between hypothermic and non-hypothermic patients.

DISCUSSION

This study recorded a high incidence of hypothermia (90.6%) among patients undergoing oesophagectomy. Nearly one-third of the patients experienced clinically significant hypothermia, and more than 60% were hypothermic before the surgery commenced. Preoperative significant weight loss and the intraoperative requirement for vasopressors were notably associated with CSH. We found no significant association between any grade of hypothermia and postoperative complications.

The incidence of intraoperative hypothermia varies depending on the surgical procedure and the definition of hypothermia. Wongyingsinn M et al.[2] reported the rate of hypothermia (<36°C) in patients receiving general anaesthesia as high as 73.5% (95% CI: 9.7, 14.3), whereas a multicentric study conducted by Yi et al.[7] found the incidence of intraoperative hypothermia (measured via a tympanic membrane temperature probe) as 44.3% in Video-Assisted Thoracoscopic Surgeries. Yamasaki et al.[3] found that the incidence of hypothermia (defined as core temperature <35°C measured via a bladder temperature probe) was 42% in 121 patients undergoing radical oesophagectomies. Amin and Ravichandran[1] found that the incidence of hypothermia (defined as core temperature <35°C) was 22%, of whom 3.1% experienced moderate (32°C–34°C) and 0.2% severe (<32°C) hypothermia in cancer surgeries. Stuart et al.[8] graded hypothermia as grade 1 (T < 36°C - 93%), grade 2 (T < 35.5°C - 62%), and grade 3 (T < 35°C - 31.9%) in 313 patients undergoing robotic thoracic surgery. Although different studies recorded core body temperature by different temperature probes, the incidence of CSH of 32.3% in our research aligned with the findings from the studies of Yamasaki et al.[3] and Stuart et al.[8]

The reasons for a high incidence of intraoperative hypothermia in oesophagectomy surgeries could be long-duration complex surgeries involving the opening of two body cavities (thoracic and abdominal), change of patients’ position during the surgery, a limited area available for applying warming blankets, and the presence of dysphagia leading to malnutrition and low BMI. This loss of muscle and subcutaneous fat makes the patients prone to developing intraoperative hypothermia.[1,9,10,11] In our study, nearly 30% of patients had a significant weight loss, and 24% had a BMI < 18.5 kg/m². Both factors were significantly correlated with the incidence of intraoperative hypothermia. Therefore, improving a patient’s nutrition before surgery may help prevent hypothermia.

Nearly 60% of patients in our study were hypothermic before surgery, likely due to a lack of warming measures in the preoperative holding area. Pre-induction procedures like the placement of an epidural catheter and attachment of monitors increase the heat loss in the OT. It is further compounded by the vasodilatory effects of anaesthetic agents. Additionally, controlled ventilation increases insensible heat loss. The EUPEMEN Protocol emphasises that pre-warming measures play a crucial role in maintaining intraoperative temperatures in oesophagectomies.[12] A systematic review by Simegn et al.[13] suggested that warming methods should be applied 1 to 2 hours before the induction of anaesthesia. Okada et al.[14] demonstrated that using a warming gown in the preoperative area helped maintain normothermia during thoracoscopic oesophagectomy.

There was an increased need for vasopressors noted in hypothermic patients. Yamasaki et al.[3] found that the doses of opioids and vasopressors were significantly higher in hypothermic patients as compared to normothermic patients. General anaesthetics impair hypothalamic thermoregulation, lowering the vasoconstriction threshold. This results in peripheral vasodilation, thus decreasing systemic vascular resistance and necessitating the use of vasopressors to maintain mean arterial pressure. Additionally, severe hypothermia can directly impair myocardial function, further contributing to the need for haemodynamic support.

Intraoperative hypothermia can lead to cardiovascular and infectious complications. Zhang et al.[6] found that the patients with standard temperature management during oesophagectomy had a higher incidence of myocardial injury, arrhythmias, and postoperative 12-hour metabolic acidosis compared to those with aggressive management. However, a large multicentric trial comparing aggressive (target 37°C) with standard temperature control (target 35.5°C) in non-cardiac surgeries observed no difference in the incidence of major cardiovascular outcomes.[15] Lu X et al.[16] demonstrated that aggressive warming, while not reducing pulmonary complications, shortened chest drainage time in elderly oesophagectomy patients. While our research found that hypothermia was associated with a greater need for vasopressors and prolonged mechanical ventilation in patients with moderate hypothermia, we did not find any association with other postoperative complications.

We acknowledge several limitations of the study. Although the study’s design was bi-directional, we were able to retrieve data reliably as the anaesthesia charts were uploaded to the electronic medical records. As a single-centre study, the outcomes may be influenced by institutional practices. Key unmeasured variables include pre-operative surface temperature monitoring in the holding area and reliable ambient OT temperature monitoring. Furthermore, we did not assess the markers of cardiac injury, acid–base balance, or total intraoperative fluids. Statistically, while goodness-of-fit and sensitivity analysis were performed, a calibration plot could not be provided due to a lack of significant factors. Finally, the patient’s subjective experiences during the preoperative or postoperative periods were not evaluated.

CONCLUSION

The study recorded a high incidence of intraoperative hypothermia during oesophagectomy surgeries. Hypothermia commenced in the preoperative period for nearly two-thirds of patients. Preoperative significant weight loss and the intraoperative requirement for vasopressors were independently associated with the incidence of clinically significant hypothermia. However, no significant association was found with postoperative complications until discharge.

Presentation at conferences/CMEs and abstract publication

Paper presentation at AAGBI, London, in September 2023. E-poster presentation at WFSA WCA 2024, Singapore.

Study data availability

De-identified data may be requested with reasonable justification from the authors (email to the corresponding author) and shall be shared after approval as per the authors’ institution’s policy.

Disclosure of use of artificial intelligence (AI)-assistive or generative tools

The authors confirm that no AI tools or language models (LLMs) were used in the writing or editing of the manuscript, and no images were manipulated using AI.

Declaration of use of permitted tools [choose any ONE]

Nil.

Authors contributions

SYP was involved in Concepts, Design, Definition of intellectual content, literature search, conduct of cases, data acquisition, data analyses, manuscript preparation, editing, review and approval, SK was involved in literature search, conduct of cases, data acquisition, data analyses, manuscript preparation, editing, MS was involved in concepts, design, definition of intellectual content, editing, review and approval; PR was involved in concepts, design, definition of intellectual content, editing, review and approval.

Supplementary material

Nil.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

Nil.