Increasing Hand Hygiene Adherence with Thermal Imaging

Hand hygiene is the first and most important defense against health care–associated infections, but all too often, health care workers (HCWs) do not follow their facility’s guidelines. From apps on a smart phone to infection preventionists on the floor with pen and paper watching the staff, countless ways have been offered as ideas on how to improve hand hygiene.

Results from a recent pilot study,¹ published in the American Journal of Infection Control, indicate that portable thermal imaging cameras could offer a new approach to assessing and improving hand hygiene practices among health care professionals (HCPs). John Boyce, MD, a renowned infectious disease specialist and hospital epidemiologist, with Richard A. Martinello, MD, medical director of infection prevention, Yale School of Medicine, devised a study on using portable thermal imaging to assess how well HCPs perform hand hygiene.

Boyce, president at J.M. Boyce Consulting LLC, spoke with Infection Control Today^® (ICT^®) on what the study comprised, how it came about, and its limitations, as well as future research plans.

ICT^®: What are the key findings of this study and why they are important?

John Boyce, MD: What I’d like to do is to start by mentioning that a lot of efforts have been made over the last few years to improve the [adherence] of health care personnel with recommended hand hygiene practices. But there has not been as much attention [paid] to how people apply the alcohol-based hand sanitizer to their hands, so-called hand hygiene technique. And because there hasn’t been as much attention devoted to that, right now, as far as I know, there’s no standardized method for hospitals to assess hand hygiene technique. So I was looking around thinking about how one could go about assessing or monitoring hand hygiene technique. And I saw an article in The New York Times about thermal imaging. I started to read a little bit about thermal imaging applications in the medical field and I ran across an article that mentioned that if you apply something like an ultrasound gel to the skin, …it cools the skin right where the gel has been applied and you could probably detect that cooling…with infrared thermal imaging. So that’s how I got the idea. Because [of] my involvement in hand hygiene over the last 20 years, I’m familiar with alcohol-based hand sanitizers. And I thought, maybe we could use thermal imaging to determine whether people have applied the sanitizer to all surfaces [on] their hands and fingers. And one of the reasons I focus on those 2 areas, …especially the fingers and the thumbs, is that those are areas that health care personnel often inadvertently miss when they’re applying a hand sanitizer. I thoughtmaybe we could look at thermal imaging [and] see if it would work in terms of assessing how well people are covering their hands and fingers. With a colleague of mine, Dr. Richard Martinello, a fellow at Yale New Haven Hospital, we did a small exploratory proof of concept study to see if that approach might work.

…This was a very small study with only 12 health care personnel…. For each person we measured and estimated their hand surface area. That’s the area over which you need to apply…sanitizer and of course some personnel have small hands, some medium, and some large hands. We estimated their hand surface area using recognized methods and then we took a thermal camera attached to an iPhone and took an image of their dominant hand, the palmar aspect of their dominant hand. Then we applied about 1.8 mLs of the hospital’s handset sanitizer and told them to rub their hands together until they feel dry. And then we immediately took a follow-up thermal image…after hand hygiene, and then…follow-up images 1 minute and 2 minutes later. Then we uploaded those thermal images to a computer for image analysis. We could also do temperature readings, which the thermal imaging allows you to do.

[One] of the findings we…noticed was, first, there is a real variability in the temperature of people’s hands. It turns out that several of the volunteers had cold hands and this was confirmed by their appearance on the thermal image. The app that we use with this thermal camera shows your hand as beinga red-brown color, if it’s warm. The cooler the parts of your skin are, the more the colors change to yellow, green, blue, or magenta. One of the volunteers’ hands was completelyred-brown with a little bit of white on his palm, which is the warmest area. [With] one or a couple of the volunteers, you could see that their fingers were not red-brown, they were more like yellow or green or magenta color, meaning that they’re cooler than their palm. So that was one thing we noticed, the baseline variability and hand temperatures, which is not really a new finding but it was new to us.

Then we found that for 11 of the 12 volunteers, you could clearly see on these visual thermal images that the temperature of their hands and fingers decreased significantly immediately after they finished doing alcohol and sanitizing because all the colors of their fingers and thumb changed. Then we noticed that on our 1-minute-later and 2-minute-later images, those cool temperatures on the fingers had started to gradually disappear. By 2 minutes afterward, much of the color had reverted toward baseline color, so the color changes don’t last very long. One of the findings was that if one wants to try to use this method to assess how well people are covering their hands with the sanitizer, you will need to probably do the image fairly quickly after they finish rubbing their hands together. If you waited 2 minutes, it would not be a very accurate assessment.

We noticed when we had one person that we told to purposely not rub their thumbs when they were doing hand hygiene [that] when we took his thermal image afterward, all the fingers had that yellow-green-blue color denoting lower temperatures and the thumb was still a brown color. It was very easy to tell that he had not covered his hands. Another finding was that 4 of the 12 volunteers didn’t have very much color change, or very much temperature change on the tip of their third finger, which is where we measured temperatures. …We measured temperatures in the middle of the palm at the end of the third finger and at the end of the thumb. In fact, the fingertips and thumbs are the areas that people often miss. A couple of our volunteers didn’t do very well on their fingertips and the other ones did fine. These things suggest that the images themselves, the color images, may be useful for assessing hand hygiene technique and the temperature readings are not necessary for teaching or monitoring hand hygiene. But it gives us a new picture of alcohol-based hand sanitizing. By having these temperature data, we noticed that people who had small or medium-sized hands, not surprisingly, had lower temperatures than people with big hands.

Another thing that we did was unique because we applied the same amount of sanitizer to each volunteer’s hands. …Because their hand surface area varies, some of them small, some of them large, what we decided to do would be to express the amount of volume applied as the amount per hand service area, [such as] milliliters per centimeter squared of hand surface area. What we found was that people with small hands had a higher number of milliliters per centimeter squared than people with big hands, because you’re applying the same amount of alcohol to a different surface area. We found a direct negative correlation between hand size and the milliliters per centimeter squared applied, which was something that, as far as I know, nobody else has looked at in that way.

ICT^®: I have small hands. I always end up getting 2 doses of sanitizer and then having too much. Is that OK or is there a problem with having too much?

JB: No, I don’t think you can apply too much sanitizer. But because HCWs are so busy, some of them have to do hand sanitizing dozens of times in a 12-hour shift. I published [findings from] one study, in conjunction with some people at Iowa, that showed that sometimes in an intensive care unit some nurses may have an indication to clean their hands 100 times in 12 hours. They’re in a hurry. Most HCWs want to hurry up and finish rubbing their hands again; they don’t want to take too long because they're in a hurry to get going and do their job. They tend to favor small volumes. In fact, some of them favor volumes that are probably too small…for a person with smaller hands or medium-sized hands. More of an issue is getting too little sanitizer on your hands. For a person with bigger hands, with a standard dispenser, it would be desirable for them to get 2 doses. For you, if you have small hands and you get 2 doses, it’s going to take you 30 to 45 seconds of rubbing [to] feel dry or you’re going to be dripping some of it on the floor or something.

In addition to not being able to assess coverage underneath the nails by our thermal imaging study, some of the other important limitations we had…, which are worth noting, is a very small sample size, only 12 people; [the study] was done in 1 room; we only used 1 brand of hand sanitizer gel; and people were a captive audiencein the room with us. And I mentioned that we didn’t assess the microbiologic efficacy. One thing that’s very clear [that] people [should] understand…if they see this interview or read the paper, is…that this was really a proof-of-concept study, and we haven’t shown that this technique is practical yet. But based on our findings, we think it really warrants a much larger study, with many more HCWs involved. Maybe in different times of the year, when it’s cold outside or not cold outside, it would be a good idea to try it with perhaps other brands of hand sanitizer, which could…conceivably yield slightly different results.

Another important thing…[is that] we’re hoping to do a follow-up study [and] use a smaller volume than what we [just used]. We delivered 1.8 mL from a hand pump of sanitizer, but many wall-mounted dispensers deliver more like 1.1 or 1.2 mL. So it’s very important that additional studies look at that smaller volume and see if thermal imaging still gives a reasonable assessment of coverage. [However,] if we’re able to do a follow-up study, we’re going to record the duration of hand rubbing and then see how that correlates with hand coverage assessed by the thermal images.

There are several potential applications [for these findings]: one would be teaching, [another] would be competency evaluations. The thing that would be the most challenging would be to use it for monitoring hand hygiene on the wards.

ICT^®: What results of the study surprised you?

The variation and baseline and temperatures really caught me off guard, although I personally have cold hands. So maybe it shouldn’t surprise me, but I thought I was a rare person. Maybe that’s the thing that surprised me the most. I wasn’t sure what to expect regarding the temperature readings because I hadn’t read any papers before where people had looked at temperature changes with alcohol sanitizer. So with the temperatures that we got, I didn’t really know quite what to expect.

ICT®: Would you like to add anything else?

JB: This whole pilot project, which again, is a very smallpilot, is designed to see if we can improve hand hygiene technique along with general compliance rates, because we all want to improve hand hygiene for the benefit of patient care. So that’s our long-term goal.

Note: This interview has been edited for length and clarity. See the video to watch the entire interview.

Reference

Boyce JM, Martinello RA. Pilot study of using thermal imaging to assess hand hygiene technique. Am J Infect Control. Published online September 14, 2022. doi:10.1016/j.ajic.2022.07.015