Temperature management in the intensive care unit: a practical survey from China

ABSTRACT

Introduction: Temperature management is an important aspect of the treatment of critically ill patients, but there are differences in the measurement and management of temperature in different Intensive Care Units (ICUs). The objective of this study was to understand the current situation of temperature measurement and management in ICUs in China, and to provide a basis for standardized temperature management in ICUs.

Methods: A 20-question survey was used to gather information on temperature management strategies from ICUs across China. Data such as method and frequency of temperature measurement, management goals, cooling measures, and temperature management recommendations were collected.

Results: A total of 425 questionnaires from unique ICUs were included in the study, with responses collected from all provinces and autonomous regions in China. Mercury thermometers were the most widely used measurement tool (82.39%) and the axilla was the most common measurement site (96.47%). There was considerable variability in the frequency of temperature measurement, the temperature at which intervention should begin, intervention duration, and temperature management goals. While there was no clearly preferred drug-based cooling method, the most widely used equipment-based cooling method was the ice blanket machine (93.18%). The most frequent recommendations for promoting temperature management were continuous monitoring and targeted management.

Conclusion: Our investigation revealed a high level of variability in the methods of temperature measurement and management among ICUs in China. Since fever is a common clinical symptom in critically ill patients and can lead to prolonged ICU stays, we propose that standardized guidelines are urgently needed for the management of body temperature (BT) in these patients.

KEYWORDS:

• Body temperature
• critical care
• fever
• management
• survey

1. Introduction

Body temperature (BT) is an important vital sign used to evaluate critically ill patients [1]. The normal temperature of the human body is between 36.0 and 37.5°C [2]. Fever is usually defined as core BT > 38°C, while hypothermia is defined as core BT < 36°C [3]. Disrupted BT is common in critically ill patients and predicts adverse outcomes [4–6], with elevated BT detected in approximately 50% of intensive care unit (ICU) adult patients [7], including those with acute neurological disease, noninfectious fever, and fever during ICU stay [5]. In an observational study of 24,204 adult ICU patients, BT ≥ 39.5°C was associated with increased mortality (20% vs 12%) [8]. However, in some studies, temperature measurement methods are usually not considered, resulting in uncertainties in accuracy and reliability [9,10].

BT measurement methods commonly used in ICU include axillary temperature, oral temperature, ear temperature, rectal temperature, and bladder temperature [11]. Pulmonary artery temperature measurement is the core BT measured by pulmonary artery catheter, which is the gold standard for BT measurement; however, this measurement is difficult to obtain routinely in ICUs [12]. Due to the critical condition of ICU patients, conventional oral or anal temperature monitoring methods are often not suitable [11]. Likewise, axillary temperature monitoring is not suitable for patients who are emetic or unable to cooperate [13]. Three observational studies conducted in ICUs in Australia and New Zealand confirmed the widespread use of various non-invasive methods [14–16], and a survey showed different attitudes of doctors and nurses towards temperature management in ICUs [17]. However, even small alterations in BT can lead to changes in inflammation and immune function, both of which may affect the prognoses of patients [18]. Therefore, possessing a method to accurately obtain the real BT is of great significance for assessing disease, establishing early diagnoses, guiding treatment, and judging the prognoses of ICU patients.

Temperature management is important in critically ill patients, and several large trials have been conducted to clarify temperature targets, management strategies, and timing. Although patient temperatures are well documented in the ICU [19], it is often unclear when and how to intervene when a patient’s temperature rises. Young et al. [20] showed that across all critically ill patients, active treatment of fever did not improve the survival rate. In the treatment of ICU patients with fever, reduced intervention and active intervention had similar prognoses. ‘Target body temperature management’ (TTM) has become a measure to intervene in each patient’s individual specific temperature, as the level of target BT may vary from case to case, to prevent fever, maintain normal BT, or lower the central BT [21]. However, indications for TTM are still based on varying levels of evidence, some of which are contradictory [18]. Especially in ischemia anoxic encephalopathy in patients with severe head injury, to reduce the brain metabolism, reduce tissue inflammation, and prevent neuronal apoptosis, hypothermia treatment is recommended [22]. However, controlling temperature to achieve low temperature treatment goals and duration remains controversial, various medical institutions there exist great differences in the working process [23].

To understand the current situation of temperature measurement and management in ICUs in China and to standardize the management of ICU temperature, we conducted a nationwide questionnaire survey to explore the methods of temperature measurement, the goals of temperature management, and the strategies of temperature management in ICU patients.

2. Methods

We designed a 20-question online questionnaire (Questionnaire Star) that was administered nationwide from August to September 2022 (Appendix; Figure 1). Survey participants included all provinces and autonomous regions in China (Figure 2). The survey was deemed exempt by the Institutional Ethics Review Board of Shandong Provincial Hospital. Patients and the public were not involved in the development of this survey. All methods were carried out in accordance with relevant guidelines and regulations. The survey included basic information about the ICUs, methods and frequency of temperature measurement, goals of temperature management, cooling measures, and recommendations for temperature management. The survey was administered to comprehensive ICUs (ie; who provide care across a wide range of diagnoses) and specialized ICUs (ie; provide diagnosis-specific care) at all levels of hospitals nationwide. We defined ‘ICU’ as a unit capable of providing invasive mechanical ventilation and organ function support, such as the use of vasoactive drugs and renal replacement therapy. Each ICU was also required to have an independent management system as a primary clinical unit in a hospital.

Figure 1. Flowchart of survey administration, quality control, and data analysis. ICU: intensive care unit.

Figure 2. The distribution of survey participants throughout China.

To ensure the integrity of the returned questionnaire, answers to all questions in the questionnaire were required; otherwise, the submission could not be completed. Duplicate submissions from the same ICU were excluded. All participants completed the questionnaire online through their mobile phone or computer browser. To ensure the authenticity of the data, survey respondents for each site were the quality control manager of the ICU, the full-time attending doctor, or the ICU doctor. All submitted data were reviewed and submitted by the hospital director.

Data were analyzed using SPSS version 22 (IBM, Armonk, NY). Descriptive and summary statistics were performed for all variables. Enumeration data were described including count, rate, constituent ratio, and mean. Categorical data were analyzed by chi-square test or Fisher exact test. When the P-value was less than 0.05, the difference was statistically significant.

3. Results

A total of 559 questionnaires were collected, of which 32 were excluded due to obvious answer errors including out of range values, inconsistent responses, or nonsensical responses. Another 102 questionnaires were excluded due to having respondents from duplicate ICUs, leaving a total of 425 valid questionnaires to be included in the study. This study included 237 ICUs from Class III Grade A hospitals (55.76%), 77 ICUs from class III grade B hospitals (18.12%), 104 ICUs from class II grade A hospitals (24.47%), and 7 ICUs from class II grade B hospitals (1.65%).

3.1. Methods of temperature measurement in each ICU

Mercury thermometers were the most widely used tool for measuring BT (82.39%), followed by electronic temperature guns (46.12%) and temperature probes equipped with monitors or ice-blanket machines (40%) (Figure 3A). The most common site for BT measurement was axilla temperature (96.47%), followed by rectal temperature (28%) and bladder temperature (11.53%). (Figure 3B). A percentage of 10.12% of ICUs took BT measurements every hour, 10.59% took them every two hours, 59.76% took them every four hours, 7.06% took them at intervals longer than four hours, and 12.47% of ICUs used continuous temperature monitoring; all used temperature probes. Among the respondents, 65.53% thought rectal temperature best reflected the actual BT of patients.

Figure 3. Body temperature (abbreviated ‘temp’) measurement instruments and sites. (A) Instruments used to obtain body temperature measurements in intensive care unit (ICU) patients. Other refers to cooling blanket probe, electronic thermometer. (B) Preferred sites of body temperature measurement in ICU patients. Other refers to forehead temperature, nasopharyngeal temperature.

3.2. Objectives of temperature management

About half of ICUs adopted TTM (225, 52.94%), among which 19.56% expected temperature control to be exerted for BT between 36.0–36.5°C, 35.56% stated that temperatures should be controlled if they are below 37°C, 29.78% placed the cutoff at 37.5°C, 11.56% placed the cutoff at 38°C, and 3.56% placed the cutoff at 38.5°C.

Only 1/4 of ICUs used preventive cooling (108, 25.41%), including 54.63% which have adopted preventive cooling measures for BT > 38°C, 35.19% for BT > 37.5°C, 9.26% for BT > 37.0°C, and 0.93% for BT > 36.5°C.

3.3. Cooling measures adopted in each ICU

Among 425 ICUs, 417 (98.12%) ICUs were treated with physical cooling, 374 (88.0%) ICUs were treated with drug cooling, 374 (88.0%) icus were treated with equipment cooling, and 329 (77.41%) ICUs were treated with all three cooling methods. The most used physical cooling measures include ice packs (96.24%), cooling blankets (69.88%), and lukewarm baths (46.35%). The most used cooling drug dosage forms included oral (57.18%), intramuscular (66.12%), intravenous (47.06%), suppository (64.0%), and 3.5% ICU would choose antifebrile patch, enema and other methods. The most frequent antipyretic drugs were compound aminobarbital injection (55.06%), ibuprofen (51.06%), indomethacin suppository (45.41%), and glucocorticoids (39.76%) (Figure 4).

Figure 4. Antipyretic drugs used for the cooling of intensive care unit (ICU) patients.

Ice blanket machines were the most used cooling equipment (93.18%), followed by blood purification equipment (64.94%). Only 4.24% of ICUs were equipped with intravascular heat exchange cooling equipment, and 3.76% of ICUs did not use cooling equipment (Figure 5). A percentage of 49.88% of ICUs would administer cooling measures for one hour, 34.59% of ICUs for 2 hours, and 10.35% for 4 hours, after which further cooling treatment would be considered if BT had not dropped significantly.

Figure 5. Cooling equipment used for intensive care unit (ICU) patients with elevated body temperatures.

3.4. Temperature control range under special circumstances

For patients who required mild hypothermia after cardiopulmonary resuscitation or nerve injury, 37.88% of ICUs reported controlling the BT between 32–34°C, 48.24% reported controlling the BT between 34–36°C, and 12.0% reported controlling the BT between 35–37°C. For the duration of mild hypothermia, 20.71% of ICUs preferred ≤24 h, 27.06% of ICUs preferred ≤48 h, 39.06% of ICUs preferred ≤72 h, and 13.18% of ICUs preferred >72 h.

3.5. Recommendations to promote temperature management in ICU

A total of 106 respondents made recommendations for improved temperature management in ICUs. We statistically analyzed the recommendations proposed and found that the most frequent key phrases used were ‘continuous temperature monitoring’ (29 times), ‘targeted temperature management’ (22 times), and ‘development of guidelines and norms related to temperature management’ (19 times) (Figure 6).

Figure 6. Frequency of key phrases used in respondent recommendations to promote temperature management.

4. Discussion

This study involved the distribution of a large-scale survey on temperature management in ICUs in China. Through the survey, we found that there are substantial differences in the methods of temperature measurement and in the strategies of mild hypothermia treatment in various ICUs. We recommend urgent establishment of relevant guidelines and norms to unify and standardize temperature management in ICUs.

This survey found that mercury thermometers remained the most widely used BT measurement tool, followed by electronic BT guns, and probes equipped with monitors or ice blanket machines. The most common sites for measuring BT were the axilla, followed by the rectum and bladder. However, a subset of the most severely ill patients may be experiencing shock or tissue malperfusion. The axillary and forehead temperatures measured by mercury thermometers or electronic BT guns can only reflect skin temperature rather than accurately reflecting core temperature. Moreover, measurements may be affected by ambient temperature, sweating, and evaporation, and it is not uncommon patients with elevated core temperatures to show normal skin temperature [24,25]. Therefore, this method of temperature measurement is not suitable for use in the ICU environment. Clinically, rectal temperature is relatively common, but the rectal probe must be introduced to a depth of 15 cm to accurately reflect core BT. For conscious patients, insertion of the rectal probe will cause considerable pain and discomfort; for unconscious patients, due to temporary paralysis of the autonomic nervous system, their rectal wall will expand and thin, causing poor correlation between rectal temperature and core temperature, and the reading can be significantly delayed [26]. The bladder temperature is very close to the core temperature and is becoming more and more commonly used in ICUs [27]. Indwelling catheters with temperature sensing can continuously measure BT. Normal urine volume has a good correlation with the core BT, but this temperature measurement method is inaccurate if the urine volume decreases due to decreased cardiac output, acute kidney injury and other factors [11].

The accuracy of temperature measurements depends on the location and method of measurement [19]. The tympanic membrane, nasopharynx, esophagus, and pulmonary artery are considered to be the most responsive sites of core BT [26,28], but they cannot be widely implemented in clinical practice due to invasiveness and complex operation. Ideally, temperature measurements should be simple, noninvasive, and harmless, reflecting core temperature as accurately as possible without being significantly affected by ambient temperature. Although we live in an era of cutting-edge medical technology, no standard technology has been developed to measure BT. Therefore, at the present stage, regardless of which method is adopted, the same measurement method and site should be repeated to understand the trend of continuous change.

Fever is a common clinical symptom in ICU, which can lead to prolonged ICU stay. High temperature (ie; >38°C) and persistent fever (ie; fever that lasts for an extended period, typically beyond a week) are associated with increased risk of death [29]. Although a patient’s temperature is well documented during an ICU stay, it is often unclear when and how to intervene when a patient’s temperature rises. To alleviate disease, improve clinical symptoms, and try to obtain a better clinical prognosis, clinicians usually apply some cooling measures, including drug cooling and physical cooling. Drug cooling aims to control BT by inhibiting the release of endogenous heat source (e.g. Interleukin-1, tumor necrosis factor). Physical cooling reduces the BT by accelerating heat dissipation [30,31]. It was found that most ICUs preferred physical cooling, including ice packs, cooling blankets, and warm water baths. It was generally believed that physical cooling was preferable to drug cooling [31]. Firstly, adverse reactions of antipyretic drugs, especially the use of non-steroidal drugs and glucocorticoids, were considered [32,33]. Second, fever that has nothing to do with a physiological temperature-increasing mechanism is less responsive to drug cooling [34]. Third, although drug cooling achieves the purpose of reducing BT by inhibiting the release of endogenous substances promoting heat production, it also interferes with the inflammatory response of the body to a certain extent [35]. There is evidence that temperature in ICU patients can be controlled with medications [32] and physical cooling devices [36]; however, it is not clear whether these measures contribute to the development of hypothermia.

TTM can be used to prevent fever and maintain normal BT. This survey found that over half of ICUs used TTM, and 85% of these ICUs kept their TTM below 37.5°C. The goal of temperature control in patients without nerve damage is still unclear. A 2019 meta-analysis of individual patient data showed that in ICU patients, more aggressive temperature management did not improve survival compared with passive temperature management strategies [20]. In the study, survival was similar across treatment subgroups, regardless of whether they were grouped according to age, disease severity, or receiving specific organ support [20]. In addition, a phase II study of active temperature reduction to normal in sedated and mechanically ventilated sepsis patients found that active temperature management reduced the early mortality of such patients compared with conventional treatment [36]. Young et al. [33] conducted a random evaluation of active control of temperature and ordinary temperature management test. This multicenter randomized clinical trial involving ICU patients with fever without acute brain injury showed that mean BT was reduced by approximately 0.5°C in the active temperature control group compared with the ordinary temperature management group [33].

Our current investigation found significant inconsistencies across units in the management of BT, including target BT and duration of mild hypothermia, for patients requiring mild hypothermia after cardiopulmonary resuscitation or neurological injury. Mild hypothermia therapy is a treatment method that reduces the BT or local brain temperature of patients by artificial physical methods, thereby reducing cerebral oxygen consumption and promoting the recovery of brain function [37]. It is suitable for such purposes as brain protection during cardiopulmonary bypass in cardiac surgery, craniocerebral injury related to decreased cerebral perfusion pressure, and ischemic hypoxic encephalopathy after cardiopulmonary resuscitation. According to the target temperature setting, mild hypothermia can be divided into four classes: mild (34.5–36.5°C), moderate (32–34.5°C), severe (28–32°C) and extreme mild (<28°C) [35]. Although an increasing number of studies have confirmed that adult out-of-hospital cardiac arrest (CA) patients treated with mild hypothermia have better neurological outcomes and lower mortality, the optimal target temperature is still a range rather than a specific value [21,37,38]. According to the guidelines issued by the European Resuscitation Committee in 2015, the recommended target temperature for mild hypothermia in patients with CA is 32–36°C [21,25]. It is not clear whether patients in different CA subgroups can benefit from lower temperature (32–34°C) or higher temperature (36°C) [39]. The American Heart Association guidelines for cardiopulmonary resuscitation published in the same year also indicated that CA patients with different characteristics may be suitable for different target temperatures [40]. In a recently published clinical randomized controlled study, there was no significant difference in the neurological outcome of patients treated with different target temperatures (32°C, 33°C, and 34°C) at 90 days after return of spontaneous circulation, and no significant difference in the neurological outcome of patients with a target temperature ≤ 34°C or > 34°C at six months. There was no significant difference between pediatric in-hospital CA patients treated with the target temperature of 33.0°C and 36.8°C [41–43]. These debates about the optimal target temperature suggest that the most effective outcomes of mild hypothermia should be achieved because of clinical bundle therapy strategies, rather than simply cooling.

There are no definitive clinical data on the optimal duration of maintaining the target temperature to answer this question. From the perspective of the brain protective mechanism of mild hypothermia, it seems that the longer the target temperature is maintained, the greater the benefit will be [13]. However, this is not the case. Clinical studies have found that mild hypothermia lasting for 48 h does not significantly improve the neurological prognosis of patients at six months compared with 24 h, but the incidence of adverse events is higher and the length of ICU stay is longer [44]. Although blood flow velocity of middle cerebral artery and oxygen saturation of jugular sinus were significantly increased at 72 h after mild hypothermia compared with the start of treatment, side effects of mild hypothermia also increased [23]. Currently, in the United States and European guidelines, the target temperature for mild hypothermia after cardiopulmonary resuscitation is 32–36°C, and the recommended maintenance time is at least 24 h [29,30]. However, for patients with craniocerebral injury, short-term (24-48 h) mild hypothermia therapy is difficult to achieve significantly better clinical effects. Mild hypothermia therapy lasting longer than 48 h can reduce the mortality of patients with traumatic craniocerebral injury. It is suggested that such patients should maintain mild hypothermia treatment for at least 3–5 days.

This study has several limitations. It may suffer from sampling bias due to its voluntary participation, potentially limiting its representativeness. The reliance on self-reported data introduces response bias, and the cross-sectional design provides only a static view of practices. Moreover, it lacks clinical outcome assessment and does not explore reasons behind practice variations or resource influences. The study’s findings could become outdated, and it does not consider language, cultural factors, or publication bias. These limitations underscore the need for caution when generalizing the results and suggest room for more comprehensive research on this topic.

In conclusion, we found that there are great differences in the methods of temperature measurement, temperature management, especially the strategies of mild hypothermia treatment among ICUs in China. At the same time, some problems in ICU temperature management were exposed, such as whether drug-based or physical cooling is preferable, the appropriate threshold for cooling, and the ideal target temperature of cooling. The goal and duration of mild hypothermia after cardiopulmonary resuscitation and in patients with craniocerebral injury are still in need of further research, and relevant guidelines and norms are urgently needed to unify and standardize temperature management in ICU.

This work highlights the need for standardized guidelines in temperature management for ICU patients due to significant variations in methods and strategies. It emphasizes the importance of adopting accurate temperature measurement methods, such as bladder temperature monitoring, and optimizing cooling measures to enhance patient care. The study also underscores the necessity of determining precise temperature control targets and the duration of mild hypothermia treatment to improve patient outcomes. In summary, the findings call for standardized, evidence-based approaches to temperature management in ICUs to positively influence patient care.

Authors’ contributions

XS was involved in the conceptualization, planning and design of the project, and in the development and execution of the survey. LM,CFW, and XL designed the questionnaire, drafted the initial manuscript, reviewed and revised the manuscript. YB, KZ and YY performed the data collection, statistical analysis and assisted in drafting and revising the manuscript. CTW assisted in the execution of the survey and assisted in the development of the manuscript. All the authors have accepted responsibility for the entire content of this submitted manuscript and approved of this submission.

Availability of data and materials

The data and materials used or analysed during the current study are available from the corresponding author on reasonable request.

Ethics approval and consent to participate

The questionnaire and methodology for this study were deemed exempt by the Institutional Ethics Review Board of Shandong Provincial Hospital, as it is a survey research. Informed consent was obtained from all participants.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by National Natural Science Foundation of China (Grant 82102274) and Natural Science Foundation of Shandong Province, China (Grant ZR202102210272), Taishan Scholars Foundation of Shandong Province (Grant tsqn202211347) and Shandong Provincial Postdoctoral Science Foundation (Grant SDCX-ZG-202202025). The funders had no role in the design and conduct of the study; collection, management, analysis, or interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Appendix

Questionnaire of ICU temperature management survey

I. Basic information

1. The name of the hospital where you work:                 

2. Hospital level

A. LevelIII Grade A

B. LevelIII Grade B

C. Level II Grade A

D.LevelII Grade B

3. ICU style?

A. SurgicalICU

B. InternalICU

C. GeneralICU

D. EmergencyICU

E. CCU

F. NICU

G. RICU

H. Other

4. ICU beds number:         

II. Current status of temperature management in ICU

5. What equipment does your department currently use to measure temperature for patients? (Multiple options available)

A. Mercurythermometer

B. Electronictemperature gun

C. Monitorbody temperature probe

D. Catheter temperature probe

E. Other

6. What method of temperature measurement is currently applied in your department? (Multiple options available)

A. Axillary

B. Oral

C. Ear

D. Bladder

E. Rectal

F. Other

7. Which method do you think is best at detecting actual temperature?

A. Axillary

B. Oral

C. Ear

D. Bladder

E. Rectal

F. Other

8. How often is temperature measured in your department?

A. Continuousmonitoring

B. Eachhour

C. Every2 hours

D. Every4 hours

E. Over4 hours

9. Whether targeted temperature management is applied in your department ?

A. Yes

B. No

10. What is the target temperature ?

A. 36.0°C—36.5°C

B. <37.0°C

C. <37.5°C

D. <38.0°C

E. <38.5°C

11. Whether preventive cooling applied in your department?

A. Yes

B. No

12. What is the temperature above which preventive cooling measures should be taken?

A. >36.5°C

B. >37.0°C

C. >37.5°C

D. >38.0°C

13. What cooling methods are applied in your department? (Multiple options available)

A. Physical

B. Medicine

B. Equipment

14. What physical cooling methods are commonly used in your department? (Multiple options available)

A. Icebag

B. Warmwater wiping

C. Alcoholwiping

D. Reducedcover on patients

E. Reduceroom temperature

15. What types of antipyretic drugs are commonly used in your department? (Multiple options available)

A. Oral

B. Injection

C. Suppository

16. What are the commonly used antipyretic drugs in your department? (Multiple options available)

A. Ibuprofen

B. Acetaminophen

C. Aspirin-DL-lysine

D. Aspirin

E. Indomethacin

F. Nimesulide

G. Diclofenac

H. Compoundaminobarbital injection

I. Hormone

J. Other

17. What cooling devices are commonly used in your department? (Multiple options available)

A. Iceblanket machine

B. Iceblanket machine

C. Intravascular heat exchange cooling

D. CVVH

18. How long after the cooling methods are taken, if the temperature does not decrease significantly, further cooling measures will be taken?

A. 1hour

B. 2 hours

C. 4 hours

D. >4 hours

19. What is the range of temperature control for patients who require mild hypothermia after CPR or nerve injury?

A. 32°C—34°C

B. 34°C—36°C

C. 35°C—37°C

D. Other

20. What are your recommendations for better temperature management in the ICU?