This’ll be the last one today but it’s been an interesting conference. Some quick ones from the panel of DSM companies.

Carter Coberley from American Healthways likes neural nets to predict cost and therefore go for intervention early.

Frank Martin from I-Trax likes to tell people about the global cost of missing work through illness. Only 1/3 is health care cost, the rest is lost productivity and re-hiring costs. I-Trax has both the predictive bit and the onsite counselling and nursing services prosucing DSM.

Chris Selecky from Lifemasters wants to go for the intervention to the right person at the right time, using more and more sophisticated technologies–start with the high risk people, and then try to get to the people who are going to be sick, BUT they are held accountable for outcomes in only one year–so they have to be effective. Incidentally they are in-source and out-source, in that more case managers are using their system at customers than in their own centers. They are also trying to engage the patients.

I asked about the recent studies showing that it didnt save money. Chris Selecky’s response — people are buying it, and they wouldn’t if it wasn’t working (and she like Sam Nussbaum spent some time slagging off of the CBO report).

One of the biggest challenges is identifying a comparison. American Healthways did a controlled cohort with BCBS MN and showed a 14% PMPM improvement — the same as Anthem showed in the study that Sam Nussbaum showed at lunch. Plus Carter says DSM has an impact on standards of care for everyone else in an osmosis type of way.

Shorter Sam Nussbaum (CMO Anthem) "DSM does work, really. But the baseline changes (sick people are not homogenous or constant–they also get better (or die). It’s worth about 14% on a PMPM basis across populations, but has an ROI of 2.85 to 1".

Brad Holmes from Forrester breaks up docs by age (under 44 makes them much more accepting of technology). Breaks up demand for software based on what docs want. Want ePresribing and charting, get Medicalogic (GE). Want diagnosis, use Next Gen. So you buy based on what you do. No one size fits all. BUT PMS systems revenues going down, EMR applications are taking off and 2004 will cross the line of PMS revenue

Blackford Middleton (ex Medicalogic, these days at Partners Healthcare in Boston) has worked out the cost-benefit model for rolling out their "LMR". Now used by 3500 of their 7000 affiliated physicians. The biggest impact is for drug savings, by using the on-formulary drugs. For CPOE there are levels; most basic is structuring data capture, second is providing some patient data, third is decision support based on algorithms and the patient record. Partners has the most advanced, and the result is that per provider saves up to 9 ADEs per year and costs savings of up to 28K per year per provider. Most valuable ones have by far the most impact ($28K per year vs $3k). For systems capitated at 14% they think this will cost solo providers a whole lot, and only save money for big groups. BUT Of all the benefits, only 11% go to the physicians. The rest goes to everyone else in the system, primarily the payers.

Gloria Austin from Brown and Toland, a 100% capitated group in San Francisco is just getting their EMR modules up (and now have lab data online) by working with affiliated hospitals, etc , etc. It’s an interesting talk (although her charts suck for readability) but I think it’s mostly irrelevant to most physicians in the US. But they are focused on implementing the EMR by concentrating on providing the physicians with the benefits, and creating the infrastructure foundation to do that. This process started for B&T in 2000 and its taken them till now to get it up. What is it costing them? $8m in first 4 years and another $10-12m going forward. Using IDX Flowcast and Allscripts Touchcast (because they were already using them and the integration was easier) plus you can import this into the Business Intelligence tool they use (Cognos). And in a nutshell this is really, really hard work, and costly organizationally to do. But she believes that you can create a seamless group that looks like a Palo Alto clinic, or a Kaiser.

Blackford said it, in reply to a question from me….there are potential operational improvements for small and solo docs (not just savings to capitated systems) but we have a major public policy problem getting these physicians to the table. And (he didn’t say) no real answers.

In the response and Q&A to Brailer, we’re told that adoption can happen in small offices and by PCPs. But CMS in California has been handing out some incentives for smaller groups (something called the Docket program) and only 256 out of 15,000 PCPs have so far expressed an interest. As the guy from CMS region said, this is disappointing.

Now the conversation has veered off as to whether we should all use the VA system (Vista). Come on people, the reason that doctors don’t use EMRs is not because the software costs too much or hasn’t been any good. In countries where they do use it a) the government paid for it all (including the hardware) and b) they were told they had to use it!

I’m at SBC Park in the bar (really!), the Wi-Fi is free and my lap top is propped on one of those big barrels of Coke cans and somewhere in the distance David Brailer is telling us that the US will exceed the achievements of those more centralized (code for socialized) systems. He thinks electronic records will happen but he’s a little less optimistic about inter-operability. No money, no mechanism and no demand for it. I’d agree that it sounds like there’s not much market for it, although there’s a hug amount of value from doing it. He calls this the "first-mover disadvantage". Like the first guy with a fax machine….

OK, Brailer’s just done. He mentioned the RFI released on Monday that is to get at the nitty gritty of the basics behind the NHIN (national health information network).  Is this pure peer to peer? Is it regionally based.

He also talked about the Regional health exchange networks (RIOs).  There are some 25 now, some 12 Federally supported.  But he is counseling against a government solution — "this is the same government that brought you HIPPA!"

He also said that if we get the EMR at the bedside but we don’t get interoperability, then we might make the problem worse.  But then he laid out all the reasons why inter-operability wont happen. I don’t think that he’s as optimistic as he thinks he is.

Thursday (Pacific time) I’ll be at the HIT meeting in San Francisco, held at PacBell SBC Park so that we don’t have to cross a picket line at the downtown hotels. I’m informed Barry Bonds won’t be there but current health care IT MYP David Brailer will, as will the rest of the good and the great of the health care IT world. There’ll be live blogging if you’re lucky and if the stadium wide Wi-Fi is all it’s cracked up to be.

The NY Times looks at the ongoing battle between J&J’s Cordis and Boston Scientific’s drug-eluting stents. Since Boston Sci has recovered from its recall problems over the summer it has been really cranking out the numbers. It’s Taxus stent sold over $600m last quarter, comfortably besting Cordis’ Cypher and putting it on course to become the best selling medical device of all time–recall that if it gets to a $1 billion quarter, only 3 drugs sell more than that in the US. Cardiologists like Taxus because it’s easier to put in and apparently the problems with early openings are behind it. However, there are some indications that Cypher might actually have better outcomes.The main issue in stents is the restenosis rate, or in English, when you put the stent in, how long does it take for the blood vessel to become blocked again? It’s been a progression as first balloon angioplasty was used, then a bare metal stent and now a drug-eluting stent. But it appears that the restenosis rate for Cypher is better than that of Taxus.

The results of some early studies comparing the two stents were reported last month at a major conference in Washington of cardiologists who perform stent procedures. The studies included one in Korea comparing Taxus and Cypher with a control group that received a bare metal stent that was not coated with any drug. The trial focused on patients with longer blockages, or lesions, in their vessels than the average patient who receives a stent, and had a high proportion of patients with diabetes or other diseases that can lead to complications. While both drug-coated stents achieved far better results than the bare metal control group, various measures showed that Cypher patients had a third to a half as much regrowth of lesions as Taxus patients. And while the number of patients who suffered such extensive reblockage that they needed a second procedure was too small to produce significant results, those data also clearly favored Cypher.

It’s unclear what “far better results than the bare metal control group” means but I’d be very interested in the long term outcomes of this type of study. Recall that last year a Stanford group compared the long term results of those who had bare metal stents with those who had a CABG, and over a 10 year period the CABG group had much better outcomes at a much lower cost. So if the restenois rate for the DES doesn’t stay good in the longer term, they too might be a failure in health services research terms while being a huge market success. As with any technology assessment the attitude of the physicians on the ground is far more important than that of the academic wonks looking at longitudinal data sets. Meanwhile an earlier article in the NY Times questioned whether any of these surgeries–including CABGs–aimed at keeping arteries functioning were a good idea.

Researchers are also finding that plaque, and heart attack risk, can change very quickly — within a month, according to a recent study — by something as simple as intense cholesterol lowering. “The results are now snowballing,” said Dr. Peter Libby of Harvard Medical School. “The disease is more mutable than we had thought.”The changing picture of what works to prevent heart attacks, and why, emerged only after years of research that was initially met with disbelief. In 1999, Dr. Waters of the University of California got a similar reaction to his study of patients who had been referred for angioplasty, although they did not have severe symptoms like chest pain. The patients were randomly assigned to angioplasty followed by a doctor’s usual care, or to aggressive cholesterol-lowering drugs but no angioplasty. The patients whose cholesterol was aggressively lowered had fewer heart attacks and fewer hospitalizations for sudden onset of chest pain.

But really that’s what Dean Ornish has proved for years. With aggressive enough diet exercise and behavior changes, heart disease can be relatively quickly reversed. So does that spell the end for stents and CABGs and a trip to the breadline for cardiologists, cardio-thoracic surgeons and their suppliers? Well probably not. The level of change required to avoid surgery is pretty extreme. I gave a talk to Ornish’s group a while back and had the lunch they provided. I had to break out for a cookie run in the afternoon after the no-fat no-sugar experience was too much for my delicate constitution. In general Americans like this type of intervention over changing their lifestyle.

Their new stents, coated with drugs to prevent scar tissue from growing back in the immediate area, are increasingly popular among cardiologists, and sales are exploding. But there is not yet any evidence that they change the course of heart disease.”It’s really not about preventing heart attacks per se,” said Paul LaViolette, a senior vice president at Boston Scientific, a stent manufacturer. “The obvious purpose of the procedure is palliation and symptom relief. It’s a quality-of-life gain.”

Those of you who’ve been paying attention may note that there’s a mini-boomlet going on again in Silicon Valley, and those of you clever enough to have bought Google stock at its IPO, or have been sleeping with the founders (and you know who you are!) are at the center of it. But there has also been a ton of venture capital pouring into social networking sites like Linked-in, Ryze, Tribe and Friendster.

These sites allow you to contact people you don’t know via a six degrees of Kevin Bacon approach. VC and rebel doc Chris Mayaud has become the most connected player in health care on Linked-in by matching his address book to the world. If he’s bugs you I apologize as I was the one who got him started on there! And even though I only have 1% of his “connections” I have had some work and interest come in from it.

There are also healthcare sites based on list-servs, or “groups” that act as bulletin boards, often for patient groups. But they tend to be long lists of email postings, and haven’t really changed much since the early 1990s, even though they are extensively used. And then again there are personalized medical record sites that funnel information to users based on their specific conditions. (That’s what my company i-Beacon’s health record used to do before the axeman came!)

Now there’s an interesting piece of software called Gencache that tries to do all three pieces — linking people, linking people to information of interest and providing messaging. If you’re interested in this sort of thing, ask Gencache Jay Limaye for a username and take a look around. He gave me a tour the other day, and although it’s early days for the software, it might be a proto-type for that all-in-one forum for that elusive medical community that we’re always hearing about.

There is a new and enormous report out, at a very attractive price, from BCC Consuting. It’s called Going Mobile: Choosing the Right Inpatient Solution. Be warned that this is a huge reference work of about 200 pages, not a quickie overview. (You should download the PDF–my link is to a overview screen). But the little I’ve dipped into so far seems to be very thorough research. Each company profiled has a 4 page review including details on their technology strategy, funding, customers, and applications.

If you are interested in this market this is a must have — and at this price, well let me say I hope that these guys don’t keep repeating this or they’ll put us all out of business.

REDMedic is a start-up building a personal health record prodcut aimed directly at end-user consumers. Their wrinkle in the space is that people can take a key-fob or card with them so that if they get into an accident an emergency room nurse can look up their information online via an emergency log-in screen. (Full disclosure: I discussed a possible consulting role with REDMedic a year or so ago but never did anything with them). Nothing wrong with this idea. I thought it was pretty good when i-Beacon (my company) did it in 2000. Dr Koop thought they had it down when they did it in 1998, and Medicalogic had the same conclusion in 1999 with their PHR. Of course the same was true for PersonalMD, HealthAtoZ, iMetrikus and about 35 other companies, including WellMed which survived and is now part of WebMD. Note that they survived and everyone else didn’t. Which may give you a hint about what I told REDMedic were the dangers in their business model.

I have no problem with REDMedic claiming this is a revolutionary idea (although it isn’t as PersonalMD had exactly the same ER room access to the web/fax “emergency card” in 1999)–after all every start up should blow its own horn. What slightly annoys me is that they’ve convinced a not very worldly journalist at Information Week of the same thing in an article called Digital Health Records Move Closer To Reality. Come on team, 2000 wasn’t that long ago. Surely someone apart from me remembers it?

I genuinely hope that REDMedic’s service aimed at consumers takes off, though at $36 a year and no one using PHRs, they are at the very bottom end of the “S” curve adoption, and likely to stay there for some time. Recasting their service as a web-based Medicalert bracelet may even work, as people are more web savvy than they were a few years back. But the same issues that stopped the other 35 companies from having success–such as the unwillingness of providers to get data from their systems into the PHR, and health plans deciding that they didn’t have to improve their web service to their consumers in order to keep them as members–have not gone away to any appreciable extent. However, the online PHR, like online banking, is one of those things that will take off at some point, and whoever is alive and kicking in the space then may make out well.

Meanwhile, if you want to get into this business easily, there’s some very nifty software sitting in a box in San Francisco that I could get into your hands cheap!