Better ways to use temperature to predict foot ulcers

Prevention of foot ulcers is an enormous challenge, with great potential in reducing the burden of diabetic foot disease. Following three very successful trials in the US, home monitoring of foot temperature has been advocated as a method to predict and prevent foot ulcers. However, this therapy is hardly used in daily clinical practice. This lack of use is probably caused by doubts concerning the working mechanism of temperature measurements, as well as the burden of doing these measurements on a daily basis using equipment that is not very user-friendly. New American research, however, shows ways forward to counter both problems.

Frykberg and colleagues tested a temperature sensing mat. All patients had to do is stand on the mat for 20 seconds, after which all data gets transferred automatically and securely to healthcare providers. This greatly solves the burden of doing foot temperature measurements and is a very promising development.

The study was diagnostic, so patients did not get feedback on their temperature measurements, and were not asked to change their behavior to prevent foot ulcers. This is an important difference with the clinical trials, and is a strong point of this study. By doing so, we can gain valuable information on the diagnostic quality of temperature measurements.

What the authors founds was a high sensitivity, but low specificity for two asymmetry thresholds (i.e., a spot on foot being 2.2 or 2.75 degrees warmer than its contralateral counterpart), and vice versa for thresholds above 3 degrees. These results are similar to other studies on temperature monitoring. What they mean is that a foot with symmetrical temperature is healthy, but predictions for a foot with asymmetry cannot be accurately made. While almost all ulcers will be preceded by an increase in temperature, an increase in temperature is also seen without an ulcer being developed.

Some words of caution are needed with this article. The authors describe an average temperature asymmetry of 3.10 degrees (2.81 in those without a foot ulcer). With such a high average (seeing the 2.2 degrees threshold), one would expect high rates of false positives. However, the authors describe that patients would only receive an average of 3 warning notifications per year when the current threshold is used. It is unclear from the article how both were calculated, but this might be the result of the use of ‘two month’ windows to calculate sensitivity and specificity. Such a window is less reflective of daily clinical practice, where patients may be asked to change their behavior as soon as the threshold is exceeded. As this was a purely diagnostic study, no such clinical information is available.

Another aspect is the involvement of the company who makes the temperature mat. Three authors on the publication work for that company, and they have been heavily involved in data analysis. The company has a clear conflict of interest with the positive results, and different analyses might have resulted in different outcomes. It is therefore important that this result is replicated in another setting, in a study where the company cannot influence data analysis.

In conclusion, this study shows promising developments, describing a better and more user-friendly system to implement temperature measurements for ulcer prevention in daily clinical practice. The findings on its diagnostics confirm previous studies, showing that feet with symmetrical temperature will not develop foot ulcers. However, for an accurate prediction of foot ulceration when a temperature difference is found, we still need to improve (and individualize) our thresholds.