< 1 min reading time

It was a real treat to have John Mattison, Chief Medical Information Officer at Kaiser Permanente, address the 10x Medical Device Conference audience in May 2014.

He answered the question on everyone’s mind: What do I have to do to get my products into Kaiser Permanente?

Click here to download the entire transcript.
John Mattison: So what do we want from device manufacturers? A lot, and these are really problematic. So this is not a wish list; these are things that are driving us nuts today.

So the old givens are that of course we want value, we want total cost of ownership to be low, we want them to be reliable, we want to have volume pricing for buying 100,000 units like we do with diabetic sensors. We want volume pricing.

We want durability so that they’re not breaking constantly and we want them to be multi-user in an intelligent way. So there’s a lot of sensors out there that act as if any data that comes from that belongs to a single individual. So we need to have very seamless transition that a single sensor can be multi-user.

So, the new givens.

[caption id=”attachment_4581″ align=”alignright” width=”250″] Meet John at Exponential Medicine[/caption]Security. So our chief information security officer and I, before he left the company for greener pastures, were going to actually come to one of these shows—this was about four or five years ago—and demonstrate how easily we could hack into any sensing device because it is very easy to do so just as a demonstration. So we need much more security.

We need transparency into the open-source components that are embedded and the open-source strategy of individual vendors because the long-term strategy of how to build proprietary IP and value surrounding an open-source core is a really critical strategic decision that every company has to make, and I’ve interviewed a lot of companies about what their strategy is. Some actually have a strategy, but not having a strategy here is dangerous.

The self-awareness of location and failure of a sensor is critical. We need the sensor itself to say, “Hey, I think I’m having a problem. I’m not sure, but take a look at me.”

Self-reporting and self-healing, and a full Internet of Things (IOT), integration compliance. So, we want to have sensor networks that do some of that aggregation very locally so that we can begin to detect problems closer to the source rather than have to take it all the way into a cloud or an aggregated cloud and then discover, “Oops, 20 minutes, an hour, two hours ago, something bad happened.”

So how do you get that kind of IOT integration? Locally it has to be very compliant, and there’s a huge IOT open-source movement. I encourage you to pay attention to it. I attended a three-day off site on that very issue just about a month ago.

And then the future givens, broader-use case coverage. So any of you know Rachel Kalmar? Anybody? I know Daniel does, but she’s the person who wears about 30 sensors and works for one of the sensor companies.

What we can’t be expecting people to do is to wear multiple sensors for every possible measurable event, but to have more integration of those functions into fewer devices.

We need context sensitivity that allows better error detection and correction within the field. We need temporal sensitivity so that, for example, we don’t need to have streaming real-time blood sugars for stable chronic diabetics where we’re just helping them learn how to eat and exercise better and adjust their meds, but we sure as heck need real-time feeds for the new-onset brittle diabetics.

Earlier disclosure of problems. So if you see a problem with a sensor that you’re working with, we want to know about it. If it’s in production, we want to know about it right away and we can at least be aware of that. We don’t want to be told that we’ve got a fix for a problem we forgot to tell you about.

We want some partnerships and test beds, and for the pre-510(k) world, and I think the FDA is interested in radically transforming how they regulate and my suggestion to them is to get out of the regulation business and get into the certification business and use global crowdsourcing of certified people to do parts of the certification.

A clear strategy for data direct to consumer as well as direct to professional. And this is a big one, no device manufacturer repository of de-identified data. So there’s this notion that device manufacturers have the right to retain all their data so that they can learn from it and that they can de-identify it. And if you read anything written by Daniel Soloff on re-identification through aggregation, it’s really quite simple, and de-identification is largely an illusion.

And then, finally, earlier disclosure of product roadmaps in volatile fields. So where are you going? So Apple is the most famous for never saying one word until Steve Jobs got on stage and announced what the new thing was, but in a field where we’re affecting people’s health we need to have a better idea of that roadmap.

Dave Sheppard: Hi John, Dave Sheppard with Ontogenesis. I work with early stage companies. So one question is, is Kaiser the kind of organization that will take on early clinical studies or you wait until something has comparative evidence two or three years down the road?

John Mattison: So Kaiser Permanente is a large organization with many different legal advisors and their risk tolerance varies from location to location and from attorney to attorney.

We know that device manufacturers need test beds in order to be able to validate things, and so what we’re interested in doing is participating earlier in the process in general. So that is something that we would like to become more involved in. And I’m an advisor to the Tricorder XPRIZE, and so we’re looking at some of those opportunities. And the FDA is actually doing some early-use waivers, for example, in the space of detecting the H7N9 virus or the MERS virus and so forth. So we’re starting to get into areas where the FDA is willing to condone explicitly pre-510(k) kinds of testing. So, they like to move in that direction.

Arundhati Parmar: You gave a very high-level talk and I’m kind of bringing you down to earth a little bit, so I apologize for that. Can you talk a little bit about the relationship between device manufacturers and integrated health systems such as yourself?

The CEO of Medtronic has talked about “we don’t want to be a medical device company anymore; we want to be a medical technology solutions partner to the hospital.” Given the sort of challenges that we face in healthcare, with the Affordable Care Act and other macroeconomics situations, what kind of help do you really want? What kind of relationship would you want with a device manufacturer going forward instead of them just selling you piecemeal a pacemaker or this or that?

John Mattison: Sure, great question. And so it’s related to the earlier question. Our accountability to providing the best and the safest proven approach to every one of our members is really high. The business model for device manufacturers is to push the envelope. So how do you safely push the envelope?

And obviously everything has to go through an IRB review, which is very substantial, and so the relationship has to be an alignment of values and an alignment of objectives. And since you did mention Medtronic, the sort of infamous reluctance to release pacemaker data directly to the consumer is not necessarily in alignment with our value of transparency of data.

So I think having discussion around the intersection of values and objectives and the neutral goals of advancing the highest possible quality of care safely is the basis of a long-term trusting relationship.

Vaughn Cook: Vaughn Cook with ZYTO. If we wanted to collaborate with Kaiser, where’s the portal to find that way in?

John Mattison: So that’s the hardest question to answer because we have seven regions, all of which want to be the first to do something, and because of that we have this sort of operational implicit strategy of “leave no vendor behind,” and so we end up having a pilot with one of everything and it’s very disruptive to our finding a path towards highly scalable and affordable solutions.

So we’re actually contracting a bit in terms of the number of simultaneous pilots that we do. So we’ve been doing pilots with lots of different vendors in lots of different areas and we’re in the process of pulling in a little bit so those we do work with we have a better chance of dedicating the resources necessary to make sure that it’s stable, reliable, scalable, and operationalizable.

And that last one is a really big one, because you are going to have a great technology, and if you don’t work out the workflow carefully, it can really fail. And a classic example is tele-X, telederm, telepsych, tele whatever.

If you don’t pay very, very close attention to the workflow and how you operationalize it, you can waste a lot of time and money and focus very quickly. So what we’re trying to do is converge on fewer pilots with the intent of bringing more of them to full scale and operationalizability.

And then, finally, the last slide. Hubert Humphrey is quoted…and this hangs on the wall very prominently at Health and Human Services in Washington, and

“It’s been said that the moral test of government is how the government treats those who are in the dawn of life – the children, those who are in the twilight of life – the elderly, and those who are in the shadows of life – the sick, the needy, and the handicapped.”

And I just want to say that it’s incumbent upon all of us that are fortunate to be where we are today and doing what we do that when we introduce new technologies, we need to think about not only how big is the market and how do we get it out there quickly and how do we get the high margins, but how do we account for the rise of the digital divide and social inequities that are associated with that and pay as much attention as we can to manage that.