Quick Thoughts: A Possible Kickstart for Disruptive Healthcare AI

Print Friendly, PDF & Email

SEE LAST PAGE OF THIS REPORT Paul Sagawa / Tejas Raut Dessai

FOR IMPORTANT DISCLOSURES 203.901.1633/.553.9827

psagawa@ / trdessai@ssrllc.com


January 31, 2018

Quick Thoughts: A Possible Kickstart for Disruptive Healthcare AI

• JPM, BRK A/B and AMZN’s healthcare joint venture will create an insurance entity covering about a million employees, with an intent to use tech to reduce costs and improve patient experience

• We expect major efforts to improve data quality, to introduce healthcare UI systems to help users self-triage before ER or office visits and to evaluate treatment options, to extend wellness programs focused on predictive diagnoses and preventative care, and to streamline processes from approvals, to payments, to fraud prevention.

• Clinical processes will be hard to change – patient data is still of low quality, the web of stakeholders (providers, government, pharmaceuticals, etc.) will resist, and the venture lacks scale (even with a million potential heads under coverage.)

• AMZN has an opportunity to establish leadership in consumer healthcare digital gateways and smart wellness programs that could be extensible beyond the venture platform. It could also build industry relationships that could benefit its hosting business. This is also a jumpstart to a move into pharmacy

We recently published a piece outlining the opportunities for deploying AI technology to address healthcare problems (http://www.ssrllc.com/publication/healthcare-ai-taking-a-long-road-one-pain-point-at-a-time/) Our conclusion was a bit unenthusiastic, given substantial obstacles for broad disruption given poor industry data quality, a byzantine and fragmented web of industry stakeholders distrustful of each other and outsiders, and the general shortage of both AI and medical talent. Still, we saw the areas of biggest potential as 1. Medical imaging, 2. Wellness, and 3. Drug development. Yesterday’s shocking announcement that Amazon would team with JP Morgan Chase, and Berkshire Hathaway to form a new company to administer healthcare for their 1 million combined employees presents a fourth category of opportunity.

Our colleague, Dr. Richard Evans, goes through obstacles that this new venture will face (http://www.ssrllc.com/publication/amzn-brk-a-b-and-jpm-look-set-to-fire-or-at-least-seriously-trim-back-the-roles-of-their-healthcare-agents/). Of these the most critical is negotiating agreements with local care providers – in many geographies the organization will be too small to independently demand parity pricing for many services. Dr. Evans also points out that established HMOs have access to broader longitudinal patient data than will the AMZN/BRK/JPM initiative. This is true, but we note that AMZN adds considerable AI and processing firepower that could help to make up for the relative narrowness of its patient records. In other areas, the business, the NewCo could likely hold its own with the traditional HMO operators. Dr. Evans has been enthusiastic for the potential for Amazon in pharmacy, and this initiative strengthens his conviction.

In this context, we are interested in the ways that Amazon might establish new business beachheads that might generate opportunity beyond the not-for-profit venture. Certainly, the new systems built for the fledgling in-house HMO will run on AWS, and the process of doing business with a range of care providers, pharmaceutical houses, etc. could help the IaaS host gain greater access to and credibility with new customers. We also see an opportunity for Amazon to build a new user interface for healthcare – one that gives consumers help in determining their best course of action when confronted with a health issue. Self-triage, supported by human doctors with help from AI systems, could greatly cut down on unnecessary urgent care, ER and office visits, while giving appropriate treatment options to the patient more quickly and with much better information. Moreover, a healthcare UI could support wellness and treatment compliance programs. Once shown to be effective, this UI could be extended to other payors, potentially offering revenues and new data for Amazon.

We agree with Dr. Evans on the pharmacy opportunity for Amazon. Existing mail order pharmacy operators are inefficient, and Amazon can certainly do better – same day deliveries, 30-day (not 90-day) fills, convenient packaging options, superior customer interface, etc. Undoubtedly, Amazon’s ambitions extend far beyond the joint venture. A million lives under coverage is a good starting point and the scale would give Amazon a better position to negotiate with suppliers and with payors as it looks to disrupt traditional pharmacies.

There are also some basic blocking and tackling issues where Amazon would add considerable value to the proposed HMO. Managing claims and payments could be made easier far more transparent to users and providers, and fraud risks could be minimized with AI powered analytics. Data security could be better assured in the cloud with AI and potentially blockchain-based defenses.

If the venture is successful, we can imagine new employers enthusiastically signing on, adding new hosting business for AWS, more data for the venture to analyze, and more power in negotiations with care providers. We didn’t think disruption was possible. Maybe now, it is.

Print Friendly, PDF & Email