An Exclusive Interview with Peter van der Grinten, GM USA/Canada of dbMotion

I'd been trying for months to work out the logistics to interview dbMotion's CEO Yuval Ofek. The time difference between the company's headquarters in Israel and the US made that difficult. After the recent UPMC announcement, they suggested I chat with Peter van der Grinten, General Manager of US and Canada for the company. I had met Pete during his short stint at QuadraMed a few years back, so that was OK with me.

Tell me about dbMotion.

Our product is an interoperability solution for both IDNs and RHIOs who are looking to integrate disparate clinical systems and provide the clinician a complete clinical view of the patient, regardless of where they were seen, who saw them, and what system was used. That’s the way the system is designed – to bring about the complete integration of the clinical information.

We deliver what we call the Virtual Patient Object or VPO, which is all the clinical information on the network that’s available about a patient. We deliver it in an SOA architecture to some entity or presentation layer. We have a clinical viewer that will display it if the user has no other system, but we hope that vendors will also be able to consume pieces of it.

Vendors or systems can also make calls to the network for specific things using the SOA architecture. If they want to find out if a particular lab test has been ordered within the last 48 hours, they can make a call to our network. We would poll other systems and bring that information back for the end users.

The product was developed in Israel in an HMO-like organization called Clalit. It has 14 hospitals and 1320 clinics, labs, and pharmacies. When it started, it covered four million people. Clalit had to decide if they wanted to convert all hospitals to Vendor X systems to get interoperability, which they couldn't afford, or to look at an interoperability solution.

It's since grown in Israel to handle almost five million patients and other facilities have joined the network.

During the recent conflict in Israel, the northern part of the country was evacuated. Some of the hospitals that use dbMotion are located there. It was a Katrina-like situation for Haifa. The dbMotion system was able to provide patient information to clinicians in Tel Aviv and other cities. It was an effective interoperability solution on a very large scale.

How did you end up working for an Israel-based company?

I’ve been with the company for two years. My background is pretty much HL7. I started at HBO a long time ago and was there for 10 years. I left to join Simborg Systems, which was the predecessor to HL7. I’ve been doing that type of work for many years. When they called me about the next step beyond HL7, I first thought, "They can’t do it." I went over there to see it and couldn’t believe it. A 4-5 second response time for all these systems was amazing.

The US is the biggest market and RHIOs were just getting started. We also looked at IDNs and saw that the push to do everything on single vendor systems wasn't working and was slowing down. It was the ideal time to look at the US market. The recent success at UPMC is the beginning.

UPMC was your first US customer?

UPMC is the first traditional US customer. EHIT, part of Montefiore Medical Center, was our first US customer, although it's more of a distribution contract. They’re working with some of the New York RHIOs. We hope to announce one of the RHIOs in the New York City area in the next 60 days. 

Describe the UPMC project. It sounds like they're creating a RHIO just to allow UPMC entities to share information.

There are some advantages to being your own RHIO. Security is easier, for instance. That’s in essence the solution we’re providing. They’ll be able to avoid a gigantic revolution of converting all systems to the same vendor, instead evolving to interoperability.

For example, you need to standardize medical terminology. You might be able to do it in a 200-bed hospital, but a 400-bed one starts to get harder. Now let's say you have different types of hospitals, like UPMC. Do you try to get everybody to agree and do everything at once as a massive single-vendor implementation? They don’t believe that’s the way to go. Interoperability lets you evolve to standard terminology and decision support criteria that are universal.

UPMC may buy a hospital, for example. They can proceed in a logical manner. They're still rolling out Cerner inpatient and Epic outpatient, requiring extensive conversion. dbMotion allows them to do it in a logical and organized fashion.

We’re the interoperability solution. We’re not the EMR system. The idea would be that Cerner or Epic or McKesson or Meditech or any other vendor would consume this VPO in the normal part of their operation. A doctor treating a patient on a med-surg unit in a Cerner hospital will to continue to operate and live on the Cerner system. At some point, whether it’s when they’re admitted or a new doctor is consulted, they’re going to want to say, "is there any information available outside the purview of this EMR that has value to me?"

Interoperability lets Cerner make a call to the network. dbMotion answers the call, goes out to network nodes, finds the answer, and delivers back a VPO of information that would be the MRI from the other hospital. The interoperability solution is in the background. If a hospital develops criteria for “what is a duplicate order,” Cerner can ask that question to the network, which would find the duplicates, perhaps in an outpatient clinic.

What's the architecture look like?

It's very tiered, in the sense that each layer serves a particular function. The integration layer is where terminologies are mapped. In this type of world, you may not have agreement on terms like CBC. We’re mapping what Hospital A says is a "CBC" and what Hospital B says is a "Complete Blood Count" into a schema. You want to try to bring as much data to the clinician as you can. Maybe we’re mapping a flat ASCII file downloaded to the network or a PDF file. You’re trying to bring back a consistent view. That’s where the evolution takes place, to better standards that reduce the amount of mapping you have to do.

The communications layer manages all communications from nodes on the network. Who’s responding, when did they last respond, etc. The business layer answers the questions. We have a presentation layer, the dbMotion Viewer, that takes VPO and displays it if there’s no display mechanism, like the MD with no EMR. We also have security layers and a management layer.

The contract with UPMC was for a lot of money. How did it break out?

It was primarily software with some implementation help, maybe 80% software and 20% implementation. UPMC has a sophisticated message router that manages the millions of transactions taking place, all the HL7 transactions that are sent all over the network. We’ll have to integrate with that because we don’t want to direct connect to all of those systems.

What kind of development will you be doing in Pittsburgh?

We'll do some joint development work with UPMC on advance features to be mutually agreed upon. UPMC is focused on decision support features. We’ll work with them to joint develop other products that UPMC will consume that will be marketed in the US and potentially in the European marketplace. We have a lot of sales activitity in Eurpore now, but no customers. UPMC is providing mostly clinical expertise, but also technical expertise, like on the message router.

What kind of investment did UPMC make in dbMotion?

Part of the deal is that they’re investing in the company. The contract is two-phased. We will relocate people there and put an office there, but we don’t yet know if it will be a full-blown branch office.

What's your headcount and where are they located?

Just under 100 total. The majority are in Israel.

Is it too late for dbMotion to get RHIO business?

I don’t think so. Some RHIOS are finally getting going. We’re participating in five or six bids at the moment. It’s not too late yet, but as you can imagine, an Israeli company coming to the US … it’s a pretty conservative market used to buying from US companies. We have a hill to climb there.

We have a site that no one can even remotely match in volume and capabilities. We have over five million patients on the network with 4-5 second response time with physicians banging on it every day. It's been battle-tested to do what we say it will. No one else in the whole world can match that. That’s why I joined them – it really works. But, you have to convince American hospitals that the patients and treatments are similar.

I don’t think the RHIOs are that far along. We’ve lost a few to others. The most successful RHIOs are those with a large IDN partner pushing the idea. In that regard, we’re early in the market.

You could argue that the RHIO market isn’t that large compaed to IDNs in terms of the number of  IT dollars being spent. From that standpoint, I think the IDN marketplace is clearly one we’ll be going after. We're talking to other IDNS and hoping that our UPMC announcement will cause them to take notice.

Any final thoughts?

If you look at healthcare today, with rising costs, IDNs and RHIOs need to make interoperability the focus. We’re on a mission to get this interoperability where it benefits patients the most. UPMC wants to give better care to their patients. If we can deliver interoperability solutions without spending billions to convert everyone, that's the way to go. It doesn’t require the massive investment of a single vendor solution, but provides patient benefit.

Who do you admire in the industry?

That's a tough one to answer. I certainly admire people like Neal Patterson who are trying to get something done and trying do the right thing, even though I don’t always agree with them. Clinicians. The great thing about healthcare IT is there’s a lot of fantastic people trying to do the right thing. In general, people are just trying to help patients to be treated better and more effectively. I have a tremendous respect for the whole industry. We’re not building Ford Explorers here.

Do you read HIStalk?

Yes. Just about everyone in my company does.