Comparison of temporal artery, mid-forehead skin and axillary temperature recordings in preterm infants <1500 g of birthweight

Comparing Axillary and Rectal Temperature Measurements in Very Preterm Infants: A Prospective Observational Study

PURPOSE

The agreement between axillary temperature (AT) and rectal temperature (RT) measurements has not been well established in preterm infants. Therefore, our study aimed to evaluate the agreement between AT and RT measurements in very preterm infants.

METHODS

Preterm infants <32 weeks of gestational age were prospectively included. The infants' body temperature (BT) was measured twice a day from day 1 to day 6. A paired t-test and the Bland-Altman method were used to analyze the difference between the AT and RT. A linear regression model was used to explore the effects of environmental factors on the differences of BT between the axillary and rectal measurements and to calibrate the RT according to the AT.

RESULTS

Eighty infants each underwent 6 paired axillary and rectal measurements. The gestational age varied from 22 to 31 weeks (mean 28 weeks). The birth weight varied from 302 to 1,770 g (mean 1,025 g). The AT was significantly lower than the RT. The difference between the RT and AT significantly increased with increasing RT. The AT and RT demonstrated poor agreement overall; however, the RT can be estimated using the AT with the following equation: RT = -4.033 + 1.116 × AT. Environmental factors, including the incubator temperature, incubator humidity, phototherapy, and application of invasive mechanical ventilation did not affect the differences between the AT and RT measurements.

CONCLUSION

AT measurements cannot be interchangeably used with RT measurements in very preterm infants.

Temperature Monitoring Devices in Neonates

Introduction: Accurate temperature monitoring of neonates is vital due to the significant morbidities and mortality associated with neonatal hypothermia. Many studies have compared different thermometers in neonates, however, there is a lack of consensus regarding which of the currently available thermometers is most suitable for use in neonates. Objectives: The aim of this review was to identify and compare current methods available for temperature monitoring of neonates beyond the delivery room, including the accuracy, advantages and disadvantages of each. Methods: A recent search and narrative synthesis of relevant studies published between January 1, 1949 and May 5, 2021 on the OVID Medline, PubMed and Google Scholar databases. Results: A total of 160 papers were retrieved for narrative synthesis. The main methods available for temperature monitoring in neonates are human touch and mercury-in-glass, electronic, infrared tympanic and other infrared thermometers. Newer innovations that are also available include liquid crystal thermometers and the BEMPU TempWatch. This paper discusses the current evidence available regarding the utility of these devices, and identifies barriers to valid comparison of different thermometry methods. Conclusion: Many methods for temperature monitoring in neonates are currently available, each with their own advantages and disadvantages. However, the accuracies of different devices are hard to determine due to variable methodologies used in relevant studies and hence, further research that addresses these gaps is needed.

Diagnostic accuracy of non-contact infrared thermometers and thermal scanners: a systematic review and meta-analysis

Infrared thermal screening, via the use of handheld non-contact infrared thermometers (NCITs) and thermal scanners, has been widely implemented all over the world. We performed a systematic review and meta-analysis to investigate its diagnostic accuracy for the detection of fever. We searched PubMed, Embase, the Cochrane Library, medRxiv, bioRxiv, ClinicalTrials.gov, COVID-19 Open Research Dataset, COVID-19 research database, Epistemonikos, EPPI-Centre, World Health Organization International Clinical Trials Registry Platform, Scopus and Web of Science databases for studies where a non-contact infrared device was used to detect fever against a reference standard of conventional thermometers. Forest plots and Hierarchical Summary Receiver Operating Characteristics curves were used to describe the pooled summary estimates of sensitivity, specificity and diagnostic odds ratio. From a total of 1063 results, 30 studies were included in the qualitative synthesis, of which 19 were included in the meta-analysis. The pooled sensitivity and specificity were 0.808 (95%CI 0.656-0.903) and 0.920 (95%CI 0.769-0.975), respectively, for the NCITs (using forehead as the site of measurement), and 0.818 (95%CI 0.758-0.866) and 0.923 (95%CI 0.823-0.969), respectively, for thermal scanners. The sensitivity of NCITs increased on use of rectal temperature as the reference. The sensitivity of thermal scanners decreased in a disease outbreak/pandemic setting. Changes approaching statistical significance were also observed on the exclusion of neonates from the analysis. Thermal screening had a low positive predictive value, especially at the initial stage of an outbreak, whereas the negative predictive value (NPV) continued to be high even at later stages. Thermal screening has reasonable diagnostic accuracy in the detection of fever, although it may vary with changes in subject characteristics, setting, index test and the reference standard used. Thermal screening has a good NPV even during a pandemic. The policymakers must take into consideration the factors surrounding the screening strategy while forming ad-hoc guidelines.

Investigation of the Impact of Infrared Sensors on Core Body Temperature Monitoring by Comparing Measurement Sites

Many types of thermometers have been developed to measure body temperature. Infrared thermometers (IRT) are fast, convenient and ease to use. Two types of infrared thermometers are uses to measure body temperature: tympanic and forehead. With the spread of COVID-19 coronavirus, forehead temperature measurement is used widely to screen people for the illness. The performance of this type of device and the criteria for screening are worth studying. This study evaluated the performance of two types of tympanic infrared thermometers and an industrial infrared thermometer. The results showed that these infrared thermometers provide good precision. A fixed offset between tympanic and forehead temperature were found. The measurement values for wrist temperature show significant offsets with the tympanic temperature and cannot be used to screen fevers. The standard operating procedure (SOP) for the measurement of body temperature using an infrared thermometer was proposed. The suggestion threshold for the forehead temperature is 36 °C for screening of fever. The body temperature of a person who is possibly ill is then measured using a tympanic infrared thermometer for the purpose of a double check.

Comparison of Axillary and Temporal Artery Thermometry in Preterm Neonates

OBJECTIVE

To compare the accuracy of infrared temporal artery thermometry with axillary thermometry in a cohort of preterm neonates between 28 and 36 weeks postmenstrual age.

DESIGN

Descriptive repeated measures design with randomization to temperature measurement order.

SETTING

Level III NICU in the Central/Southeastern United States.

PARTICIPANTS

Sixty-eight neonates born between 28 weeks and 36 weeks postmenstrual age cared for in incubators or open cribs.

METHODS

Neonates were randomly assigned to temperature measurement order (axillary followed by temporal artery or temporal artery followed by axillary). Temperature pairs were taken once during the day shift and once during the night shift. Behavioral states were assessed before, during, and after temperature measurement.

RESULTS

Neonates were predominantly female (64.7%) with a mean age of 6.6 days and a mean gestational age of 32.7 weeks, and most were cared for in incubators (n = 55). Noninferiority was observed between the two temperature methods (Holm-Bonferroni criterion = .025, p < .001). There was no statistically significant difference in the behavioral states of the neonates between the two temperature methods. It took nurses significantly longer to use the axillary thermometer than to use the temporal artery thermometer (p < .001).

CONCLUSION

Temporal artery temperature measurements were as accurate as axillary temperature measurements in low-birth-weight neonates in the NICU. Nurses spent less time measuring with the temporal artery method than with the axillary method.

Association between Fever and Primary Tooth Eruption: A Systematic Review and Meta-analysis

Aim

To perform a systematic review and meta-analysis to establish if fever is associated with primary tooth eruption.

Materials and methods

Literature searches involved Pubmed, MEDLINE, Web of Science, Scopus and Cochrane. The potentially relevant studies had the full text analyzed. Only studies concerning fever during eruption period of primary tooth in humans were included. Papers in non-English language, and papers that included syndromic patients or patients with any disease were excluded. The meta-analyses were performed with Review Manager (version 5.3). Only studies that reported the results as dichotomous data were analyzed with Cochran-Mantel-Haenszel test in meta-analysis function of Review Manager 5.3. The fixed-effects model was used to evaluate the association between tooth eruption and fever.

Results

Search identified 83 potential studies. After exclusion of the duplicated studies, or were not related to the criteria of inclusion only 6 studies were selected for the systematic review. In the overall meta-analysis, no association was found [OR = 1.32 (0.88-1.96)] between fever and primary tooth eruption. However, in the subgroup analysis, when the method used to measure fever was the rectal temperature there was an association [OR = 2.82 (1.55-5.14)] between fever and primary tooth eruption.

Conclusion

There are few suitable studies in the literature regarding the association between primary tooth eruption and fever. However, our study found an association between fever and primary tooth eruption only when rectal temperature was performed

How to cite this article

Nemezio MA, De Oliveira KMH, Romualdo PC, Queiroz AM, Paula-e-Silva FWG, Silva RAB, Kuchler EC. Association between Fever and Primary Tooth Eruption: A Systematic Review and Meta-analysis. Int J Clin Pediatr Dent 2017;10(3):293-298.

Temporal Artery Temperature Measurement in the Neonate

Objective We compared an infrared temporal artery thermometer with our clinical standard axillary thermometer for temperature measurements in neonatal patients. Study Design We measured temporal artery (T_ta), axillary (T_ax, clinical standard), and rectal (T_r, gold standard) temperatures of 49 infants. The difference between T_r and T_ta was compared with that between T_r and T_ax, and the data were analyzed based on bed type and postmenstrual age. Results The mean T_ta, T_ax, and T_r were 37.16 (SD 0.36) °C, 36.61 (SD 0.30) °C, and 36.82 (SD 0.30) °C, respectively. The measurements by these methods were all significantly different. The mean T_r-T_ax was 0.21 (SD 0.26) °C, and the mean T_r-T_ta was -0.34 (SD 0.37) °C, indicating that T_ax was closer to T_r than was T_ta (p < 0.0001). T_ta agreed more closely with T_r for infants in cribs than for those in incubators. Adjusting for bed type and body weight, with each week of postmenstrual age, the discrepancy between T_r-T_ta and T_r-T_ax decreased by 0.005°C (p = 0.034). Conclusion Compared with the gold standard, T_r, T_ta is not more accurate than T_ax. The temporal artery thermometer was less accurate for infants in incubators than for infants in cribs. The accuracy of temporal artery temperature increased with postmenstrual age.

Comparison of Temporal Artery Thermometry with Axillary and Rectal Thermometry in Full Term Neonates

OBJECTIVE

To assess the efficacy of temporal artery thermometer in febrile and hypothermic neonates in comparison to axillary thermometer.

METHODS

It was a cross sectional observational study. Study participants included 210 neonates admitted in neonatal intensive care unit of a tertiary care teaching hospital, divided into three groups of 70 each, namely normothermic, febrile and hypothermic. Temperatures were measured using temporal artery, axillary and rectal thermometers in each patient.

RESULTS

Mean rectal temperature was found to be comparable to mean temporal artery temperature in normothermic babies. Temporal artery thermometer had a better sensitivity to diagnose fever, than hypothermia. Also, temporal artery temperature showed a good correlation with rectal temperature in normothermic and febrile group and not in hypothermic neonates.

CONCLUSIONS

Temporal artery thermometer can accurately detect temperature in febrile and normothermic fullterm neonates but not in hypothermic neonates. Further studies are required before advocating temporal artery thermometry as a replacement of rectal thermometry among this group of population.

Method Comparison (Agreement) Studies: Myths and Rationale

Unprecedented technological growth in the last quarter of twentieth century has resulted in improved health care and opened new domains of health care research. This technological leap also facilitated the paradigm shift from hospital care to home care through development of 'point of care' devices. As early diagnoses and timely referral is a key to health management, these devices play an important role in improving health. Validation of the new technology in different settings is necessary before adopting it to practice. For a binary result like pregnancy test, it is trivial to use statistical tools like sensitivity, specificity etc. For a continuous variable like blood glucose level the analysis is not straightforward. Many of us misinterpret 'association' as 'agreement'. This misinterpretation is reflected in studies which have compared two different technologies. The findings of well conducted studies do not contribute to the evidence base just because of wrong analysis strategy. We delineate on finer nuances of analysis and interpretation of method comparison studies.

Digital Axillary and Non-Contact Infrared Thermometers for Children

Axillary digital thermometers (ADTs) and non-contact (infrared) forehead thermometers (NCIFTs) are commonly used in pediatric settings, where an incorrect body temperature measurement may delay treatments or lead to incorrect diagnoses and therapies. Several studies comparing ADT or NCIFT with other methods have found conflicting results. To investigate whether ADT and NCIFT can be used interchangeably, a comparative observational study was conducted involving 205 children aged 0 to 14 years who were consecutively admitted to the pediatric emergency department. The Bland–Altman plot illustrated agreement between the two methods. A total of 217 pairs of measurements were compared; axillary measurements showed average values significantly higher than forehead measurements (37.52°C and 37.12°C; t = 7.42, p = .000), with a mean difference of 0.41°C between the two methods (range = −1.80 and +2.40). In this setting and population, ADT and NCIFT cannot be used interchangeably.

A practical comparison of temporal artery thermometry and axillary thermometry in neonates under different environments

AIM

Thermoregulatory stability and monitoring are crucial in neonatal care. However, the current standard of temperature measurement using Axillary Thermometry (AT) poses multiple limitations. Temporal Artery Thermometry (TT) is a promising new method, which thus begs the question: Can TT replace AT in neonates? Previous studies reveal conflicting results, with none involving a Southeast-Asian multi-ethnic neonatal population under different environments.

METHODS

A 6-month prospective comparative study involving neonates managed in a tertiary neonatal centre. Subjects were divided into 4 groups based on the required nursing environment: A) Room air B) Phototherapy C) Radiant warmers D) Incubators. Six hundred and sixty-one paired TT and AT temperature readings were obtained, with concurrent FLACC scoring to evaluate the discomfort associated with each thermometry method.

RESULTS

TT readings were higher than AT in all groups. The mean temperature difference between both methods (TT-AT) was lowest in Group A (0.10 ± 0.19°C), followed by Groups B (0.50 ± 0.33°C), C (0.97 ± 0.76°C) and D (1.15 ± 0.57°C) respectively. Bland-Altman analysis revealed good clinical agreement (± 0.5°C) between both methods in Group A (7-0.27,0.47). However, Groups B (-0.14,1.13), C (-0.51,2.45) and D (0.03,2.27) showed poor agreement. Multiple GEE analysis revealed Malay ethnicity to be an additional predictor of decreased TT-AT ( β = -0.13, p = 0.012). Compared to TT, AT was associated with higher discomfort levels (p <0.001).

CONCLUSIONS

Given the good agreement and increased comfort with TT use, our study confirms that TT is comparable to AT for neonates nursed in room air. TT is therefore recommended for the temperature screening and monitoring of neonates nursed on ambient air. Its use in other environments and factors predictive of comparability of both methods requires further study.

Performance of a non-contact infrared thermometer in healthy newborns

OBJECTIVE

To evaluate the performance of a non-contact infrared thermometer (NCIT) in comparison with digital axillary thermometer (DAT) and infrared tympanic thermometers (ITT) in a population of healthy at term and preterm newborns nursed in incubators.

SETTING

1 level III maternity hospital, and its intensive neonatal care unit.

PARTICIPANTS

119 healthy at term newborns and 70 preterm newborns nursed in incubators were consecutively enrolled. Exclusion criteria were unstable/critical conditions, polymalformative congenital syndromes and severe congenital syndromes.

INTERVENTIONS

Body temperature readings were prospectively collected. Each participant underwent bilateral axillary temperature measurement with DAT, bilateral tympanic measurement with ITT and mid-forehead temperature measurements using NCIT.

PRIMARY OUTCOME MEASURES

Degree of agreement between methods was evaluated by the Bland and Altman method.

RESULTS

714 measurements in 119 healthy at term newborns and 420 measurements in 70 preterm newborns nursed in incubators were performed. Clinical reproducibility of NCIT was 0.0455 °C for infants in incubators and 0.0861 °C for infants outside an incubator. Bias was 0.029 °C for infants in incubators and <0.0001 °C for infants outside an incubator. Zero outliers were recorded. The mean difference between methods was good both for newborns at term (0.12 °C for NCIT vs DAT and 0.02 °C for NCIT vs ITT) and preterm newborns in incubators (0.10 °C for NCIT vs DAT and 0.14 °C for NCIT vs ITT). Limits of agreement were 0.99 to -0.75 and 0.78 to -0.75 in at term newborns and were particularly satisfactory in preterm newborns in incubators (95% CI: 0.48 to -0.27 and 0.68 to -0.40).

CONCLUSIONS

Our results with Bland and Altman analysis demonstrate that NCIT is a very promising tool, especially in preterm newborns nursed in incubators.

TRIAL REGISTRATION

The study was approved by the Careggi University Hospital Ethics Committee (07/2011).

Measurement of body temperature in 300 dogs with a novel noncontact infrared thermometer on the cornea in comparison to a standard rectal digital thermometer

OBJECTIVE

To assess the accuracy of obtaining body temperatures in dogs with a noncontact infrared thermometer (NCIT) on the cornea compared with a rectal digital thermometer (RDT).

DESIGN

Prospective single center study.

SETTING

University teaching hospital.

ANIMALS

Three hundred dogs presented with low, normal, or high body temperatures.

INTERVENTIONS

Three body temperature readings were measured by an RDT and by an NCIT on the cornea of the left eye by 2 investigators (experienced and inexperienced). Results obtained by the 2 methods were compared.

MEASUREMENTS AND MAIN RESULTS

Median body temperature measured by the experienced investigator with the RDT and the NCIT were 38.3°C (range 35.5°C-41.1°C; 95% CI: 38.2-38.4°C) and 37.7°C (35.9°C-40.1°C; 95% CI: 37.7°C-37.9°C), respectively. Measurement of RDT as well as of NCIT correlated well between both investigators (rRDT = 0.94; rNCIT = 0.82; respectively, P < 0.001 for both methods). Mean RDT and NCIT-temperature correlated poorly (r = 0.43; P < 0.001) when taken by the experienced investigator and even less by the nonexperienced investigator (r = 0.38; P < 0.001). Repeatability of the NCIT revealed an unsatisfactory value (0.24°C) compared to RDT measurement (0.12°C). Agreement between both devices in measuring low, normal, and high values, calculated by Cohens-Kappa, was unsatisfactory (к = 0.201; P < 0.001). Calculating the receiver operating characteristic curve to determine the best threshold for fever (defined as RDT temperature >39.0°C) showed an area under the curve of 0.76. Mean discomfort score was significantly lower using NCIT compared to RDT measurement (P < 0.001).

CONCLUSIONS

There was poor agreement between body temperatures obtained by RDT and NCIT. The corneal NCIT measurement tends to underrecognize hypothermic and hyperthermic conditions. Although the use of the NCIT yields faster results and is significantly more comfortable for the dog than the RDT measurement, it cannot be recommended in dogs at this time.

A comparison of mid-forehead and axillary temperatures in newborn intensive care

OBJECTIVE

To evaluate accuracy of mid-forehead (MFH) thermometry compared with digital axilla (DAT) temperatures in infants in newborn intensive care.

STUDY DESIGN

A comparative study of MFH and DAT temperatures of newborn infants receiving tertiary-level intensive care. All admissions were considered and the following exclusion criteria applied: 'in extremis', hypoxic ischemic encephalopathy or non-English-speaking parents. Foot temperatures, infant and environmental variables were measured.

RESULT

In all, 783 readings were obtained in 100 infants with a birth weight range 515 to 4885 g (mean 2152 g). The between-person correlation was 0.30 (P < 0.001) and the within-person correlation was 0.52 (P < 0.001). Bland-Altman plots showed wide 95% confidence intervals in the differences between MFH and DAT measurements (-0.87 to 1.16 °C). Differences were affected by infant variables measured. MFH more accurately predicted DAT measurements in smaller neonates and were less accurate in neonates requiring Bubble Continuous Positive Airway Pressure (CPAP).

CONCLUSION

MFH thermometry is not able to replace DAT temperature recording in the newborn intensive care.

Temperature measurement in the preterm and term neonate: a review of the literature

The maintenance of a constant body temperature is important to all humans but even more so for newborn babies (neonates), especially those born pre-term. Because accurate measurement of body temperature is an important component of thermoregulation management in the neonate, a review of the literature was undertaken to determine the most appropriate method and site of temperature measurement in both the preterm and term neonate. The available evidence indicates that the axilla remains the most common place for temperature measurement.

Clinical accuracy of tympanic thermometer and noncontact infrared skin thermometer in pediatric practice: an alternative for axillary digital thermometer

INTRODUCTION

The aim of this study was to compare the body temperature measurements of infrared tympanic and forehead noncontact thermometers with the axillary digital thermometer.

METHODS

Randomly selected 50 pediatric patients who were hospitalized in Dr Behcet Uz Children's Training and Research Hospital, Pediatric Infectious Disease Unit, between March 2012 and September 2012 were included in the study. Body temperature measurements were performed using an axillary thermometer (Microlife MT 3001), a tympanic thermometer (Microlife Ear Thermometer IR 100), and a noncontact thermometer (ThermoFlash LX-26).

RESULTS

Fifty patients participated in this study. We performed 1639 temperature readings for every method. The average difference between the mean (SD) of both axillary and tympanic temperatures was -0.20°C (0.61°C) (95% confidence interval, -1.41°C to 1.00°C). The average difference between the mean (SD) of both axillary and forehead temperatures was -0.38 (0.55°C) (95% confidence interval, -1.47°C to 0.70°C). The Bland-Altman plot showed that most of the data points were tightly clustered around the zero line of the difference between the 2 temperature readings. With the use of the axillary method as the criterion standard, positive likelihood ratios were 17.9 and 16.5 and negative likelihood ratios were 0.2 and 0.4 for tympanic and forehead measurements, respectively.

DISCUSSION

The results demonstrated that the infrared tympanic thermometer could be a good option in the measurement of fever in the pediatric population. The noncontact infrared thermometer is very useful for the screening of fever in the pediatric population, but it must be used with caution because it has a high value of bias.

Comparison of three different thermometers in evaluating the body temperature of healthy young adult individuals

The aim of this study was to compare the measurement values obtained with a non-contact infrared thermometer, a tympanic thermometer and a chemical dot thermometer. The research population was composed of students studying in two departments of a university in Ankara. A total of 452 students who fit the inclusion criteria of the study and volunteered to participate were included in the sample. Body temperature measurements with different thermometers were performed by the same researcher at the same room temperature. Data were analyzed in a computerized environment by SPSS 15.0 statistical program pack and Bland-Altman graph. Mean age of healthy young adults participating in the study was 19.66 ± 0.94, and 55.1% of them were female. The agreement limits for non-contact infrared and chemical dot was between -1.30 and 0.32°C; for non-contact infrared and tympanic was between -1.26 and 0.13°C; and for chemical dot and tympanic -0.89 and 0.74°C. It was determined that, although the measurement values of the tympanic membrane and chemical dot thermometers conformed with each other, the conformity of the non-contact infrared thermometer was weak.

Accuracy of infrared thermometers in very low birth weight infants and impact on newborn behavioural states

AIM

To study the impact on newborn behavioural states and accuracy of three infrared thermometers compared with digital axillary thermometer measurements in very low birth weight infants.

METHODS

Single-centre prospective observational study. Preterm infants born <1500-g birth weight were eligible. Infants were observed for pre-measurement behaviour state using a five-point neonatal behaviour observation tool. One infrared temperature was taken from each of the devices, followed by an axillary measurement. Further behaviour-state observations were recorded following infrared and axillary measurements.

RESULTS

One hundred measurements were collected from each infrared device among a cohort of 42 very low birth weight infants. Only one infrared device showed satisfactory agreement with bias -0.071 (95% limits of agreement -0.68 to 0.54). The other two devices demonstrated poor agreement: bias -1.34; 95% limits of agreement -2.62 to -0.5 and bias -0.56; 95% limits of agreement -1.38 to 0.25. Neonatal behavioural scores showed only minimal changes when infrared measurements were performed but increased significantly following axillary measurements. The difference between the two modalities was statistically significant with a mean increase of 1.44 points following axillary measurements (95% confidence interval 1.21 to 1.67, P < 0.001).

CONCLUSIONS

Temperature measurements taken with infrared thermometers demonstrated less disruption to preterm infants' behavioural state, however accuracy of devices varied.

Comparison of temporal artery to rectal temperature measurements in children up to 24 months

This descriptive study compared temporal artery (TA) and rectal temperature measurements, patient comfort during temperature measurements, and nursing time required to obtain temperature measurements. Study participants (n = 40) included children 0-24 months old with fever higher than 38°C, admitted to a freestanding children's hospital in the Midwest. Statistical analysis of 450 paired TA and rectal temperature measurements revealed a 0.776 correlation, mean difference of 0.03°C, and 94.7% of measurements differing by less than 1.0°C. Patient comfort, measured via the Face, Legs, Activity, Cry, Consolability Scale, was enhanced with TA thermometry compared to rectal. TA thermometry resulted in an 87% savings of nursing time.

Measurement of body temperature in adult patients: comparative study of accuracy, reliability and validity of different devices

BACKGROUND AND AIMS

We compared a range of alternative devices with core body temperature measured at the pulmonary artery to identify the most valid and reliable instrument for measuring temperature in routine conditions in health services.

METHODS

201 patients from the intensive care unit of the Candelaria University Hospital, Canary Islands, admitted to hospital between April 2006 and July 2007. All patients (or their families) gave informed consent. Readings from gallium-in-glass, reactive strip and digital in axilla, infra-red ear and frontal thermometers were compared with the pulmonary artery core temperature simultaneously. External factors suspected of having an influence on the differences were explored. The cut-off point readings for each thermometer were fixed for the maximum negative predictive value in comparison with the core temperature. The validity, reliability, accuracy, external influence, the waste they generated, ease of use, speed, durability, security, comfort and cost of each thermometer was evaluated. An ad hoc overall valuation score was obtained from these parameters for each instrument.

RESULTS

For an error of ± 0.2°C and concordance with respect to fever, the gallium-in-glass thermometer gave the best results. The largest area under the receiver operating characteristic (ROC) curve is obtained by the digital axillar thermometer with probe (0.988 ± 0.007). The minimum difference between readings was given by the infrared ear thermometer, in comparison with the core temperature (-0.1 ± 0.3°C). Age, weight, level of conscience, male sex, environmental temperature and vaso-constrictor medication increases the difference in the readings and fever treatment reduces it, although this is not the same for all thermometers. The compact digital axillar thermometer and the digital thermometer with probe obtained the highest overall valuation score.

CONCLUSION

If we only evaluate the aspects of validity, reliability, accuracy and external influence, the best thermometer would be the gallium-in-glass after 12 min. The gallium-in-glass thermometer is less accurate after only 5 min in comparison with the reading taken after being placed for 12 min. If we add the evaluation of waste production, ease-of-use, speed, durability, security, patient comfort and costs, the thermometers that obtain the highest score are the compact digital and digital with probe in right axilla.