• 
• 

Three Part Question

In [a 5 month old boy with a fever] how [accurate is tympanic thermometry] as [a measure of core body temperature]?

Clinical Scenario

A 5 month old boy attends the emergency department with a history of fever given by his mother. His temperature as taken with a tympanic thermometer is 37.5oC. His mother says he is hot to the touch. He has no focus for his fever on examination. The departmental protocol recommends a full septic screen in this age group if the temperature is above 38oC. You would like to know how accurate temperatures taken by this method are and whether you should check the temperature using another method.

Search Strategy

Original Research - SumSearch
SumSearch "temperature measurement ", "child", "fever" and filter "diagnosis"

Search Outcome

57 individual articles, 2 directly relevant. Secondary Sources - 0 Systematic Reviews - 1

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Duce SJ, 1996, UK	20 relevant research studies	Systematic review (1a)	Evaluation of infrared tympanic thermometry with a 8mm probe	Tympanic thermometry was found to be an inaccurate, inconsistent and insensitive method of core temperature measurement in neonates, infants and children.	Narrative Review
Lanham DM et al, 1999, USA	241 sets of temperature measurements from 178 paediatric patients (mean age 18.6 months)	Prospective cohort (4)	Rectal temperature of 38 oC compared to tympanic temperature	mean diff between rectal and tympanic measurements was -0.6 oC. Tympanic >37.4oc. Sensitivity 80% Specificity 85%. LR+ 5.3 LR- 0.23. PPV 87%. NPV 85%.	Age and presence of fever significantly affected the rectal-tympanic difference, which suggests that the tympanic method may not accurately measure temperature in younger, febrile children
Wilshaw R et al, 1999, USA	120 infants 59 with and 61 without fever	Case-control (4)	Rectal temperature of 38 oC compared to tympanic temperature	mean diff between rectal and tympanic measurements was +0.05 oC. Tympanic >38 oC. Sensitivity 100% Specificity 58%. LR+ 2.4 LR- 0.	Relationship between tympanic and rectal measurements was affected by age and sex

Comment(s)

The systematic review reported 44 studies addressing the use of different methods of temperature measurement including, axillary, sublingual, tympanic and rectal. Two further studies directly addressed the question of how representative tympanic measurements are of core temperature measurement. Both studies showed a correlation between tympanic and rectal methods of temperature measurement although it was not strong enough to use as a basis to make decisions regarding clinical management.

Clinical Bottom Line

The diagnosis of fever without a focus should not be made based on tympanic thermometry as it is not an accurate measurement of core temperature. Rectal temperature measurement remains the clinical gold standard for diagnosis of fever in infants and children.

References

1. Duce SJ. A systematic review of the literature to determine optimal methods of temperature measurement in neonates, infants and children. DARE Review 1996;4:1-124.
2. Lanham DM, Walker B, Klocke E, Jennings M. Accuracy of tympanic temperature readings in children under 6 years of age. Pediatric Nursing 1999;25(1):39-42.
3. Wilshaw R, Beckstrand R, Waid D, Schaalje GB. A comparison of the use of tympanic, axillary and rectal thermometers in infants. Journal of Pediatric Nursing 1999;14(2):88-93.