Evidence limited for predicting, monitoring COPD exacerbations via wearable devices

Key takeaways:

• Available studies on wearable devices and COPD exacerbations have differing data collection methods.
• Patient adherence to a device and app should be considered in testing.

PHILADELPHIA — Despite the growing prevalence of wearable devices, more research is needed before they can be used to predict and monitor COPD exacerbations, according to a presentation at the 2024 GOLD COPD International Conference.

“An output from PubMed indicates the rapid rise of publications in the field, with more than 8,000 papers on the topic of wearables published recently to date, and a similar trajectory is evident, albeit with much lower numbers, specifically in people with COPD,” Narelle Cox, PhD, associate professor and researcher in physiotherapy and respiratory research at Monash University, said during her presentation.

In addition to predicting and monitoring exacerbations, Cox highlighted five other factors that wearables could potentially measure in patients with COPD: physical activity and exercise, symptom monitoring, quality of life, physiological monitoring and medication adherence.

Using findings from a systematic review and meta-analysis published in npj Digital Medicine that assessed 37 studies focusing on wearable technology in adults with COPD, Cox outlined how this technology has been found to impact this patient population.

According to Cox, both daily step count and 6-minute walk distance significantly went up with use of wearables.

“It’s worth noting that the amount of these improvements did not exceed the minimum clinically important difference in the case of 6-minute walk distance, and was by an amount that may or may not achieve clinically meaningful improvements in step count,” Cox said. “As well, the impact was very short lived.”

When evaluating the use of wearables for COPD exacerbation avoidance and prediction (10 studies; n = 3,660; mean age, 69 years; 69% men), results from the meta-analysis were mixed, Cox said.

Of these 10 studies, three analyzed exacerbation prediction. When taking a closer look at the three studies, Cox highlighted some potential limitations including small sample size, use of bespoke modelling and high rate of patient dropout. Additionally, each study used a different timeframe for collecting data (once daily vs. overnight, continuous monitoring, data transfer every 1 to 3 days), and factors for prediction were unclear when looking at their findings together.

“Collectively, this appears to tell us that we don’t yet have confirmation on the best way to collect data, the best frequency of data collection or which factors are most important for predicting exacerbation risk,” Cox said.

Based on the current data on use of wearables for exacerbation prediction, Cox noted limited proof of “effect, benefit and usefulness.”

Another important factor surrounding wearable technology in this patient population is adherence. To showcase this factor, Cox highlighted a 6-month pilot study of a wearable COPD self-management app published in BMC Medical Informatics and Decision Making, in which patients (n = 34) were given a smartphone and smartwatch. In this study, researchers observed that as time progressed, use of the app went down.

“Active users in this study were defined as those who used the app at least 50% of their study enrollment days,” Cox said.

“There was never 100% engagement with the app, and the proportion of active users started to decline after month one,” she continued. “Even amongst the active users, the app usage continued to decline over the 6-month period, starting at around twice a day at 9 or so minutes in the first month and declining down to just once a day for around 3 minutes at month six.”

Notably, Cox pointed out that high proportions of patients reported having WiFi at home (88.2%) and use of a computer at home (88.2%), whereas smaller proportions of patients reported current or ever smartphone use (64.7%) and current or ever smartwatch/wearable use (35.3%).

“At present, the widespread availability and use of such devices perhaps seems to be slightly less relevant to our patient population than what it is to the general public,” Cox said. “Whether this can change over the next decade or so, as more native digital users become the patient population in question, only time will tell.”

Other factors Cox noted to remember as wearables become more popular in health care include cost and data security and privacy.

Additionally, clinicians and researchers should keep in mind that data collected during wearable device use will need to be interpreted.

To illustrate this point, Cox showed that the data output from a wearable device used by one patient totaled more than 37,000 line items of data.

“These data needed cleaning and then analyzing in order for us to, first of all, make sense of it and then be able to do anything clinically meaningful with it,” Cox said.

References:

• Shah AJ, et al. npj Digit Med. 2023;doi:10.1038/s41746-023-00962-0.
• Wu R, et al. BMC Med Inform Decis Mak. 2024;doi:10.1186/s12911-024-02461-y.