Your browser is no longer supported

For the best possible experience using our website we recommend you upgrade to a newer version or another browser.

Your browser appears to have cookies disabled. For the best experience of this website, please enable cookies in your browser

We'll assume we have your consent to use cookies, for example so you won't need to log in each time you visit our site.
Learn more

In depth

What is the evidence base for the assessment and evaluation of body temperature?

  • Comment

Measurements and decisions about  body temperature are still based on traditional ideas. Examples in this review encourage practice based on evidence



Märtha Sund-Levander, PhD, RNT, is assistant professor and senior researcher, Unit for Research and Development, Hoegland Hospital, Sweden; Ewa Grodzinsky, PhD, Reg. BLS, is associated professor, Linköping University and head of The Research and Development Unit in Local Health Care, County Council of Östergötland.


Sund-Levander M, Grodzinsky E (2010) What is the evidence base for the assessment and evaluation of body temperature? Nursing Times; 106: 1, early online publication.   

In clinical practice measurement and management of body temperature is often based on tradition and personal ideas and beliefs rather than evidence based knowledge. This article summarises the literature to provide a guide to evidence based assessment and evaluation of body temperature in clinical practice.

Keywords: Assessment, Body temperature, Evidence base, Fever, Measurement

  • This article has been double-blind peer reviewed



Practice points

When assessing and evaluating body temperature use scientific knowledge instead of traditional practice and personal beliefs:

  • It is not appropriate to compare temperature readings taken at different sites and readings should not be adjusted;
  • Evaluate body temperature in relation to individual baseline temperature, which should be noted;
  • Axillary temperature measurement is not recommended to assess core temperature in adolescents and adults;
  • Use the same site on each occasion for individual patients, and ensure technique is appropriate to the equipment used;
  • Time and site of measurement and medication and use of antipyretic medication should be noted in patients’ records;
  • Recognising that older care home residents may have a low baseline body temperature may prevent delayed diagnosis of infection.



Assessment and evaluation of body temperature is one of the oldest known diagnostic methods and still has a huge impact on decisions about medical diagnosis, nursing care, treatment and ordering laboratory tests. The definition of normal body temperature as 37.0°C and fever as 38.0°C or above originates from the middle of the 19th century when thermometers were introduced into medical practice (Musher et al, 1987). At that time there was no knowledge about thermoregulation and the influence of hormones, cellular metabolism and physical activity on body temperature. The measurements were performed in a non-standardised way on patients, indicating that a large number were febrile, and the axillary site, which gives a poor estimation of the core body temperature, was used.

Today there is a general acceptance that normal body temperature is a range rather than a fixed temperature, but there is still widespread confusion surrounding the assessment and evaluation of body temperature in adults (Mackowiak and Wassermann, 1995), especially in older people (Bentley et al, 2001).

Scientific knowledge and evidence based experience should be the base for decisions in nursing and medical care. Yet, in clinical practice, measurement and decisions about body temperature are still based more on tradition and personal ideas and beliefs(Mackowiak and Wassermann, 1995). This was explored in a systematic review that showed a lack of studies performing temperature measurements, and little evidence of a standardised approach to measurement(Sund-Levander et al, 2002). This article aims to summarise the literature, to guide evidence based assessment and evaluation of body temperature in clinical practice.

Literature review

To perform an evidence based assessment and evaluation of body temperature, the influence of thermoregulation, gender, ageing and site of measurement have to be considered. The following review is based on a literature search in MEDLINE and CINAHL for English or Swedish language textbooks, original papers, reviews and scholarly papers, and manually from identified articles’ reference lists(Sund-Levander and Grodzinsky, 2009).



The purpose of thermoregulation is to maintain the core body temperature within an individual temperature range, so called “set point”. This is sustained by thermo-sensitive neurons in the hypothalamus, which integrate information from the surrounding blood and peripheral receptors and compare this information with the set point. The body temperature is then adjusted, that is, heat loss responses are stimulated when the thermal area in the hypothalamus is sensed as warmer than the set point, and heat production responses are activated when the neurons are cooled. In a warm environment heat loss responses are increased and heat production is inhibited, while a cold environment evokes heat production and inhibits heat loss mechanisms. Shivering is an acute response aimed at equalising the peripheral temperature with the core (Van Someren et al, 2002). The skin and the subcutaneous tissues are important insulators of the body.

Due to the diurnal rhythm, which is consistent within individuals both in health and disease, early morning temperature is lower and varies less compared with afternoon and evening body temperatures(Van Someren et al, 2002). Also, cellular metabolism (Van Someren et al, 2002), exercise(Nielsen and Kaciuba-Uscilko, 1998) and ambient temperature influence the individual variation in body temperature. 


Women are thought to have a lower baseline metabolic rate, and in general have a thicker layer of subcutaneous fat, insulating the core from heat gain during hot conditions (Fox et al, 1969). Fertile women have a higher average body temperature than men (Chamberlain et al, 1995), due to female hormones (Baker et al, 2001), while recent research indicates that postmenopausal women have a lower body temperature compared with premenopausal women (Sund-Levander et al, 2004).


Several physiological factors, such as less heat-producing cells, decrease in total body water, delayed and reduced vasoconstriction and vasodilation response and decreased sweating and metabolic rate influence body temperature in older people (Kenney and Munce, 2003). Increased frequency of hypothermia (Mercer, 1998)and  changed shivering response have been observed in this age group (Frank et al, 2000). Also, an inactive lifestyle might decrease heat production (Van Someren et al, 2002). No differences in the average normal body temperature between older women and men have been found (Sund-Levander and Wahren, 2002). Lower body temperature is found in frail older people, those with dementia, those with dependency in activities in daily living and a body mass index <20. Daily medication with analgesics is associated with a higher temperature (Sund-Levander and Wahren, 2002), reflecting the possibility that pain in older people is related to chronic low-grade inflammation with increased circulating levels of pro-inflammatory proteins and consequently causes temperature to rise (Bruunsgaard et al, 2001).

Body temperature measurement

The aim of measuring body temperature is to estimate the core temperature (the temperature in the thoracic and abdominal cavities and some of the muscles and the brain) and not the peripheral temperature (of the skin and subcutaneous tissue). The temperature in the pulmonary artery is generally considered the gold standard of core temperature, but in clinical practice the rectal site has been viewed as an effective minimally invasive method to estimate the “true” body temperature. Traditionally the oral and axillary readings are adjusted to the rectal temperature by adding 0.3°C and 0.5°C, respectively (Betta et al, 1997).

The digital electronic thermometers for invasive measurement and rectal, oral and axillary devices have a sensor that produces electronic signals, reflecting the tissue temperature. The temperature is displayed as an unadjusted or adjusted value either in a steady-state mode after the sensor has reached equilibrium, or a predictive mode (Sund-Levander, 2004).The infrared radiation ear device (IRED) estimates the infrared heat waves from the tympanic membrane (Blatties, 1998).

The rectal site

The rectal site is assumed to estimate the temperature of the deep viscera. As the measurement is influenced by the low blood flow and high isolation of the area and heat-producing activity of microorganisms in faeces (Blatties, 1998) it is, in normal circumstances, higher than at other sites (Smitz et al, 2000).

However, rectal temperature significantly lags behind temperature changes at other sites, and may be either higher or lower than the core temperature during rapid temperature changes, for example, cooling of the skin, exercise and fever (Blatties, 1998). Hard stool obstructing adequate placing of the thermometer and inflammation around the rectum influence the reading (Varney et al, 2001). It is recommended that the thermometer is inserted to a depth of 4cm to avoid error (Blatties, 1998).

The oral site

The oral temperature follows changes in core temperature because of proximity to the external carotid artery (Blatties, 1998), but the sublingual temperature differs between the posterior pocket and the front area (Erickson, 1980), as well as between the right and left posterior pockets (Modell et al, 1998). Other influencing factors are salivation, previous intake of hot or cold food and fluids, gum chewing, smoking and rapid breathing (Blatties, 1998; Rabinowitz et al, 1996). A fall in oral temperature during the onset of fever may occur due to reduced blood flow.

The axillary site

The accuracy of the axillary site is strongly affected by ambient temperature, local blood flow, underarm sweat, inappropriate placement of the probe or poor closure of the axillary cavity, and the time it takes to record a reading (Blatties, 1998). Furthermore, a temperature difference of 1.4°C has been found between the right and left axilla (Howell, 1972).  Axillary measurements, even with careful positioning, slowly register changes in core temperature and the readings widely deviate from other sites (Robinson et al, 1998), especially during fever. These factors make the axillary site unsuitable for estimating core body temperature (Metlay et al, 1997).

The ear site

The tympanic membrane and hypothalamus share their blood supply from the internal and external carotid arteries but as the IRED probe is not placed in direct contact with the membrane, the reading is a mix of heat from the tympanic membrane and the aural canal (Fraden, 1991). To compensate for the deviation, IRED instruments include an offset system(Lefrant et al, 2003). The ear temperature can be measured in the unadjusted mode, as well as in the adjusted mode, in which the manufacturer has readjusted the value in order to equalise the ear temperature with the oral, rectal or axilla temperatures (Betta et al, 1997). There are considerable variations in how manufacturers have calculated these mathematical algorithms for adjusting to other sites.

The accuracy of IRED for reflecting the tympanic membrane temperature and changes in the core temperature during physical exercise and warming are good (Shibasaki et al, 1998). Readings may be influenced by ambient temperature (Betta et al, 1997), and, although the literature is inconsistent, cerumen (Rabinowitz et al, 1996) and otitis media (Chamberlain et al, 1991). A narrow ear cavity may affect repeatability, because of difficulty positioning the probe correctly. The difference in temperature found between the left and right ear indicates the impact of operator performance on the temperature readings. Duplicate or triplicate ear temperature measurements have been suggested to improve accuracy but recently it was shown that repeated measurements do not improve reproducibility and that one measurement is sufficient when the ear, oral or axillary temperature is measured (Sund-Levander et al, 2004). In practice, the value of IRED measurements is not consistent, due to observed differences between left and right ear and poor repeatability (Craig et al, 2002; Chamberlain et al, 1995).


Comparison between sites

Many studies have focused on the degree of uniformity between different sites to define the best choice for estimating the core temperature using a non-invasive method (Sund-Levander et al, 2002). A study simultaneously measured rectal, oral, ear and axillary temperatures and compared them in healthy adults. The deviations were;

  • —0.7°C to + 2.8°C for rectal-ear temperatures;
  • – 1.4°C to + 2.3°C for rectal-axillary;
  • - 1.5°C to + 2.3°C for rectal-oral (Sund-Levander et al, 2004).

This demonstrates it is not possible to accurately convert temperatures recorded at one site to estimate that at another. The lack of agreement between measurements illustrates it is not possible to interchange temperatures recorded at different sites or to claim there is one range of core body temperature.

It is important to remember that all temperatures outside the hypothalamus are themselves estimates of core temperature with their own variability (Erickson and Meyer, 1994).

Examples from clinical practice

The following stories from clinical practice illustrate the consequences when evaluation of body temperature measurement is guided by routine and tradition rather than evidence based knowledge.

Case study 1

A 23 year old man with high fever, shivering, a sore throat, joint pain, dry cough, headache and problems with breathing visited A&E with his mother. Two days earlier, when his mother had called the GP, the nurse recommended painkillers (paracetamol) and to wait and see.

When they went to A&E, there was a four hour waiting time. The nurse there measured the man’s body temperature and noted it as 36.4 °C. She decided he did not have a fever and advised the patient to go home and contact his GP the next day. The patient’s mother refused to leave A&E and demanded a second opinion from another nurse, who measured his body temperature again, and noted 41.7°C. The doctor assessed the patient’s condition as nonspecific septicaemia and he was admitted to hospital for treatment.


Suspecting fever

Case study 1 illustrates how custom and practice mean that a fever is only suspected if a temperature is 38°C or above. More importantly, although the patient was seriously ill, the nurse’s decision was determined by the measurement on the thermometer. The story does not tell us at what site the body temperature was measured, but it is obvious that the nurse used an incorrect technique. Furthermore, it is not noted if the young man had taken paracetamol, as recommended by the nurse at telephone contact, before visiting A&E. Nevertheless, the first nurse in A&E should have reacted differently, as the patient was presenting with serious signs of infection.

Learning points:

  • The same site of measurement should be used as far as possible;
  • Note the time, site of measurement and medication with antipyretics;
  • Use a correct measuring technique.


Case study 2

This is an excerpt from the medical record of a 93 year old woman with severe dementia, living in a care home. Due to her deteriorated condition she is referred to a GP. The day after she is seen again and assessed as having pneumonia.

Day one: poor condition in the morning, pale, wheezing and restless. Legs and feet are warm. Rectal temperature 37.1°C. No fever.

Day two: difficulty in breathing, wheezing, cough. Temperature in the morning: in the ear 37.7°C and rectum 38.3 °C. A diagnosis of pneumonia is made.  


Identifying infection

This woman in Case study 2 died two days after the second day of documentation. It is obvious from the case record that the absence of fever, as assessed by the doctor, is crucial for her further care. Was it possible to identify the infection earlier? Frail older people are at risk of having a low baseline body temperature (Sund-Levander and Wahren, 2002), in this case her normal morning temperature was 35.6°C rectally and  35.0°C in the ear. In this case a temperature of 37.1ºC indicates a fever. Atypical symptoms are common in older people with infections (Sund-Levander and Grodzinsky, 2009) and this can lead to delayed diagnosis and treatment (Metlay et al, 1997), and subsequent mortality. In addition, almost 50% of care home residents take paracetamol every day, lowering temperature in fever, a fact that seriously influences the evaluation of temperature readings and identification of the presence of fever (Sund-Levander, 2004).

Learning points:

  • Evaluate body temperature in relation to the individual baseline temperature;
  • Individual baseline temperature in care home residents should be noted;
  • Medication with antipyretics should be noted;
  • To prevent a delayed diagnosis of infection in care home residents, nurses should be aware of a low baseline body temperature.


Case study 3

Case study 3 is about cooling of the skin to lower body temperature in a patient with severe head injury (Sund-Levander and Wahren, 2000). The record of care is recorded in Fig 1.

At the beginning of the observation the gradient between the central (tympanic) and peripheral (tip toe) temperature is about 6°C and then increases to 17°C. The patient starts to shiver and the gradient is reduced to 2°C and then again increases to 14°C. During these temperature changes the patient has a period of severe shivering, which triggers the nurse to take action by measuring the axillary temperature and to cool the skin with a fan and alcoholic wraps.

Cooling the skin

This case underlines the importance of knowing the physiology of thermoregulation and consequences of nursing actions. The gradient between the central and peripheral temperature is increasing due to a rising set point associated with fever. Because of the increased temperature gradient between the core and peripheral temperature, the patient started to shiver to equalise central and peripheral temperature. This caused nursing staff to measure the axillary value, which is an unreliable technique for estimating core temperature (Metlay et al, 1997).

As the temperature was >38.0°C, the nurse started cooling the skin with a fan and alcoholic wraps, to lower the core body temperature. The thermoregulatory shivering response has caused the skin temperature to increase and as a consequence the temperature gradient has reduced to only 2°C. However, as the nurse did not consider this physiological response, the cooling of the skin with the fan once again increased the temperature gradient, which triggered shivering once more. For patients, shivering is extremely strenuous, releases stress hormones, is unpleasant and related to fatigue, exhaustion and feelings of helplessness. To prevent shivering, patients should be given antipyretics before surface cooling starts (Mackowiak, 2000).

Learning points:

  • Base the site of body temperature measurement on thermoregulatory mechanisms;
  • Base nursing actions in fever on thermoregulatory mechanisms;
  • Do not use the axillary site as an assessment of core body temperature in adolescents and adults.


Clinical implications

The range of body temperature is wider than traditionally thought, and varies with age, gender, site of measurement and operator techniques. As normal body temperature shows individual variations, it is reasonable that the same should hold true for the febrile range. It is, therefore, important to know patients’ baseline morning temperature to be able to assess an increased temperature as fever. A lower baseline temperature may be related to a lower febrile response in older care home residents. As a basis for assessing fever, therefore, nurses should establish individual baseline body temperature in the morning on admission to care home facilities. It is also important to note medication which lowers temperature in fever.





  • Comment

Have your say

You must sign in to make a comment

Please remember that the submission of any material is governed by our Terms and Conditions and by submitting material you confirm your agreement to these Terms and Conditions. Links may be included in your comments but HTML is not permitted.