Photo 2. Kardia Mobile device records a single-lead ECG of the
author’s heart rhythm.
November 2018 Dallas Medical Journal 21
Using the cyclic variation of the reflected light plus the motion
information detected by the device’s accelerometer, the device
can calculate the HR via an algorithm. In the hospital setting,
pulse oximeters use the same principle to determine HR and
pulse ox readings. Most fitness trackers sit on the wrist, but
newer ones are attached to the temple and earbuds, so to avoid
excessive arm motion, which can cause artifacts and erroneous
HR calculations.
ACCURACY OF WRIST-BASED VS CHEST STRAP
The accuracy of the methods and devices that measure HR has
been questioned. The HR accuracy of chest strap electrodebased
monitors under various degrees of rest and exertion
have compared favorably with medical Holter monitors as the
gold standard. I have found — as have many of my weekend
warrior colleagues — that the PPG, optically based HR monitors
(Fitbit, Apple smartwatch) may be accurate at rest, but at higher
HR, especially with higher levels of exertion and excessive
arm motion, the device has much poorer correlation than with
electrode-based chest strap monitors (e.g., Polar).
Several factors can cause a wrist-based optical band to be
inaccurate with exercise:
• Fast-paced arm movements can cause the monitor to move on
the wrist, leading to errors in capturing reflected light and thus
errors in determining the HR
• A band that is too tight can limit blood flow, impairing the
ability of the sensor picking up reflected light from blood.
Alternatively, if it is too loose, space and air can prevent the
optical monitor from detecting blood flow. A band too high or
too low on the wrist can provide inaccurate information.
In 2017, JAMA published a research letter detailing a study
from the Cleveland Clinic comparing the accuracy of five
common HR monitors. Four of the most popular wrist-based
optical HR monitors were compared for accuracy with the Polar
electrode chest strap as subjects wore standard limb lead ECGs
on a treadmill stress test. The Polar H7 chest strap performed
the best, with a concordance correlation coefficient of 0.99. The
four wrist-based devices ranged from 0.83 to 0.91. The best
performing wrist device had an HR error range of –27 bpm to
+29 bpm. The wrist-based devices were most accurate when the
wearer was at rest, and accuracy diminished with exercise. For
patients with heart conditions who are rehabbing with HR-based
exercise prescriptions, the accurate assessment of HR is even
more imperative. In this setting, electrode-based chest strap HR
determinations are essential.
DISEASE SENSORS
The biometric use of wrist-based wearables for cardiac
monitoring purposes thus seemed to be doomed to semi-accurate
HR determinations at low to moderate physical activity for
low- to moderate-level fitness enthusiasts. That seems to be
changing with the emergence of the Apple watch Series 4, which
is embedded with electrode technology from a company named
Alivcor to record electrical rhythm signature of the heart in the
form of a single-lead ECG. Until several years ago, electrodebased
technology for ambulatory cardiac monitoring mostly was
limited to HR determinations in a chest strap. That changed in
2011 when Alivcor was founded based on its 1998 patent for
wireless transmission of ECGs in hand-held devices. In 2012
it developed an FDA-approved smartphone case that works
as a single-lead ECG. Embedded on the back cover were two
electrodes upon which the user would place two fingers from
each hand for about 30 seconds. The electrical signal would be
transmitted via Bluetooth to the app, where the tracing would
be filtered and recorded, and voila, you have a medical-grade
single-lead ECG. The case was coupled with a proprietary app
called Kardia. The company later released a credit-card sized
device version called the Kardia Mobile. I keep mine in my
wallet. (See photo 2.)
I purchased the case and the card when they first came out.
Initially, I used them more as a type of parlor trick to record
ECG tracings of friends and family members. But on at least
two occasions, I recorded tracings of friends of my parents
and identified unrecognized atrial fibrillation. For my patients
with paroxysmal atrial fibrillation who need ongoing disease
management, I recommend they purchase the Kardia Mobile
card.
INTERPRETATION OF RESULTS
In the initial study of 54 healthy subjects in 2012, the device
diagnosed WPW, Vtach and SVT. It especially was useful in
identifying and managing atrial fibrillation. These early versions
required a physician to interpret the tracing, which limited its
adaptability. Most people can appreciate the significance of
an excessively slow or fast HR, but with a rhythm strip, most
people cannot translate the images into meaningful information;
a physician still was needed for interpretation. Initially, and for
a small fee, the Alivcor app would send the tracing to the cloud,
where a staff physician would read it within a few minutes.
While its first ECG devices relied on physicians to analyze the
readings, in 2015, the company received multiple FDA