The recent spread of coronavirus has to led to an increased interest in using infrared thermometers to scan for fevers in people. While scanning the forehead or ear canal of a person can yield a relatively accurate reading of their body temperature when using a properly calibrated medical infrared thermometer, these instruments should only be used for preliminary scanning and not for medical diagnosis. Studies [1,2] have shown poor accuracy of many infrared thermometers when used for diagnosing fever, even in children whose temperatures are typically easier to read with this technology.
Skin Temperature is not Body Temperature
The reason is that skin temperature is only loosely correlated with core body temperature. Typical human forehead skin temperature can easily range form the mid 80F’s to the mid 90F’s. Ambient air temperature and the activity level of a person may increase or decrease their skin temperature by several degrees, even if the person is healthy and has a stable and normal core body temperature.
The areas of the body that have a fairly stable temperature and are accessible to an infrared thermometer are the area of the temporal artery of the forehead and the ear canal. And even those fluctuate in temperature quite a bit in adults, and can only be relied on in young children for getting a useful measurement result.
In short, the use of infrared thermometers for body temperature measurements should be limited to preliminary scanning to identify potential outliers within a group of people or within a specific environment. For example, every person walking into the front of an office building during the same outdoor weather conditions, or everybody in a specific location with the same environmental conditions. This is because skin temperature can fluctuate greatly due to environmental conditions which will have a direct impact on the extrapolated body temperature reading of a medical infrared thermometer.
Medical Infrared Thermometers
That being said, when using an infrared thermometer to measure skin temperature with the intent to determine body temperature one should use an infrared thermometer designed and approved for medical use. These body temperature infrared thermometers have a limited temperature range that only covers a few degrees above and below the normal body temperature (e.g. 32..43°C/90..110°F), and because of that they can achieve higher accuracy in that limited range (e.g. +/- 0.3°C).
Many models automatically extrapolate body temperature from the skin surface temperature readings using a built-in formula. Readings are typically taken in the area of the temporal artery of the forehead, or in the ear canal.
Other Infrared Thermometers
The other types of infrared thermometers, including the ennoLogic infrared thermometers eT650D and eT1050D, are designed to cover a much wider temperature range, typically several hundred degrees. These IR thermometers are used in home, cooking, food safety, HVAC, automotive and many industrial applications. Their accuracy is typically +/- 2°C/3.5°F in the range of room and body temperature.
This wide tolerance makes these types of infrared thermometers less useful for determining body temperature by scanning a person’s skin. However, these thermometers may be used for preliminary scanning of skin temperature similarly to how they are used in food safety applications, so long as proper procedures are followed and the limitations of the instrument are understood.
Human skin is, in fact, a surface that is fairly easy to measure with infrared thermometers. It has an emissivity of about 0.98, is non-reflective when dry, and readings of skin temperature with any infrared thermometer produce fairly repeatable results. How these results correlate to body temperature is an entirely separate issue (see above).
Even though an industrial IR thermometer cannot produce precise measurements (due to its ± 2°C/3.5°F tolerance specification), it can be used to produce repeatable results and used for scanning for potential outliers. In other words, a sampling of healthy subjects could be used to determine a baseline value, e.g. forehead temperature of 33°C(91.4°F), and scanning for outliers could then be performed based on an agreed difference. (The values chosen here are arbitrary and only used for illustration of the principle of scanning for outliers based on relative values that are not necessarily accurate absolutes.)
Detection Requires Confirmation
Whenever you are using an infrared thermometer to take non-contact temperatures of people, even when using a medical infrared thermometer, the reading you get shpould only be seen as an indicator and requires a contact measurement to confirm it. So, the first step after identification of an outlier should be the confirmation of the infrared thermometer reading with a medical contact thermometer.
Using this method, it needs to be understood that each individual infrared thermometer will have its own unique measurement error. So the baseline value of healthy subjects might be x°F for one thermometer, and y°F for another thermometer – both used in the exact same application. Each user must be familiar with the ‘normal’ readings of his/her unit. So two units may be working side by side and reading slightly higher or lower than each other, but are providing accurate readings relative to their own measurements.
Implementation should essentially follow the same protocol used in the food safety industry, i.e. all items in a shipment are scanned on arrival and individual items which fall outside a predetermined specific temperature range are identified, pulled and further inspected and tested with a direct contact thermometer to verify if they are actually outside of the safety range.
By delivering non-contact surface temperatures, infrared thermometers can provide critical data without the contamination risks associated with contact thermometer measurements, so long as they are used with the correct knowledge and understanding of their limitations.
Factors that can Impact Non-Contact Infrared Thermometer Readings
The sensor of an infrared thermometer is quite reliable, and should deliver accurate readings without drift over time, as long as the instrument is not mishandled or abused.
Factors that can affect the accuracy of the reading include direct contact with steam. Readings generally return to normal once the unit has had time to dry completely. Likewise, smoke, or dirt in the air, exposure to gas fumes and other volatile fumes can all affect the operation of the sensor.
Other factors that could potentially impact accuracy include extreme cold or hot ambient temperatures. For that reason, always bring the unit to room temperature before using it. The digital IR thermometer is an electronic instrument, so of course it can be damaged by severe rough handling, submersion in water, etc.
An infrared thermometer designed for home and industrial applications is not approved for medical use. Its accuracy tolerance is typically ± 2°C/3.5°F and it covers a temperature range of several hundred degrees. This needs to be kept in mind when considering the use of an industrial infrared thermometer any time it’s being used to measure skin temperature.
A body infrared thermometer has an accuracy tolerance of ± 0.3°C/0.6°F and a temperature range of only about 10°C/20°F.
A high-quality medical body temperature infrared thermometer should ideally be FDA-certified for medical use. High-quality infrared thermometers for home and industrial use are only FDA-certified for laser safety.
All non-contact surface temperature readings which are critical to human health or safety should always be confirmed through appropriate contact thermometer temperature readings in order to make a final diagnosis.
 Infrared ear thermometry compared with rectal thermometry in children: a systematic review. Lancet. 2002 Aug 24;360(9333):603-9.
 In a systematic review, infrared ear thermometry for fever diagnosis in children finds poor sensitivity. Journal of Clinical Epidemiology, Vol. 59, No. 4, 2006, p. 354-357.