KardiaBand is more accurate than Apple Watch 4 in diagnosing AF

The difference likely lies in how the software interprets the EKG tracer. How busy clinicians operate based on the information is unclear.

Wearable devices promise the ability to instantly alert patients about potentially dangerous changes in heart rhythm, but new data from the SMART WARS study suggests that not all technologies are created equal.

Led by Christopher Ford, MD (Eastern Health Clinical School, Box Hill, Australia), researchers found that KardiaBand (AliveCor) outperformed the Apple Watch 4 (Apple) when it came to accuracy and sensitivity for detecting atrial fibrillation (AF) in a group of 125 patients tested with both devices. The differences do not stem from the ability to read the electrocardiogram, they say, but rather appear to result from the proprietary algorithms used to interpret the information.

Their findings were recently published online in JACC: Clinical Electrophysiology.

“As cardiologists in the current era of readily available smart devices, the data that patients collect from these devices is not frequently presented, but we do not have objective scientific studies to guide us on its accuracy or how to interpret the data, and whether Teh, senior author Andrew W.T., A MBBS (Eastern Health Clinical School and Austin Hospital Clinical School, Melbourne, Australia), he told TCTMD:

While the idea of ​​using these devices as diagnosing aids for atrial fibrillation is attractive, their accuracy requires objective validation, he noted in an email.

Teh said that while they were studying, he would have expected to see similar performances by the two producers they auditioned for. The Apple Watch 4 and KardiaBand, an accessory that comes with the Apple Watch, are both capable of taking single-lead ECG recordings. Their manufacturers software analyzes these automatic cadence ratings, and reports their results to users.

Exactly what caused the difference, he said, was a “complex question”. But given the fact that ECG tracings were obtained in a consistent manner, and “because the cardiologists who reviewed the tracers from both devices were able to accurately diagnose atrial fibrillation and normal (sinus rhythm) rhythm, the software algorithm used by each A device that most likely appears to be the hallmark that caused the Apple Watch’s low resolution.”

Kalyanam Shivkumar, MD, PhD (University of California, Los Angeles), Editor-in-Chief JACC: Clinical Electrophysiology, said the magazine was drawn to the SMART WARS study because its topic is so timely and “of great interest to everyone”. Prior to comparing the two devices, there was no reason to believe that one performed better than the other, he commented to TCTMD. “We had no entry bias.”

Shivkumar stressed that it is important to know, regardless of the specific brands, whether these “over-the-counter medical technologies in the diagnostic space” provide reliable information. “More and more of these devices are going to be used, so shall we get complete garbage?” he asked, adding that this data shows it’s “not rubbish.”

Shivkumar said the KardiaBand’s automated algorithm has already outperformed that of the Apple Watch 4. “This is a useful point to keep in the back of our minds.”

He noted that “the joy of what’s going to happen in the future” is seeing how artificial intelligence, machine learning, and other approaches are informing innovation in the field. Automated rhythm detection algorithms are a work in progress.

Possibility of missing diagnoses

For the SMART WARS study, researchers took consecutive outpatient recordings from 125 patients (median age 76 years; 62% men) using an Apple Watch 4 and KardiaBand plus 12 ECG leads. They analyzed both instrumental diagnoses and two cardiologists’ blind interpretation of wearable device findings.

The diagnosis of atrial fibrillation was confirmed in 27 patients (24 persistent and three paroxysmal) using 12-ECG, and atrial flutter in four patients, all of whom had previous diagnoses. The rest of the patients were in sinus rhythm.

When assessing all recordings made by wearable devices, 66% of those obtained by the Apple Watch 4 and 74% of those obtained by KardiaBand were accurate. Exclusion of inconclusive readings from the analysis increased the diagnostic accuracy to 93% and 94%, respectively. Presenting a clinician’s judgment to instrumental readings if the device provided no diagnosis – a hybrid approach – resulted in accuracy rates of 87% and 91%, respectively.

Diagnostic accuracy versus 12-lead AF-detection EKG

sensitive

Quality

PPV

net present value

Apple Watch 4

automated

hybrid

50%

68%

100%

93%

100%

75%

92%

90%

KardiaBand

automated

hybrid

96%

94%

93%

90%

84%

76%

99%

98%

The researchers say the agreement between the KardiaBand algorithm and 12-lead ECG for diagnosing atrial fibrillation was “excellent,” with an ak value of 0.82. Between the Apple Watch 4’s algorithm and ECG, the agreement was “fair” with a value of 0.64. Adding the clinician’s input above these readings produced k-values ​​of 0.75 and 0.78, respectively.

Overall, they conclude, “These findings suggest that although traceability of these devices is of sufficient quality, instrumental diagnosis alone is not sufficient for clinical decision making about the diagnosis and management of atrial fibrillation.”

Ford et al highlight the Apple Watch 4’s 50% sensitivity only through automated readouts, citing the risk of undiagnosed. They say the only remedy when the device doesn’t provide a clear diagnosis is manual interpretation of the traces. “This has the potential to create a significant workload for clinicians, as they may not be able to rely on the device’s automated diagnosis of sinus rhythm.”

Patients seem to agree on the need for physician intervention. A survey of study participants found that 87% said they would consider seeking medical advice if their devices consistently had abnormal readings, while most expected that a cardiologist (85%) or a general practitioner (74%) would review their smartphone results . Furthermore, they expected reactions within an average of 16 days (average of 1 day).

Shivkumar also drew attention to the Apple Watch 4’s lack of sensitivity in this study, although he said this was less important in a screening tool than a device designed to monitor patients with known disease. “Some of these types of wearables will get to the point where you can actually use them for medical indications, where you get really nervous to monitor the burden of arrhythmias and so on, where you’re going to make treatment decisions,” he said. .

As of now, “What these technologies do is bring a problem to medical attention. But once it gets attention . . . the way you act on this information will likely require an upper limit.”

Shivkumar says he often hears from patients who have questions about the results of their smartwatches. “All the time people come up with this, and #1, #2, [and] #3 Respond when we see [them] Come it: Do you really have an arrhythmia? . . . This is where the privacy figure gets interesting. In general, something was causing it,” he explained, but it often turns out that something is a normal sinus rhythm.

“Ultimately, we as clinicians will act on the information, and what matters to us are the real positives,” said Shivkumar.

Software and hardware development

AliveCor stopped selling the KardiaBand watch accessory in 2019, several months after the Apple Watch 4’s ECG feature entered the market, but the company continues to sell several KardiaMobile devices that interact with smartphone apps. In April 2021, AliveCor also”File a complaint with the US International Trade Commission (ITC), alleging Apple infringement of three AliveCor patents,” notes a company press release. The ITC investigation began in May 2021. Late last month, a federal judge refuse to kick out An antitrust lawsuit alleges that Apple tried to monopolize the market for heart rate analysis.

Looking to the future, Teh predicted that with further advances in hardware and software, the diagnostic accuracy of wearable devices would improve. “However, the health care professional’s ability to review information will still be important when basic clinical decisions such as the diagnosis of atrial fibrillation are needed, which could have significant implications for treatment. Currently, the devices cannot be relied upon to routinely replace standard diagnostic tools such as the 12-device ECG and the 24-hour Holter monitor.”

Teh said he respects the investment involved in developing advanced heart rate assessment techniques, and as such appreciates why the details are proprietary. However, he added, “I think our study provides some objective data, which companies may use to improve their algorithms.”

The paper notes that if software could be updated without consumers having to purchase new hardware for each release, the algorithms would likely be improved more regularly. “However, this in turn makes validating these devices in the real world very difficult, as each iteration of the program will require its own validation study, which is expensive and time-consuming.”

The investigators wrote that with the latest generation of smartwatches also arriving on an annual basis, it can be difficult to keep up. All this leaves the doctors With so much uncertainty surrounding these devices” Accuracy, despite the growing expectations of our patients to adapt their management to the results.”

Leave a Comment

Your email address will not be published.