SIA Asks FDA For Crackdown on Non-Compliant Temperature Systems
By Conor Healy, Published Nov 03, 2020, 03:05pm ESTIPVM worked with the Security Industry Association (SIA) on a letter sent to the FDA calling on them to take action "protecting public health and safety by enforcing the guidelines."
This 1-minute video explains the issues involved, and why a crackdown is needed:
SIA Letter Overview
The letter calls out solutions not following FDA screening guidelines, and those that sell them:
I am writing to bring attention to the issue of some companies and solutions that are not following FDA guidelines and the standards within while claiming to detect elevated human temperatures for the purpose of mitigating the spread of disease. These non-compliant systems have the potential to endanger public safety. [emphasis added].
Further stating:
Some products introduced during [the pandemic] are missing the mark when it comes to enhancing public safety. These failures come as a result of not following the guidance.
In particular, the letter identifies lacking enforcement of guidelines by the FDA as a cause:
Without enforcement of the FDA guidelines, the result has been confusion in the market which can lead to a false sense of public safety.
It further states concern over the risk to the industry at-large:
The proliferation of solutions that cannot function as marketed also negatively affects makers, resellers and users of solutions that do follow the guidelines.
Wall Street Journal Report on SIA Letter to the FDA
On November 3, 2020, the Wall Street Journal covered SIA's letter in a piece exploring problems with the market for temperature screening, in particular, the consequences of the FDA's decision to not enforce guidelines for these medical devices during the pandemic:
The move prompted many new sellers, including many with little or no experience with medical devices. Many of these sellers rely on inexpensive imports from China that they rebrand, sometimes using misleading marketing materials to target school districts, airports, hospitals and businesses at a steep markup, longtime vendors and scientists say.
The piece also covered the $1 million purchase of Hikvision thermal cameras without blackbodies and intending to screen 30 per second, reported in an investigation by IPVM.
IPVM Research on Thermal Cameras
In the past 6 months, IPVM has tested 26 of these systems.
Guidelines for Usage: FDA Gives Guidance on 'Coronavirus' Thermal Fever Detection Systems, and World IEC Fever Screening Standards Explained.
Comments (7)
It's interesting that almost none of those FDA cleared thermal cameras can actually meet the specifications in the FDA guideline!
#1, Which devices and which guidelines do you believe they do not meet?
FDA guideline published on April 2020 states "Accuracy...is less than or equal to +/-0.5C" and "Both stability and drift are less than 0.2C". I didn't see these were met by any system on the market.
Virtually everyone is claiming that. To be clear I am not saying that they all do but can you cite specific examples where the published specs are inferior to the guidelines?
There are many many. 2 examples below for your reference. You can find more if you do a little research.
Flir E54-EST, "Accuracy [Drift] in Screen-EST Mode±0.3°C (±0.5°F)". Drift is over 0.2C. No stability data. And no blackbody support.
Seekscan, there is no spec about stability or drift while it claims "Meets FDA guidelines" on their datasheet. Lower resolution than IEC spec.
And no blackbody support.
FLIR previously commented on blackbody vs no blackbody usage, copied below:
FLIR cameras are engineered and calibrated with automatic ambient drift compensation that eliminate the need for a black body reference. Reasons for this include:
- Camera calibration is part of the production process and is based on multiple high-end black bodies sources.
- The cameras integrate internal temperature sensors that compensate for a possible calibration shift.
- A shutter between the camera detector and the lens is used as a reference to perform non-uniformity corrections when the camera environment changes.
This proprietary mix of technologies ensures the thermal camera measurements remain stable and constant meeting
the standard set forth in IEC 80601-2-59:2017, which states thermal cameras “may use SELF-CORRECTIONS to maintain the drift within acceptable limits… to allow for substitution of the CALIBRATION SOURCE”.
As for:
Drift is over 0.2C. No stability data.
I'll ask FLIR to comment. FLIR does have FDA 510(k) clearance for specific models.
Seekscan, there is no spec about stability or drift
I think systems should provide specs on stability and drift but even if they do (and let's say they specify within the FDA guidelines), the much bigger practical problem is the actual performance issues and the terribly bad deployments.
Lower resolution than IEC spec.
Yes but that sensor is comparably huge compared to ompanies selling truckloads of temperature tablets using 32 x 32 sensors or less. I agree resolution should be a factor but we've actually tested a wide variety of systems and Seek Scan, relative to the vast array of rushed out systems this year, performed far more accurately and far less prone to missing elevated temperatures.
Net/net, I agree with your emphasis on complying to guidelines but the ones you are picking are not the ones that post the highest risk to public safety (i.e., that would be bad deployments, marketing bad deployments, rigged systems, super low res systems, etc.). Agree/disagree?
the standard set forth in IEC 80601-2-59:2017, which states thermal cameras “may use SELF-CORRECTIONS to maintain the drift within acceptable limits… to allow for substitution of the CALIBRATION SOURCE”.
wow, I don't know who is making this statement, but this is missing key text, and absolutely does not mean what is stated in that sentence:
original text from iec:
A screening thermograph may use self-corrections to maintain the drift within acceptable limits. An automated service test mode may be utilized to perform these measurements to allow for the substitution of the CALIBRATION source for a FACE
calibration source is mandatory for iec, and there is no exemption in the iec standard to avoid using it.
also mentioning that a shutter between sensor and lens is enough to achieve better than 0.2 drift with enviromental temperature changes sounds pretty optimistc, as it does not allow to compensate the temperature changes of the optics....
