Measuring heart rate at home using connected sensors
Heart rate is a measurement that researchers and clinicians increasingly want to collect remotely, especially as COVID-19 disrupts healthcare systems. Did you know there is a choice of two types of sensors to deploy? The Elektra Research Team explains the tradeoffs in accuracy and usability between electrocardiograms (EKG/ECG) and photoplethysmograms (PPG) to guide decision-making.
Written by: Christine Manta, Hannah Curren, and Manny Fanarjian
Remote heart rate (HR) monitoring is gaining popularity in clinical care and research. Between 2008 and 2018, there was a 580 percent increase in published articles containing the words “wearable” and “heart rate”. Before connected technologies like wearables are deployed, clinical and research teams should understand which type of sensor is best for their needs. Improper selection can lead to poor data quality, poor patient adherence and wasted time and resources.
In this post, we highlight the value of capturing HR data and discuss tradeoffs in accuracy and usability of two common sensors: electrocardiogram (EKG/ECG) and photoplethysmogram (PPG).
Why collect heart rate data?
Heart rate (HR) is a simple and effective parameter for monitoring patients outside of the hospital. It can be used for many purposes, including:
- Detecting arrhythmias occurring naturally or as a medication side effect
- Adjusting medication doses in chronic conditions based on patients’ resting HR (e.g., beta blockers in congestive heart failure, CHF)
- Approximating stress response
- As a diagnostic marker in many medical conditions, from CHF to anorexia nervosa
Remote HR monitoring is valuable to:
- Researchers because it provides much more data to extract low-level signals, allowing for accelerated decisions about study design and feasibility.
- Clinicians because it means more reliable diagnostic information, reflective of a patient’s day-to-day experience.
Want to learn more about remote vs in clinic data collection? Check out this primer on digital measurement.
Heart Rate Variability (HRV), a common parameter derived from HR, is a measure of the time variation between heart beats. It is affected by age, respiratory rate, stress, cardiovascular health, and autonomic nervous system activation.
Measuring HRV can aid prognosis: low HRV is associated with poorer cardiopulmonary fitness and negative outcomes in cirrhosis, heart failure and sepsis. It can also be used to predict conditions like diabetic neuropathy before patients begin experiencing symptoms.
Remote HR monitoring is gaining attention during the COVID-19 pandemic. Two recent FDA guidances recommend remote monitoring for ongoing clinical trials and clinical care. Research is under way to determine how HR and HRV can monitor or predict COVID-19 (e.g., Oura Ring projects). While there are anecdotal reports, more evidence is needed.
How can you measure heart rate?
The two main approaches are electrocardiogram (EKG/ECG) or photoplethysmogram (PPG).
An EKG, which uses an electrode to measure the heart’s electrical activity, is the gold standard in traditional clinical settings, and is used in many connected technologies. Most EKG products have a single-lead configuration. For more detailed arrhythmia analysis, some newer wearables incorporate multi-lead designs.
- Chest straps: Polar and Garmin are popular among consumer fitness enthusiasts.
- Portable monitors: AliveCor’s KardiaMobile involves a person placing fingers from each hand on portable electrodes to capture a single-lead tracing on their phone. The KardiaMobile 6L uses an additional electrode on the left leg for a six-lead tracing.
- Patches: MC10 and CamNTech have been used in research settings.
PPGs, sometimes called optical heart rate sensors, dominate the wearable space, and are present in almost all fitness trackers. They use LED lights to measure blood volume pulse (BVP), which is the change in blood volume determined by the amount of light passing through the skin. The light sensor determines heart rate by calculating the amount of time between the changes in blood volume. PPGs are used by researchers and consumers, including:
- Bands for wrist-worn products like Fitbit or arm-worn products like Everion
- Clips worn on the chest such as Health Tag
- Rings like Oura Ring
As shown in the figure below, the two sensors measure different parameters, resulting in unique waveforms. This is an important distinction for researchers looking to perform analysis on signal data to calculate HRV or diagnose arrhythmias.
Which one’s better?
There are two key considerations:
- Accuracy
- Usability for clinical or research teams and for patients or study participants
Accuracy is impacted by skin tone, activity and motion artifacts. Usability is impacted by data volume from continuous versus episodic monitoring, form factor and wear location.
There is mixed evidence supporting the accuracy of PPG on darker skin. Green lights from PPG sensors are better absorbed by melanin, blocking the green light’s path, suggesting that individuals with darker skin may have less accurate readings. However, multiple studies have found no statistical difference in measurements for individuals with darker skin. Since EKGs do not rely on light, variations in skin tone are a non-issue, eliminating a potential form of bias.
Both PPG and EKG accuracy is complicated by motion artifacts. Both large artifacts such as body movement during high intensity exercise and small artifacts like typing on a computer cause less accurate measurements. Most consumer products have algorithms to automatically identify and delete PPG measurements that are coupled with large amounts of motion data. Generally, heart rate measurements taken at rest and during low-intensity activities are more likely to be of higher quality.
Usability for clinical and research teams is impacted by the volume of data collected in continuous vs episodic monitoring. PPGs are useful for continuous monitoring to assess trends, such as early identification of deterioration in heart failure patients. EKGs are useful for episodic measurement such as screening for atrial fibrillation. Regardless of sensor or monitoring time, the ability to access direct electrical or blood volume pulse signals will vary. Especially for PPG products marketed to consumers, additional analysis on signal data may not be possible.
Usability for patients and study participants is impacted by form factor and wear location. PPGs are often unobtrusive wristbands. One in five Americans wear a smartwatch, so individuals who already own PPG products may be less likely to switch tools for a study. Bring Your Own Device (BYOD) clinical trials can increase adherence, but pose technical challenges. EKG patches or chest straps may not be comfortable long term, especially for seniors with frail skin. If continuous monitoring is not required portable EKG monitors are user-friendly.
So, which sensor should you use?
The answer is: “it depends.” Research and clinical teams must weigh the tradeoffs in accuracy and usability to make the best decision for their needs.
Curious to learn more about HR measurements? Check out a recent journal club from the Digital Medicine Society, discussing inaccuracies in PPG sensors
Did we miss an important tradeoff between the two sensors? Let us know in the comments!
Acknowledgements: For reviewing, editing and providing illustrative examples we gratefully acknowledge Adam Conner-Simons, Chuck Schlaff, Dena Mendelsohn, Andy Coravos, Jeesoo Sohn and Tristan Rasmussen.
Elektra Labs is advancing healthcare by enabling the safe, effective, and personalized use of connected products in real-life settings.
Technology has always moved faster than our ability to safeguard ourselves from its consequences. Connected health products are no exception. The increase in FDA clearances for these products has accelerated their adoption in clinical trials and routine care, but not without major concerns regarding data security, accuracy and biases.
Similar to pharmacies, which safely and effectively evaluate and dispense drug components, Elektra Labs offers free and paid services to better evaluate and dispense the >750 connected products on the market today for clinical trials and remote health monitoring.
Since Elektra’s inception in 2017, the company’s mission, values and progress have garnered attention in several top publications, and funding from several prominent health, bioscience and tech investors as well as the National Science Foundation and the Harvard Business School Rock Center for Entrepreneurship.
Sign up for an account at ElektraLabs.com to get access to our Academy and learn more about digital biomarkers, decentralized clinical trials, novel endpoints and deploying safe and effective connected technologies in research and clinical care.
If you’re an engineer, security researcher, or research collaborator with a passion for digital medicine we’re hiring! We’re looking to grow the team, and we need your help in building the infrastructure for safe, effective and ethical tools in the digital era of medicine
