By Pat Anson, PNN Editor

You can monitor a lot of health conditions with an Apple Watch, everything from your heart rate and blood oxygen levels to fitness and sleep patterns. Researchers at Duke University have found the watch could also be useful in predicting pain levels in people with sickle cell disease (SCD).

In a small study recently published by JMIR Formative Research, Duke researchers used the watches to collect health data from 20 adults with SCD and used machine learning computer models to predict their pain scores.

SCD is a genetic disorder that causes red blood cells to form in a crescent or sickle shape, which creates unpredictable and painful blockages in blood vessels known as vaso-occlusive crises (VOCs). About 100,000 Americans live with SCD, primarily people of African or Hispanic descent.

Because VOC’s can lead to life-threatening infections, strokes and organ failure, knowing their intensity could lead to earlier treatment and save lives. VOCs are typically treated with pain medication and intravenous saline solutions to promote hydration.

The sickle cell patients in the study were all admitted Duke University’s SCD Day Hospital while experiencing a VOC and provided with an Apple Watch Series 3, which was worn for the duration of their visit. Data collected from the watch included their heart rate, heart rate variability and calorie consumption, which were then matched with pain scores and vital signs collected from their electronic medical records.

In all, a total of 15,683 data points were collected, which were then analyzed using three different machine learning techniques. The best performing one was the “random forest” model, which predicted pain scores with an accuracy of nearly 85 percent.

“The strong performance of the model in all metrics validates feasibility and the ability to use data collected from a noninvasive device, the Apple Watch, to predict the pain scores during VOCs,” wrote lead author Rebecca Sofia Stojancic, who works in the Sickle Cell Comprehensive Care Unit at Duke University Hospital. “It is a novel and feasible approach and presents a low-cost method that could benefit clinicians and individuals with sickle cell disease in the treatment of VOCs.”

The idea of using mobile health apps and wearable technology to predict pain scores isn’t a new one. The idea was first explored in 2019 by Duke researchers using a Microsoft Band 2 to collect data from sickle cell patients.

“The Microsoft Band 2 allowed easy collection of objective, physiologic markers during an acute pain crisis in adults with SCD. Features can be extracted from these data signals and matched with pain scores. Machine learning models can then use these features to feasibly predict patient pain scores,” researchers reported.

Could wearable devices be used someday to predict pain flares from other chronic health conditions? A handful of clinical studies have explored the use of smartphones, Fitbits and other devices to predict migraines and hospital readmissions for high-risk patients, but no results have been posted so far.