Does your health monitor have device bias? – . Health Blog

In recent years, the number and type of health monitoring devices available in smartphones and fitness apps has grown dramatically.

Your smartphone is likely to track the number of steps you take, how far and fast you walk, and how many stairs you climb each day. Some phones log your sleep, heart rate, how much energy you burn, and even "gait health" (how often are both feet on the ground? How even are your steps?). Of course, wearables and fitness equipment without a phone are also available, e.g. B. Devices to measure your heart rhythm, your blood pressure or your oxygen level. The accuracy of these devices varies – and in some cases, your skin tone can make a difference.

How accurate are health monitors in general?

I know from my experience with hospital monitors that they are not always accurate. False alarms from EKG monitors often have medical staff scurrying into patient rooms only to find the patient feels good and surprised at the excitement. A particularly common false positive is a dangerous and unstable heart rhythm on a continuous heart monitor that may be due to the movement of a patient brushing their teeth.

High-stakes devices with monitoring features like defibrillators and pacemakers are extensively tested by their manufacturers and verified by the FDA, so their accuracy and reliability are generally quite good.

But what about home health monitoring devices that are intended for consumption that have not been extensively tested by the FDA? Ever spend a few minutes counting your steps to see if your phone's match matches? Or climb a few flights of stairs to see if you get full credit for not taking the elevator?

The accuracy of consumer devices depends in part on what is being monitored. For example, one study assessed the accuracy of heart rate monitors and energy expenditure calculators in phones and health apps. The accuracy was quite high for heart rate (often in the 95% range) but much less accurate for energy use. Accuracy can also vary depending on who is being monitored.

Device Bias: What It Is and Why It Occurs

While no health gadget is perfect, some users get more reliable results than others. For example, if you wear nail polish, a pulse oximeter – a device that attaches to the tip of your finger to measure blood oxygen through your skin – may not work well because the polish interferes with the light sensor from working properly. There is a simple solution to this situation: remove the polish.

In other cases, however, the solution is not easy. We are increasingly realizing that certain medical devices are less accurate depending on a person's skin color, a phenomenon known as device bias.

• Pulse oximeter. Although they are generally considered to be very accurate in the healthcare sector and are widely used in the healthcare sector, their accuracy tends to be less in people of color. This is because the device relies on light shining through the skin to recognize the color of the blood, which varies based on the oxygen levels. The amount of pigment in the skin can change the way light behaves on its way to blood vessels and lead to inaccurate results. The FDA has issued a warning about this and other restrictions on the use of pulse oximeters.
• Measurement of bilirubin in newborns. Bilirubin is a breakdown product of red blood cells. Newborns are screened for high levels as this can cause permanent brain damage. If phototherapy (light treatment) is detected, it can help the baby get rid of the excess bilirubin, thereby preventing damage to the brain. Screening includes examining a newborn's skin and eyes for jaundice (a yellow coloration caused by increased bilirubin levels) and a light meter test to check for high bilirubin levels. However, the accuracy of this test is less in black newborns. This is especially important because jaundice is more difficult to detect in infants with darker skin and dangerously high bilirubin levels are more common in this population.
• Heart rate monitors in smartphones. According to at least one study, smartphone apps may also be less accurate with colored people. This is in turn due to the fact that the more skin pigments there are, the more stray light sensors detect pulsations in the blood flow that reflect the heartbeat.

Why device preload is important

Sometimes a measurement error has no immediate health consequences. An error rate of 5% to 10% in measuring heart rate may be of little concern. (In fact, one might be wondering why someone would need a heart rate monitor when they could just count your pulse for 15 seconds and multiply by 4!)

However, pulse oximeter readings are used to help decide if a person needs to be hospitalized, who needs to go to intensive care, and who needs additional testing. If the oxygen levels in people of color are consistently overestimated, the likelihood of under-treatment is higher than those of others whose readings are more accurate. And that can exacerbate pre-existing healthcare disparities.

These examples add to the growing list of prejudices in health care and other cases where disregarding different people has serious consequences. If you are using a health device it is reasonable to wonder if it has been tested on people like you. It is also reasonable to expect that people who develop medical and consumer health devices would expand the test subject demographics to ensure that the results are reliable for all users before they are released to market.

Sometimes a change in technology, e.g. For example, using a different type of light sensor can make health-related devices work more accurately for a larger group of people.

Or, there may not be a simple solution, and user characteristics need to be considered in properly interpreting the results. For example, a device could offer the user a selection of skin tones that match the skin color. Based on extensive data from previous tests on people with different skin colors, the device could adjust the results accordingly.

The final result

The urge to monitor our bodies, health and life experiences continues to gain momentum. Hence, we need to test and validate health-related devices to make sure they are suitable for different people before declaring them suitable for the general public. Even then, the device bias does not go away: bodies vary and technology has its limits. The key is knowing that it exists, repairing what can be repaired, and interpreting the results accordingly.

Follow me on Twitter @RobShmerling