Development and validation of a risk prediction scale for hypothermia during cesarean section: A prospective study

Highlights

• • What is already known
• •Hypothermia has an adverse effect on post-obstetric recovery and affects neonatal outcomes.
• •The patient indications for and the timing, methods, and duration of active heat preservation during temperature management in this population remain unclear. The timely identification of parturients at a high risk for hypothermia during cesarean section is crucial for providing personalized insulation measures.
• •Currently, there is no specific scale for predicting hypothermia in parturients during cesarean section.
• • What this paper adds
• •The predictive factors in the model developed in this study were temperature after anesthesia, body mass index, subclinical hypothyroidism during pregnancy, intraoperative fluid loss, amount of intraoperative abdominal cavity irrigation, duration of active insulation, and room temperature before operation.
• •This predictive model can identify women who are at a high risk of hypothermia during cesarean section.

Abstract

Background

Evidence shows that active insulation can reduce the incidence of hypothermia during cesarean section; however, compliance is lower than recommended. Moreover, several aspects of temperature management via active heat preservation remain unclear, including patient indications, timing, methods, and duration. Therefore, promptly identifying parturients at a high risk for hypothermia during cesarean section is crucial.

Objective

To develop and validate a scale for predicting hypothermia in parturients during cesarean section.

Design

Prospective study.

Setting

Three grade A hospitals in Hunan Province, China.

Participants

The prediction scale was developed based on data from 369 parturients who underwent cesarean section from July 2018 to January 2019. Inclusion criteria were as follows: cesarean section under lumbar anesthesia, epidural anesthesia, or combined lumbar and epidural anesthesia; voluntary participation in the study and completion of the informed consent form; age >18 years.

Methods

Univariate and multivariate analyses were used to determine factors influencing hypothermia and establish the predictive model for hypothermia risk during cesarean section. The Hosmer–Lemeshow test was used to determine the goodness of fit of the prediction tool, and the area under the receiver operating characteristic curve was used to determine the predictive ability of the proposed scale. The cutoff value of the prediction scale was determined according to the Youden index.

Results

In the logistic regression prediction model, the Hosmer–Lemeshow goodness-of-fit test yielded a p-value of 0.425. The area under the receiver operating characteristic curve was 0.888. The model exhibited a good fitting effect and discriminant validity. Total risk scores for hypothermia ranged from 0 to 11. A score of 7 was used as the diagnostic cutoff value. Parturients during the operation who had total scores of ≥7 and <7 were considered the high-risk and low-risk groups, respectively. The area under the receiver operating characteristic curve for the scale was 0.891. The authenticity evaluation indicated that the incidence of hypothermia was significantly higher in the high-risk group than in the low-risk group.

Conclusions

The risk prediction scale developed in this study exhibits moderately predictive efficiency, sensitivity, and specificity for identifying parturients at high risk of hypothermia during cesarean section. Implementing this scale in clinical practice may help to decrease the incidence of hypothermia in such patients.

Tweetable

 

abstract

This new predictive model can identify women who are at a high risk of hypothermia during cesarean section.

1. Background

Intraoperative hypothermia, defined as a core body temperature <36°C, is a common complication of surgery, and 20–90% of patients undergoing surgery are prone to hypothermia (Bartosz and Małgorzata, 2014; Duff et al., 2014; Torossian, 2008). Cesarean section is widely regarded as a safe method for addressing obstructed labor, and can improve maternal and infant outcomes. According to the China's National Maternal and Child Health Statistics data, the rate of cesarean section in China is high (36.7%) (Li et al., 2020). Recent implementation of the two-child and three-child policies in China has resulted in an increase in the number of older women giving birth. With the continuous development of perinatal medicine and improvements in surgery and anesthesia technology, medical staff may recommend cesarean section for older patients based on the status and risk of the pregnancy to ensure maternal and child safety (Liu et al., 2017). The main indications for caesarean section included previous caesarean section (24.2%), cephalopelvic disproportion (22.6%), fetal distress (20.5%), and breech or other abnormal presentation (12.5%). Indeed, for women with evident indications, cesarean section can reduce maternal and neonatal mortality (Lumbiganon et al., 2010).

During cesarean section, spinal anesthesia affects the central thermoregulation and blocks local sensory, motor, and sympathetic nerves, thus inhibiting the feedback mechanism of thermoregulation. Meanwhile, room temperature, surgical heat loss, and fluid expansion also exert varying degrees of influence (Torossian, 2008). In other countries, the incidence of perioperative hypothermia in women undergoing cesarean section under lumbar anesthesia ranges from 32–80% (Chakladar et al., 2014; Hess et al., 2005). In China, the reported incidence rates were 33.3–56.6% and 47.5% for hypothermia and shivering during cesarean section, respectively (Li et al., 2017; Mo and Qiu, 2017). Moreover, the parturient is generally awake during a cesarean section, which leads to a low comfort level and may cause a series of complications, including shivering, surgical site infection, increases in blood loss (the main cause of maternal death worldwide), and hypothermia in the newborn, among others (Cobb et al., 2016; Jun et al., 2019). Munday et al. (2018) have confirmed that although the incidence of hypothermia during cesarean section is high, it can be reduced through effective prevention. Sultan et al. (2015) suggested that medical staff should provide active thermal insulation to attenuate changes in body temperature and improve thermal comfort in parturients undergoing cesarean section; however, clinical compliance remains low, and evidence to support this strategy is lacking. Some scholars have also proposed that, although active thermal insulation can reduce the incidence of hypothermia during cesarean section and improve maternal comfort, hypothermia and shivering may still occur, highlighting the need for appropriate thermal insulation (Cobb et al., 2016).

International scholars believe that the maintenance of body temperature in women after cesarean section is determined by various complex factors, including heat loss, room temperature, infusion, epidural or spinal anesthesia, damage to the thermoregulatory mechanism, and body temperature redistribution (Torossian et al., 2015). Heat loss during cesarean section under spinal anesthesia occurs through two mechanisms. First, in the first hour after anesthesia, maternal vasodilation below the level of sensory block leads to heat loss. This is followed by a decrease in the core-peripheral temperature gradient and subsequent blood redistribution (Hynson et al., 1991; Matsukawa et al., 1995). Intraspinal anesthesia also results in vasoconstriction above the block area and a decrease in the shivering threshold (by approximately 0.5°C), although vasoconstriction above the level of the block does not prevent a decrease in core temperature (Kurz et al., 1993; Ozaki et al., 1994; Saito et al., 1998). The risk factors for hypothermia in surgical patients are reported in the guidelines for hypothermia prevention developed by the United States Association of Perioperative Registered Nurses, the American Society of Peri Anesthesia Nurses, and Britain's National Institute for Health and Care Excellence; however, these guidelines focus on intraoperative risk factors and levels of evidence, and they do not consider risk factors for hypothermia according to the type of surgery. Thus, several groups have highlighted the need to develop guidelines and prediction models based on the study of risk factors in patients with perioperative hypothermia (Association of Operating Room Nurses Recommended Practices Committee, 2007; Hooper et al., 2009). Only one previous predictive model of hypothermia during cesarean section has been developed (Desgranges et al., 2017). Although the authors also identified factors protective against hypothermia during cesarean section, they did not construct a scale for predicting hypothermia risk during cesarean section to distinguish patients at high and low risk of intraoperative hypothermia.

The 2030 sustainable development goals adopted by all United Nations member states in 2015 include goals related to the promotion of maternal health. Specifically, the goals mention that it is necessary to reduce the maternal mortality rate for live births, improve early-warning systems, and reduce health risks in developing countries (The United Nations, 2015). However, no models for predicting hypothermia during cesarean section have been designed. Thus, this study aimed to develop and validate a prediction scale for assessing the risk of hypothermia during cesarean section.

2. Methods

2.1. Design and participants

In this prospective study, a list of grade A hospitals in Hunan Province was obtained through an online search, and three hospitals were randomly selected from the list using the random number table method (serial numbers A, B, and C, respectively). Pregnant women who underwent cesarean section under lumbar anesthesia, epidural anesthesia, or combined lumbar and epidural anesthesia from July 2018 to January 2019 were randomly selected from among the three hospitals and were divided into the hypothermia group (body temperature at the end of surgery < 36°C) and the normothermia group, according to the diagnostic criteria for low temperature and hypothermia (Hasankhani et al., 2007). After obtaining approval from the hospital managers, we collected information for parturients who underwent elective cesarean delivery using the hospital electronic medical record system. Preoperative visits were conducted for parturients who met the inclusion criteria. We explained the purpose, method, content, and meaning of this study prior to obtaining written informed consent from participants. The researchers who conducted the preoperative visits differed from those who collected the data.

To estimate the study's sample size, we followed the logistic regression sample size calculation method, which requires 15 to 20 times the number of influencing factors. Considering a 15% loss of sample, we ultimately required 360 parturients undergoing cesarean section.

2.2. Eligibility criteria

The inclusion criteria were as follows: (1) voluntary participation in the study and completion of the informed consent form; (2) age >18 years; and (3) cesarean section. The exclusion criteria were as follows: (1) impaired thermoregulation; (2) receipt of mild hypothermia treatment; (3) body temperature >38°C three days before surgery; and 4) a definite diagnosis of sweat gland dysfunction. The elimination standards were as follows: (1) change from intraoperative anesthesia to general anesthesia; (2) major bleeding or other critical condition changes during the investigation; and (3) >5% of data missing from the questionnaire or an inability to trace important data to the source (signs and symptoms in pregnant women at the time).

2.3. Data collection and sources of information

Among the primary outcome variables, body temperature was monitored using an infrared ear thermometer when the parturient entered the operating room, after anesthesia, before skin incision, and at the end of the operation. An infrared tympanic ear thermometer (Omron TH839S; accuracy ± 0.2°C) was used for temperature measurements. After inspecting and cleaning the external auditory canal cerumen, we obtained and averaged two measurements in the same ear to minimize measurement error at each time point. Study nurses were specifically trained in thermometry based on the manufacturer's instructions, and binocular vision was normal in all nurses. The pinna was pulled back, following which the probe of the infrared thermometer was inserted into the external auditory canal and directed forward. The probe was kept in the same position until the researcher heard a beep. The same thermometer was used for all parturients and maintained and calibrated in accordance with the manufacturer's guidelines.

For the independent variables, we collected general information, such as maternal age, gestational age, previous cesarean section, and pregnancy complications.

A self-administered questionnaire on risk factors for hypothermia during cesarean section was used to collect the following information:

• •During the preoperative visit on the day before surgery, we collected data related to maternal age; gestational age; previous cesarean section; body mass index; and pregnancy complications, such as hypertension, gestational diabetes mellitus, gestational hyperthyroidism, premature membrane rupture, subclinical hypothyroidism during pregnancy, intrahepatic cholestasis during pregnancy, eclampsia, and other complications.
• •Prior to anesthesia, we collected data related to the type of surgery, duration of fasting and water deprivation, body temperature, anesthesia grade, method and level of anesthesia, and room temperature before operation.
• •During surgery (from skin incision to the end of the operation), we collected data related to operative time, blood transfusion volume, fluid infusion volume, amount of wound washing, amount of abdominal cavity irrigation, and intraoperative fluid loss including urine volume, amniotic volume, and blood loss. As we could not distinguish amniotic fluid loss from blood loss, the data were collected together. The vast majority of amniotic fluid and blood loss was determined based on the volume in the liquid waste collector and disposable sterile sheet with a liquid waste collection bag used during surgery. Additionally, the nurses squeezed out volumes of fluid or blood contained in gauze and saline pads at the end of the surgery, but the remaining blood volume contained in them were not weighed. In addition, factors associated with temperature management, such as measurement, quantity, and duration of active insulation, were also recorded.
• •After surgery (after suturing the skin), we collected data related to postoperative body temperature.

2.4. Statistical analysis

Data were analyzed using the genuine authorized version of SPSS 26.0 purchased by Central South University and authorized by IBM corporation. First, the Kolmogorov–Smirnov test was used to test the normality of continuous variables, which were expressed as the mean ± standard deviation or median. The measurement data that did not conform to the normal distribution were represented using the median and interquartile range, and the relevant comparisons between groups were performed using the rank-sum test. Enumeration data were described in terms of cases and percentages, and the relevant comparisons between groups were performed using the Chi-square test or Fisher's exact test.

A logistic regression equation with backward elimination was used to construct a risk prediction model for hypothermia during cesarean section. The score of each variable included in the multivariate logistic regression model was calculated based on the regression coefficient (β) value of that variable. The scale for predicting hypothermia risk during cesarean section was established according to the variable scores and analyzed using a receiver operating characteristic curve. The cutoff value for classifying parturients into high-risk and low-risk hypothermia groups was determined using the Youden index.

The goodness-of-fit of the prediction tool was determined using the Hosmer–Lemeshow test. P-values > 0.05 indicated appropriate goodness-of-fit, and there was no significant difference between the predicted and observed values.

The discriminant accuracy and predictive ability of the prediction tool were evaluated according to the area under the receiver operating characteristic curve. The curve was used to evaluate the effects of the diagnostic experiment, with larger area values indicating greater diagnostic value. The area under the curve values were classified as follows: < 0.5 (no clinical significance), 0.5–0.7 (low clinical significance), 0.7–0.9 (medium clinical significance), and > 0.9 (high clinical significance).

2.5. Authenticity evaluation

The parturients in the database were divided into low-risk and high-risk groups. The incidence of intraoperative hypothermia was calculated in both groups, and the difference was examined using the Chi-square test. P-values < 0.05 were considered significant.

2.6. Ethical and legal considerations

The study was approved by the local ethics committee (2018028) and was conducted under the supervision of the hospital ethics committee to protect the rights and interests of participants. Prior to data collection, a letter of introduction was submitted to obtain permission from hospital managers, and written informed consent was obtained from all participants to ensure that each participant understood the purpose, content, risks, and benefits of the research. Each patient participated in the study voluntarily and could exit at any time. All data were used only in this study, and no personal information is disclosed.

3. Results

Altogether, 369 parturients participated in this study. Maternal age ranged from 20 to 49 (32.1±4.6) years, while gestational age ranged from 202 to 290 (267.0±15.5) days. There were 168 (45.5%) patients undergoing their first cesarean section and 201 (54.5%) undergoing their second cesarean section. There were 28 patients (7.6%) with hypertension during pregnancy, 106 patients (28.7%) with gestational diabetes, and 21 parturients (6.0%) with premature membrane rupture. A total of 83 (22.5%) parturients developed hypothermia during cesarean section and were thus included in the hypothermia group. Meanwhile, 286 (77.5%) parturients did not develop hypothermia during cesarean section and were assigned to the normothermia group.

The variables associated with hypothermia during cesarean section (p < 0.05) were selected for the univariate analysis, as follows: temperature upon entering the operating room, after anesthesia, and before skin incision; body mass index; intraoperative blood transfusion volume; intraoperative fluid infusion volume; intraoperative fluid loss; amount of intraoperative abdominal cavity irrigation; active insulation; duration of active insulation; quantity of insulation measures; infusion heating; use of an air-forced insulation device; room temperature before and during the operation; and subclinical hypothyroidism during pregnancy (Table 1).

Table 2.

Multivariate analysis of factors influencing hypothermia during cesarean section.

Predictor	Coeff.	p-value	Odds ratio (95% CI)
Constant	−5.700	<0.001	0.003
Temperature after anesthesia (℃)	2.828	<0.001	16.912 (8.076–35.415)
Body mass index	0.763	0.044	2.144 (1.019–4.511)
Subclinical hypothyroidism during pregnancy	0.958	0.039	2.607 (1.048–6.487)
Intraoperative fluid loss (ml)	1.035	0.004	2.818 (1.402–5.659)
Amount of intraoperative abdominal cavity irrigation (ml)	1.085	0.006	2.958 (1.372–6.378)
Duration of active insulation (min)	0.966	0.013	2.628 (1.230–5.618)
Room temperature before operation (℃)	1.488	<0.001	4.428 (2.040∼9.613)

CI, confidence interval.

Total scores on the risk prediction scale for hypothermia during cesarean section ranged from 0 to 11 (Table 3).

Table 3.

Assessment of hypothermia risk during cesarean section.

Predictor		Beta coeff.	p-value	Odds ratio (95% CI)	Score
Temperature after anesthesia (°C)	<36.8	2.828	0.000	16.912 (8.076–35.415)	4
	≥36.8				0
Body mass index	≤25.05	0.763	0.044	2.144 (1.019–4.511)	1
	>25.05				0
Subclinical hypothyroidism during pregnancy	Yes	0.958	0.039	2.607 (1.048–6.487)	1
	No				0
Intraoperative fluid loss (ml)	≤980				0
	>980	1.035	0.004	2.816 (1.402–5.659)	1
Amount of intraoperative abdominal cavity irrigation (ml)	≤775				0
	>775	1.085	0.006	2.958 (1.372–6.378)	1
Duration of active insulation (min)	≤33	0.966	0.013	2.628 (1.230–5.618)	1
	>33				0
Room temperature before operation (°C)	≤23	1.488	0.000	4.428 (2.040–9.613)	2
	>23				0

CI, confidence interval.

The Hosmer–Lemeshow test yielded a Chi-square value of 8.090 and a p-value of 0.425. The regression equation indicated appropriate goodness-of-fit (Table 4).

Table 4.

Hosmer-Lemeshow test of hypothermia risk during cesarean section.

Step	Chi-square value	Degree of freedom	p-value
1	10.288	8	0.249
2	7.342	8	0.500
3	10.460	8	0.234
4	11.443	8	0.178
5	8.866	8	0.354
6	7.340	8	0.500
7	8.141	8	0.420
8	7.416	8	0.492
9	6.227	8	0.622
10	6.243	8	0.620
11	3.795	8	0.875
12	3.730	8	0.881
13	5.418	8	0.712
14	8.090	8	0.425

Table 1.

Single-factor analysis of factors influencing hypothermia during cesarean section.

Variable	Hypothermia group (n=83)	Normothermia group (n=286)	Chi-square /Z value	p-value
Maternal age (year)a)	32.0 (29.0, 36.0)	31.0 (29.0, 35.0)	-0.944b	0.345
Gestational age (day)a)	271.0 (260.0, 275.0)	270.0 (263.0, 274.0)	-0.185b	0.853
Temperature upon entering the operating room (℃)a	36.7 (36.6, 36.9)	36.9 (36.8, 37.1)	-6.825b	<0.001
Temperature after anesthesia (℃)a	36.6 (36.5, 36.7)	36.9 (36.7,37.0)	-9.263b	<0.001
Temperature before skin incision (℃)a	36.5 (36.3, 36.6)	36.8 (36.6, 37.0)	-9.659b	<0.001
Body mass indexa	26.4 (24.8, 28.1)	27.1 (25.3, 29.3)	-2.470b	0.013
Duration of fasting (h)a	12.0 (7.0, 15.0)	11.0 (7.5, 14.0)	-0.877b	0.381
Duration of water deprivation (h)a	11.0 (9.0, 14.0)	8.3 (6.0, 11.0)	-0.630b	0.529
Intraoperative fluid infusion volume (ml)a	800.0 (600.0, 1200.0)	800.0 (600.0, 1100.0)	-0.036b	0.971
Amount of intraoperative wound washing (ml)a	300.0 (200.0, 400.0)	300.0 (200.0, 300.0)	-2.036b	0.042
Amount of intraoperative abdominal cavity irrigation (ml)a	200.0 (0.0, 1000.0)	200.0 (0.0, 500.0)	-2.629b	0.009
Intraoperative fluid loss (ml)a	1100.0 (900.0, 1300.0)	950.0 (750.0, 1250.0)	-2.766b	0.006
Operative time (min)a	50.0 (42.0, 62.0)	50.0 (42.0, 60.0)	-0.513b	0.608
Duration of active insulation (min)a	0.0 (0.0, 40.0)	40.0 (0.0, 65.0)	-3.486b	<0.001
Anesthesia method			5.787c	0.055
Spinal anesthesia	57 (68.7)	154 (53.9)
Epidural anesthesia	2 (2.4)	11 (3.8)
Combined lumbar epidural anesthesia	24 (28.9)	121 (42.3)
Planned repeat cesarean delivery	50 (60.2)	151 (52.8)	1.437c	0.261
Type of surgery			1.470c	0.314
Emergency operation	3 (3.6)	21 (7.3)
Selective operation	80 (96.4)	265 (92.7)
Active insulation	26 (20.5)	155 (54.9)	13.464c	<0.001
Quantity of insulation measures			14.872c	0.001
None	57 (68.7)	131 (45.8)
One	10 (12.0)	39 (13.6)
Two	16 (19.3)	116 (40.6)
Infusion heating	17 (20.5)	118 (41.3)	11.970c	<0.001
Air-forced insulation device	23 (27.7)	152 (53.1)	16.693c	<0.001
Room temperature before operation (°C)			15.962c	<0.001
≤23	65 (78.3)	154 (53.8)
>23	18 (21.7)	132 (46.2)
Room temperature during operation (°C)			12.080c	0.001
≤23	64 (77.1)	160 (55.9)
>23	19 (22.9)	126 (44.1)
Intraoperative blood transfusion volume (ml)	9 (10.8)	12 (4.2)	5.297c	0.030
Hypertension	3 (3.6)	25 (8.7)	2.411c	0.158
Gestational diabetes mellitus	19 (22.9)	87 (30.4)	1.781c	0.215
Subclinical hypothyroidism during pregnancy	19 (8.4)	31 (16.8)	7.977c	0.010
Gestational hyperthyroidism	1 (1.2)	11 (3.8)	1.426c	0.313
Premature membrane rupture	3 (3.6)	18 (6.3)	0.860c	0.432
Other complications	18 (21.7)	60 (21.0)	0.019c	0.889

a) d, M (P₂₅, P₇₅); b) Z value; c) Chi-square value.

Based on these variables, we performed a multivariate analysis using the backward stepwise procedure. The final model was established by including temperature after anesthesia, body mass index, subclinical hypothyroidism during pregnancy, intraoperative fluid loss, amount of intraoperative abdominal cavity irrigation, duration of active insulation, and room temperature before surgery (Table 2).

The total score of the risk prediction scale for hypothermia during cesarean section and the occurrence of hypothermia were considered the test and state variables in the receiver operating characteristic curve analysis, respectively. The analysis yielded an area under the curve of 0.891, p-value of 0.000, and 95% confidence interval of 0.854–0.929, indicating that the scale had good discriminative validity and could accurately distinguish parturients with intraoperative hypothermia (Fig. 1).

Fig. 1.

The receiver operating characteristic curve for the total score of the risk prediction scale for hypothermia during cesarean section.

The cutoff value of the risk prediction scale for hypothermia during cesarean section was determined according to the Youden index. The score corresponding to the maximum Youden index was taken as the cutoff value (Youden index = sensitivity + specificity −1) and used to classify parturients into the low-risk and high-risk groups. When the maximum Youden index was 0.592, the corresponding score was 6.5, indicating that a total score ≥ 6.5 would be considered positive. The cutoff value was determined to be 7, meaning that patients with total scores ≥7 were classified into the high-risk group, while those with total scores <7 were classified into the low-risk group (Table 5).

Table 5.

Determination of the cutoff value for predicting hypothermia risk during cesarean section.

Cutoff value	Sensitivity	Specificity	Youden index
−1.000	1.000	1.000	0.000
0.500	1.000	0.948	0.052
1.500	1.000	0.850	0.150
2.500	0.988	0.633	0.355
3.500	0.964	0.444	0.520
4.500	0.880	0.308	0.572
5.500	0.819	0.231	0.588
6.500	0.711	0.119	0.592*
7.500	0.542	0.035	0.507
8.500	0.349	0.014	0.335
9.500	0.133	0.003	0.130
10.500	0.012	0.000	0.012
12.000	0.000	0.000	0.000

*Maximum Youden index

In the authenticity evaluation, a total score of 7 points was used as the diagnostic threshold, and parturients in the database were divided into the low-risk and high-risk groups for intraoperative hypothermia. The incidence of intraoperative hypothermia in both groups was calculated, and the difference was tested. The incidence of hypothermia was significantly higher in the high-risk group (63.4%) than in the low-risk group (8.7%) (Table 6).

Table 6.

Incidence of intraoperative hypothermia in different risk groups.

Group	Hypothermia group (%)	Normothermia group (%)	Total number	Chi-square value	p-value
High-risk group	59 (63.4)	34 (36.6)	93 (25.2)	119.582	0.000
Low-risk group	24 (8.7)	252 (91.3)	276 (74.8)
Total number	83 (22.5)	286 (77.5)	369 (100)

3.1. Diagnostic efficiency

The sensitivity, specificity, positive prediction rate, and other indicators were calculated to determine the diagnostic efficiency of the scale for predicting hypothermia risk during cesarean section. The evaluation indices of the diagnostic effect of the test were as follows: sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of the scale were 71.1% (59/83), 88.1% (252/286), 63.4% (59/93), 91.3% (252/276), and 84.3% (59+252/369), respectively.

4. Discussion

At present, medical staff rely only on experience or maternal complaints for intraoperative temperature management, and there are no tools for identifying patients at high risk of maternal hypothermia during cesarean section, impairing the ability to provide personalized warming measures. When the operation turnover is fast or the circuit nurse is busy during the operation, the medical staff can easily overlook observation and management of maternal body temperature. Moreover, as studies have shown that most parturients do not routinely receive temperature monitoring during the operation, it is impossible to objectively determine whether maternal hypothermia has occurred. The most important result in our study is the development of a prediction scale for hypothermia during cesarean section. The area under the receiver operating characteristic curve for our scale for predicting hypothermia risk during cesarean section was 0.891, indicating a good prediction effect, with sensitivity, specificity, and accuracy values of 71.1%, 88.1%, and 84.3%, respectively. These rates indicate appropriate goodness-of-fit and confirm the validity of the scale for predicting maternal hypothermia during cesarean section. The predictive factors involved in this scale included temperature after anesthesia, body mass index, subclinical hypothyroidism during pregnancy, intraoperative fluid loss, amount of intraoperative abdominal cavity irrigation, duration of active insulation, and room temperature before operation. Desgranges et al. found that obesity, active forced-air warming, oxytocin augmentation during labor, core temperature < 37°C upon entering the operating room, temperature and < 36.6°C at the start of the operation, and a fluid infusion volume > 650 ml during the operation were predictive of maternal hypothermia during cesarean section (Desgranges et al., 2017). However, in their retrospective study, they utilized only one study site including 359 parturients, which may have resulted in sampling bias. Further, after anesthesia, a forced-air warming blanket was placed on the parturient's upper body with the warming unit set at 43°C, but their department lacked specific recommendations regarding the use of forced-air warming to prevent hypothermia, which may be viewed as a potential source of bias (Desgranges et al., 2017). As all parturients were given oxytocin in equal doses after delivery in our study, we did not include oxytocin augmentation as a factor. When the volume of fluid infusion was >1,000 ml during cesarean section, 36.59% of the parturients were treated with fluid warming. Thus, no significant difference was observed in this aspect.

In the current study, subclinical hypothyroidism during pregnancy was an independent risk factor for hypothermia during cesarean section. Some studies have shown that patients with hypothyroidism have a greater risk of intraoperative hypothermia (Giuliano and Hendricks, 2017; Huang and Hu, 2016), given that the thyroid gland is related to energy metabolism, which can affect the body's heat production and heat dissipation. Body mass index (≤ 25.05) was an independent risk factor for preventing hypothermia during cesarean section, which is consistent with previous research (Munday et al., 2013; Negishi et al., 1996; Xie et al., 2016).

Intraoperative fluid loss exceeding 980 ml was also associated with a higher risk of maternal hypothermia in the current study. In most previous studies, intraoperative fluid loss in patients only referred to intraoperative blood loss, with results indicating that intraoperative blood loss cannot be considered a predictor of intraoperative hypothermia (Desgranges et al., 2017; Pu et al., 2011). Compared with other surgical patients, parturients lose large amounts of both amniotic fluid and blood when delivering the fetus and placenta, which may in turn lead to maternal heat loss. Intraoperative peritoneal irrigation volume (> 775 ml) was also identified as an independent risk factor for hypothermia during cesarean section in our study, which is consistent with the findings of Yi et al. However, the cutoff value for intraoperative peritoneal irrigation volume was > 500 ml, which may explain why 49.1% participants received active insulation in our study versus 14.17 in the study by Yi et al. (2017).

Our results also indicated that body temperature (< 36.8°C) after anesthesia was associated with a decreased risk of hypothermia. When the difference between body temperature and peripheral ambient temperature is not significant, the redistribution of temperature caused by intraoperative anesthesia will be reduced accordingly (Xie et al., 2016; Zhang et al., 2015). Maintaining a preoperative room temperature above 23°C can also improve the thermal comfort of the parturient and slow down the trend of perioperative hypothermia, which is consistent with the findings of other studies (Duryea et al., 2016; Xie et al., 2016; Zhang et al., 2015). This may be because a higher room temperature can reduce the temperature gradient, convection, and radiation between the maternal body and the environment (Horosz and Malec-Milewska, 2013). However, this result was consistent with the study of Desgranges et al., as room temperature remained almost constant at approximately 22°C in their study (Desgranges et al., 2017).

Before recommendations can be made for routine use of active warming to prevent hypothermia during cesarean section, additional studies are required to demonstrate greater efficacy and determine the optimal active warming technique (Cobb et al., 2016). Our findings indicated that active insulation for over 33 min can help to maintain a stable body temperature during cesarean section, in accordance with previous results (Shin et al., 2015). In this study, of the parturients, 49.1% received active insulation, while only 7.05% in the hypothermia group received active insulation, which showed that the rate of patients receiving active insulation measures during surgery was relatively low, in accordance with previous results (Wang et al., 2017). Preoperative pre-warming combined with intraoperative active insulation can reduce the risk of hypothermia by 1.8 times when compared with active insulation alone (Menzel et al., 2016). Therefore, identification of women at risk for perioperative hypothermia may allow for more targeted active warming strategies that can optimize maintenance of perioperative normothermia.

This study's novelty lies in the following features: First, the scale developed in this study exhibits a good ability to predict hypothermia of parturients during cesarean section when facing diverse clinical circumstances, which establishes a system for hypothermia management to reduce health risks in parturients and promote maternal and newborn health. Making motherhood safer is a human rights imperative and a top priority for the United Nations Fund for Population Activities, which works at all levels to promote universal access to sexual and reproductive health care and rights, including by promoting international maternal health standards and providing guidance and support to health systems.

Second, the risk prediction scale can realize early and dynamic evaluation of the risk of hypothermia during cesarean section, which assists clinical staff in selecting effective measures and timing to prevent hypothermia, such as prewarming, active warming, and maintaining a constant room temperature. For high-risk hypothermia groups, we should eliminate the risk factors of hypothermia, take insulation measures, and focus on preventing hypothermia. Moreover, for low-risk hypothermia groups, we should avoid excessive heat preservation, so as to maximize the utilization of thermal insulation resources, save on warming costs, optimize resource allocation, alleviate the workload of medical staff, and reduce the impact of hypothermia on parturients and newborns. It is of great clinical significance to promote postoperative rehabilitation of parturients and improve the quality of nursing in operating room.

Third, when the research objects, research methods, and statistical methods are different due to the heterogeneity between the studies, the clinical data obtained are different, and the final risk factors are also different. When we apply the hypothermia risk prediction scale, the applicable population, prediction efficiency and consistency of the risk prediction scale should be considered, and the specific analysis should be combined with clinical problems. Therefore, the scale for predicting maternal hypothermia risk developed in this study is more suitable for parturients during cesarean section, which has guiding significance for the prevention of intraoperative hypothermia in parturients during cesarean section.

Fourth, this was a prospective study. It could comprehensively and accurately evaluate the related risk factors affecting the occurrence of hypothermia in parturients during cesarean section, which can avoid the limitations of retrospective research, such as lack of data. Moreover, the data collection was scientific and repeatable. Boet et al identified nine theoretical domains influencing temperature management by medical staff during the perioperative period, including knowledge, beliefs about capabilities, beliefs about consequences, reinforcement, memory/attention/ decision-making, environmental context and resources, social or professional role/identity, social influences, and behavioral regulation (Boet et al., 2017). Therefore, we analyzed factors concerning temperature management as well as maternal, anesthesia factors, and surgical factors, resulting in more precise clinical prediction results in this study. Moreover, it provides reference and inspiration for the development of risk prediction scales of intraoperative hypothermia in patients undergoing other surgery type.

4.1. Limitations

The risk factors of intraoperative washing of the abdominal cavity and intraoperative fluid loss cannot be accurately assessed preoperatively; therefore, during the operation, the clinical medical staff should dynamically assess the risk of maternal hypothermia based on each maternal situation and provide timely and individualized adjustment of heat-preservation measures. Meanwhile, the sensitivity and positive predictive value of our scale were low, which may be related to the high heat-preservation consciousness and low incidence of hypothermia in this study, highlighting the need for a large-scale multicenter study for clinical validation. In addition, the need to monitor temperature during caesarean section is also related to comfort (Gustafsson et al., 2021). In this regard, it can be argued that most patients are at risk of heat loss if preventative measures are not taken. Thus, all patients should be asked about their thermal comfort, including those with normal temperatures or at low risk of hypothermia.

5. Conclusions

The scale for predicting maternal hypothermia risk during cesarean section developed in this study has excellent predictive efficiency as well as high sensitivity and specificity. As the scale can identify parturients who are at a high risk of hypothermia during cesarean section, it may aid in the implementation of clinical practices to decrease the incidence of hypothermia in such patients.

Availability statement

All data included in this study are available upon request by contacting with the corresponding author.

Declaration of Competing Interest

The authors have no conflict of interests to declare.