A change of heart: InfoBionic’s cheaper, better way to analyze heart arrhythmias

InfoBionic is using cheaper wireless technology and cloud computing to revolutionize the diagnosis of heart arrhythmias

InfoBionic shows how advances in cloud computing and wireless devices enable more seamless, efficient diagnosis and monitoring of heart arrhythmias, which until now have been costly and labor intensive diseases to diagnose. By simplifying the diagnosis of heart arrhythmias with just one device, InfoBionic is creating an opportunity for itself to use machine learning on huge, novel datasets to help doctors and patients better manage this challenging condition.

Heart arrhythmias are episodic, irregular heartbeats that occur in several million Americans,[1] leaving them at greater risk for stroke and heart disease and contributing to 130,000 deaths each year in the United States.[2] Unfortunately, heart arrhythmias are very difficult to diagnose as they occur unpredictability with not yet fully understood triggers that vary across patients.[3] Currently, diagnosis works in three stages, where patients whose condition is not identified in an earlier stage moving on to a costlier, more involved method that culminates in mobile cardiac telemetry where real-time patient data is sent to humans who flag unusual heart activity to doctors. The current system has the disadvantages that it requires numerous patient visits to try and trade-in the different diagnostic devices, requires doctors to have three types of devices on hand and be expert at analyzing their different kinds of reports, and costs a lot to the healthcare system (up to $1,000 for the most involved test).[4]

InfoBionic is a start-up taking advantages of two powerful aspects of ongoing digital transformations – the rapid decrease in the cost and size of wireless devices [5] and the computational power available through cloud computing [6] – to create a breakthrough product called MoMe Kardia that helps doctors better diagnose arrhythmias and capture more of the value of their services.

MoMe Kardia’s wireless technology provides several benefits to its doctor customers and patient end users. The small device that patients can wear on their belts captures all the different data that the three existing arrhythmia diagnostic tools use, saving doctors from having to have different devices on hand.

MoMe Kardia’s device:

MoMe Kardia’s device: Source: InfoBionics, “The System.” https://infobionic.com/the-system/, accessed November 2016.
Source: InfoBionics, “The System.” https://infobionic.com/the-system/, accessed November 2016.

Based on advances in wireless technology, the MoMe Kardia streams the data it collects over cellular networks, eliminating patients’ need to carry additional wireless transmission devices or return to the doctor’s office to upload their data.[7] Additionally, doctors can wirelessly change the data collection settings of the device to switch between the three different diagnostic tools currently in use, again saving patients’ time from having to visit the doctor’s office to swap out different devices.[8]

MoMe Kardia’s use of cloud computing power creates easy-to-use data-analysis for doctors, eliminating lots of previously manual labor and helping doctor’s capture more of the value of the care they deliver. By using the cheap, abundant processing power available through cloud computing, InfoBionic can apply automated algorithms to analyze the data the MoMe Kardia collects.[9] This allows MoMe Kardia to integrate the different kinds of data collected in the different diagnostic modes in one integrated app, eliminating doctor’s current need to make sense of three different reports for each diagnostic method. Additionally, the cloud-enabled algorithms are powerful enough to eliminate the current need for expensive ($1,000 a month), manual review of heartbeat data in mobile cardiac telemetry.[10] As a result, doctors can now get reimbursed for the “technical” part of mobile cardiac telemetry, capturing more of the value of the service. Currently, the vast majority of the insurance reimbursement for the $1,000/month diagnostic treatment goes to the humans scanning the heartbreak data as it comes in, with only $30 shared with the doctor.[11]

MoMe Kardia charges doctor’s a monthly fee of around $500 for the device and data analysis. InfoBionic believes doctor’s will be able to diagnose 1.5 to 2 patients a month with the device, replacing the three different devices doctor’s currently use which have dramatically different costs (from $100 to $1,000) [12]. MoMe Kardia’s 3-in-1 device with a flat fee is appealing to doctors given the move to value-based pricing in healthcare. [13] It allows doctors to more confidently forecast the cost of diagnosing arrhythmias when entering into contracts to provide those services.

InfoBionic should aggressively apply machine learning, another digital transformation, to the enormous datasets MoMe Kardia is collecting. Machine learning could improve MoMe Kardia’s algorithms to more accurately and quickly spot arrhythmias. This would leverage MoMe Kardia’s unique position of having datasets from all three types of traditional diagnostic methods and having that data in scale on the cloud as opposed to more locally stored with individual doctor’s practices. MoMe Karida also has a great opportunity to move from the diagnosis of arrhythmias to help with their management by better forecasting arrhythmias. By applying machine learning to its unique data, MoMe Kardia could help doctors better understand what may be triggering their patients’ arrhythmias and how those triggers might vary across different patient populations.

 

797 words (excluding citations)

[1] Zoni-Berisso, M. et al. “Epidemiology of atrial fibrillation: European perspective.” Journal of Clinical Epidemiology. Volume 6, 2014, pg. 214.

[2] Stuart, Mary. “InfoBionic: Set to Disrupt Arrhythmia Monitoring.” The MedTech Strategist. Vol. 3, No. 12, pg. 47. August 24, 2016.

[3] Ibid.

[4] Ibid.

[5] Iansiti, Marco and Lakhani, Karim R. “Digital Ubiquity: How Connections, Sensors and Data Are Revolutionizing Business.” Harvard Business Review. November 2014, 4.

[6] Kim, Eugene. “This One Chart Shows the Vicious Price War Going on in Cloud Computing.” Business Insider. January 14, 2015. http://www.businessinsider.com/cloud-computing-price-war-in-one-chart-2015-1.

[7] InfoBionics, “The System.” https://infobionic.com/the-system/, accessed November 2016.

[8] Ibid.

[9] Stuart, Mary. “InfoBionic: Set to Disrupt Arrhythmia Monitoring.” The MedTech Strategist. Vol. 3, No. 12, pg. 49. August 24, 2016.

[10] Ibid.

[11] Ibid.

[12] Ibid

[13] Gerhardt, Wendy et al. “The road to value-based care: Your mileage may vary.” Deloitte University Press. March 20, 2015. https://dupress.deloitte.com/dup-us-en/industry/life-sciences/value-based-care-market-shift.html, accessed November 2016.

Previous:

The new Thanksgiving weekend activity … “Spying Saturday”

Next:

Airbnb – Evolving from Rooms to Experiences, Places and Meet Ups

6 thoughts on “A change of heart: InfoBionic’s cheaper, better way to analyze heart arrhythmias

  1. Thanks for this informative post about InfoBionic’s MoMe Kardia, RG. The ability to downsize and combine the 3 modes in this cardiac telemetry equipment is certainly an advantage for consumers/patients. 24-hour Holter machines used to be bulky devices that require multiple sticky electrodes pasted on the body, which was clearly inconvenient for patients. Although I appreciate tapping into the processing power of cloud technology and the huge databases of algorithms, it would be interesting to find out how the diagnostics abilities of these algorithms vs. a more conventional method of having a technician/physician interpret the readings. Would certain rarer arrhythmias slip through the algorithms or would the fatigue of a technician looking at 1,000 telemetric readings a day be more dangerous for patients?

  2. RG,

    This post was particularly relevant for me, as one of my uncles passed away without warning due to arrhythmia at a relatively young age. What was most frustrating was that there was not an immediate answer as to what he could have done differently to prevent his early death – while he was aware that he had arrhythmia his doctor had not expressed much concern and proper monitoring was lacking.

    For this reason, I thought your proposition that MoMe Karida “move from the diagnosis of arrhythmias to help with their management by better forecasting arrhythmias” to be particularly strong. Reflecting on our class discussion about IBM’s Watson, it seems that machine learning has the most potential when applied to questions we currently do not have answers to. As you mention, arrhythmias vary across different patient populations and there is little consensus on how to treat these effectively: “by applying machine learning to its unique data, MoMe Kardia could help doctors better understand what may be triggering their patients’ arrhythmias and how those triggers might vary across different patient populations.”

    You mention that using MoMe will result in significant cost-savings as it reduces the amount of humans scanning heart data. If I understood correctly, it would cost doctors ~$250/patient per month to use MoMe. This made me think about the Public Goods question we discussed in class and whether MoMe will eventually be seen as having a responsibility to provide cheaper services. Considering this is a life/death situation I wonder whether they will have to increase accessibility as the services would cost ~$3,000/year per patient (assuming doctors price at breakeven). Perhaps MoMe can reduce prices as they expand their consumer base in a subscription/type model focused on preventive care with reliable cash flows.

    Thanks for sharing!

  3. As someone who has a strong passion for the biomedical sciences, it is always exciting to learn about new devices that can help to diagnose/treat patients more efficiently and effectively, so thank you for sharing this innovation with us!

    My question for MoMe would be one regarding the vast amounts of data that they are collecting (wirelessly!) about the patients that are using their devices. Does this data ‘belong’ to the patient, the doctor or MoMe? Is this something that patients would be concerned about given that it is personal medical data? As you suggest, this data collection by MoMe could serve a good and higher purpose if geared towards becoming better at diagnosing patients. But there is also always the fear that MoMe could exploit this data in some way – e.g., selling this to insurance companies. Are customers comfortable with this aspect of data collection and ownership given that it pertains to personal medical information? Even if we assume that MoMe follow data protection rights stringently, can we trust that there will be no breach/hack to their systems that may release identifiable medical data to the public? Although cloud-based data serves many benefits, it also opens ourselves up to greater risk as consumers to a a breach of privacy.

  4. Thanks for a really fascinating article! I think this is a really exciting product in many ways – I like that it allows for holter, event, or mobile telemetry monitoring all in one device. I’m also very optimistic about the idea of generating an arrhythmia database (assuming patient consent) and think this is a great space in medicine where machine learning can be applied – ECG interpretation particularly in the setting of arrhythmias is a difficult skill and requires a lot of experience and nuance. I have some concerns about the pricing, which seems to be justified relative to the cost of mobile cardiac telemetry. In my experience I’ve never seen mobile cardiac telemetry used; there are very few instances in which a patient is healthy enough to avoid hospitalization but requires real time ECG interpretation. As a practice manager I’d like to see a pricing tier that doesn’t charge me for 24/7 data analysis. I’d also push back against the assertion that this device would be able to identify triggers, as in many cases arrhythmias are triggered by health conditions external to the heart, which wouldn’t be captured by ECG data.

  5. All of us consent to third parties storing and to some extent using our data. Indeed, in order to be able to use many social media or e-commerce platforms, we are required to do so. Oftentimes, the data isn’t even required for the company to deliver the service we are seeking. For this and other reasons, some of us find this required consent to be a “hard sell.” That is, we are reluctant to give up our privacy, but in the end we do because the value we get from the service is high.

    I believe InfoBionic would have an easier “sell” than social media or e-commerce companies. While some patients might be initially wary to consent to use of their medical data, they can be comforted in the fact that this data will be used to save lives. As we saw in the Watson case, more quality data leads to better results when it comes to machine learning. Thus, each person contributing their data would directly lead to a smarter device that can improve accuracy of diagnoses. Since consenting would directly lead to saving lives, I believe people will willingly contribute their data for the benefit of all.

  6. Thanks for this post. I have a family member who has suffered from arrhythmia in the past and currently has a pacemaker, so this post caught my eye.

    While I see the cost benefit of replacing three different devices, I wonder if there is data to support the ability for doctors to better diagnose this condition. I believe in the power of data and but I worry that automated algorithms may allow for doctors to relax their own personal judgement/analysis which may be needed to catch a black swan event. Perhaps MoMe could be a “cyborg” which balances machine intelligence with human judgement (Shivon Zillis from Bloomberg Beta talks more about the potential use cases for cyborgs: https://techcrunch.com/2015/11/26/machine-intelligence-in-the-real-world/).

    I also wonder how MoMe can gain traction if the hardware is dependent on integration with the existing three diagnostic tools. While I am not familiar with the existing tools, I would think that those companies may want to integrate and analyze their own data rather than have an external party come in and take that margin.

Leave a comment