Hacker News Reader

Comment by ihnorton | Hacker News Reader

Interesting move, but: this, folks, is why hospitals buy from $BIGCORP (Epic, GE, Siemens).

edit: to clarify, by "this" I mean "stability". One path to build a medical software company has been to start by targeting individual practices, clinics, and small hospitals. That is still possible, and there are plenty of companies doing it. But that path is getting harder to take in the US because of medical industry consolidation. Even "independent" practices now usually have referring affiliations, and more often than not this means that they use the hospital network's existing software infrastructure.

View on HN

Is it?

I have no solid idea, but my impression from glancing at solutions for healthcare years ago wasn't that Hospitals are choosing Big Co so much as they have little choice.

As I understand it, smaller companies don't try / give up on keeping up wit the costs of required certifications (possibly the wrong term) and the like.

This also seemed to help explain the slow pace of advances and updates to such systems as patches and changes have their own onerous process.

Yes. Three reasons:

1. Healthcare is heavily regulated. HIPAA is the most well-known law, but there are also countless local and state laws hospital systems must comply with.

2. Any new software needs to be sufficiently integrated with several existing systems and databases. What this means is that you can't just develop a stand-alone solution that deals with just one part of the process.

3. Many hospital staff are technophobic. Doctors especially.

I'm oversimplifying of course, but this is basically why incumbents (big co) are preferred. They have tons of domain expertise in healthcare, are deeply familiar with processes and requirements AND know plenty of people in the right places to grease the wheels when necessary. This makes it very difficult for startups to break into the industry.

To OP's original point, healthcare orgs tend to think in terms of 5 year chunks, at minimum, when planning. They figure that the likelihood of any small podunk startup being around in 5 years is minimal, so even if the product is 10x or better than the existing solution, there's that additional barrier to adoption.

Not really. It happens to be that most 'image viewing' tech like this is bundled with the hardware hospitals buy. Example: If you buy a GE/Fuji/etc PACS your usually stuck with their image viewer.

MedXT is not a viewer. It's an exchange product with a viewer. Large COs like GE (Siemens), Fuji, McKesson, are not even close to having the exchange capabilities like MedXT. It's a super strategic move for Box to buy them, Radiology is slowly moving into collaboration, and Box is now the largest vendor of collaborative Radiology.

Thanks for the correction :) Interesting.

Question: How does Box end up being the largest vendor of collaborative Radiology tech now? Do they already have assets that do similar?

collaboration in radiology requires at least a viewer, up to this point they partnered with companies like DICOM Grid to provide that functionality. The other thing you'd need is to reach into the infrastructure, behind the firewall and get the images off the legacy PACS (Picture Archive and Communication Systems). This is done via the DICOM protocol. Box didn't have this part, instead, they asked doctors to UPLOAD images to Box to view them. In the above press release, Levie specifically talks up legacy integration. This is what he's talking about. Being able to reach in and get the images, most of the time automatically, view them in the web, add to that already existing collab features Box has been building (documents, notes, etc), and you've got what could become the next-gen radiology platform - at least a more effective medical image exchange.

Disclosure, I was chief architect at DICOM Grid and built similar functionality there.

FWIW I'm not sure I agree that it's getting harder (though sure, consolidation is a challenge). That said, for the moment let's concede that it is and I'll argue that's fine.

Harder isn't a bad thing. For the last decade in healthcare we've had a small array of large players with strong footholds and really shitty projects, and a huge array of small players with small footholds and really shitty products.

Harder means you have to do a very good job to win, which will probably lower the noise floor among the $NOTBIGCORP crowd and may make that road for growth even more plausible for folks doing something really excellent.

I don't quite get what you're saying. If you had made the "risky" move here and used MedXT your solution will continue to work and will improve. You're still better off than with the clumsy desktop tools that you were using before.

Big companies do mergers all the time (including GE and Siemens selling off and buying business units). Change will always happen.

Reply on news.ycombinator.com