Search the KHIT Blog

Tuesday, November 19, 2013

Healthcare "Big Data" will not be a panacea

Risk Calculator for Cholesterol Appears Flawed
By GINA KOLATA, New York Times, November 17, 2013

Last week, the nation’s leading heart organizations released a sweeping new set of guidelines for lowering cholesterol, along with an online calculator meant to help doctors assess risks and treatment options. But, in a major embarrassment to the health groups, the calculator appears to greatly overestimate risk, so much so that it could mistakenly suggest that millions more people are candidates for statin drugs.

The apparent problem prompted one leading cardiologist, a past president of the American College of Cardiology, to call on Sunday for a halt to the implementation of the new guidelines.

“It’s stunning,” said the cardiologist, Dr. Steven Nissen, chief of cardiovascular medicine at the Cleveland Clinic. “We need a pause to further evaluate this approach before it is implemented on a widespread basis.”...

Making sure we come to scientifically accurate and effective conclusions across the vast expanse of health care conditions are treatments will not be easy. Given increasingly easy access to "big data," there will be much naive quantitative enthusiasm, some of it merely wasteful, some of it dangerous to deadly.


apropos of data, I joined this Meetup group. Our first meeting is tonight.

First Meeting- discuss format of meetings, network
Wednesday, November 20, 2013
7:00 PM to 9:00 PM

Hacker Dojo
599 Fairchild Drive, Mountain View, CA

Hi everyone,

With the help of a meetup member Steve Banville, we have a conference room reserved at Hacker Dojo for Nov 20th.  Thanks Steve for your help.  I have also been speaking to a few of the members offline and it seems that it would be best if we start off with a meeting to discuss the agenda, the objectives, the format, frequency of meetings, etc of the Meetups.  I want this Meetup to be a real learning experience rather than a gloss over of methodologies.  It would be very useful to discuss how we think we can make this relevant for us.  I will send out a more detailed set of topics that I'd like to discuss in the 1st meeting. 

In the meantime please rsvp to let me know whether you can make it.  We may need to adjust the conference room reservation based on the number of people that rsvp.

Thank you, Seema.
Should be interesting.
This group is for anybody who practices Statistics, loves data and wants to grow their skills. Objectives of the meetups would be hearing from guest speakers and some deep discussions on statistical methods and statistical and programming techniques. All statisticians are welcome, from Big Data Scientists to Survey Statisticians and analysts.
A couple dozen stats people have signed up to attend. Link to the group is here.

Whose data is it [sic] anyway?

A common and somewhat unique aspect to EHR vendor contracts is that the EHR vendor lays claim to the data entered into their system. Rob and I, who co-authored this post, have worked in many industries as analysts. Nowhere, in our collective experience, have we seen such a thing. Manufacturers, retailers, financial institutions, etc. would never think of relinquishing their data to their enterprise software vendor of choice.

It confounds us as to why healthcare organizations let their vendors of choice get away with this and frankly, in this day of increasing concerns about patient privacy, why is this practice allowed in the first place?...

Prior to the internet-age the concept that any data input into software either on the desktop, on-premise or in the cloud (AKA hosted or time sharing) was [sic] not owned entirely by the users was unheard of. But with the emergence of search engines and social media, the rights to data have slowly eroded away from the user in favor of the software/service provider....

Of course this is not a good situation when we are talking about healthcare, a sector that collects the most personal data one may own. EHR purchasers need to take a hard detailed look at their software agreements to get a clear picture of what rights to data are being transferred to the software vendors and whether or not that is in the best interests of the HCO and the community it serves....

The second data ownership challenge to be very careful of is the increasing incorporation of patient generated health data into the healthcare delivery system. We project an explosion in the use of biometric devices, be it consumer purchased or HCO supplied, to monitor the health of patients outside of the exam room. Much of this data will find its way into the EHR. Exactly who owns this [sic] data and what rights each party has is still debatable. It is critical that before HCOs accept user data they work out user data ownership processes, procedures, and rights....
Yeah. We've been arguing about this all the way back to DOQ-IT and beyond. ePHI data ownership; a hopeless perplex of IP case law, muddled federal regulation, and widly varying state statutes. Who can "own" facts about me? It's not at all clear.

"Whose data are they anyway."

Well, that's ugly, I have to admit, my pedantic defense of "data are" notwithstanding.

Nevada Health Information Exchange Taps Orion Health HIE to Enhance Care Coordination Across the State
November 18, 2013,

Orion Health, a company focusing on eHealth technology, the Nevada Health Information Exchange (NV-HIE) and Nevada'sDepartment of Health and Human Services (DHHS), announced that the NV-HIE Board of Directors selected the Orion Health HIE to power and enable a statewide electronic health information exchange.

According to a release, Orion Health HIE was selected to provide the technology that will support NV-HIE services and programs that advance trusted information exchange for the coordination and continuity of health care for all Nevadans - anywhere, anytime...
DHHS is responsible for the State HIE Cooperative Agreement grant* awarded to Nevada as part of the 2009 federal stimulus bill. Under a grant sub-award from DHHS, the NV-HIE is utilizing state HIE grant funds to establish core HIE services that facilitate the trusted electronic exchange of personal health information, support the adoption of electronic health records (EHRs), and enable intra- state, interstate, and nationwide HIE.

DHHS is already offering Orion Health Direct Secure Messaging, as Nevada DIRECT (NV DIRECT), which enables standards-compliant communications between various health care providers and organizations. NV DIRECT will transition to NV-HIE and become part of its core HIE service offerings by the end of 2013.
I'm having trouble understanding how Nevada will be able to support two HIE's. The state, while geographically huge, has a relatively small population, more than 90% of which is clustered in two population centers: the 2 million-plus Las Vegas area, and the smaller Reno-Tahoe area.

HealtHIE Nevada (which yours truly named, btw -- shot these pics also) has had boots on the ground for more than a year. DHHS has had this grant money for nearly four years. Just now getting around to deploying?

Interestingly, DHHS does have the statutory authority to "regulate" HealtHIE Nevada out of business (SB43, now part of the Nevada Revised Statutes).
Sec. 6.1. The Director shall establish or contract with not more than one nonprofit entity to govern the statewide health information exchange system. The Director shall by regulation prescribe the requirements for that governing entity...

4. The Director shall by regulation establish the manner in which a health information exchange may apply for certification and the requirements for granting such certification, which must include, without limitation, that the health information exchange demonstrate its financial and operational sustainability.
Should be interesting to see how all of this plays out. Orion versus Optum. The bare-knuckle barroom brawl that is Nevada politics.

Below, my former colleague Rachel in a Vimeo promo video.

Nv-HIE website banner:

Maybe they should have contracted with Mana Health for website design.

"© 2013 Nevada Health Information Exchange (NV-HIE)" Seriously? This is all taxpayer money. You can't "copyright" anything.

Gotta love this. Go to the home page. What do you see today (11/20/13)?

Really Johnny-on-the-spot. July 3rd, 2013? Is anyone home?

But, perhaps the vendor's product will be just fine.

Appears that someone at Nv-HIE is reading my blog. Home page today:

Well, good.


Ten days 'til we find out whether HHS Secretary gets tossed under the bus. Below, from an NPR story (they got some stuff wrong) the fiasco in one graphic (my annotations in red).

I've posted the entire "Red Team" deck here (ugly scanned pdf). I am reminded of my earlier post on the "Agile Software Development" hokum.

"KatrinaCare" indeed.

AND THE HITS JUST KEEP ON COMIN' website 'didn't have a chance in hell'
The failure rate for software development projects is high generally, particularly large ones like, says Standish Group data
Patrick Thibodeau, October 21, 2013 (Computerworld)
WASHINGTON -- A majority of large IT projects fail to meet deadlines, are over budget and don't make their users happy. Such is the case with

The U.S. is now racing to fix, the Affordability Care Act (ACA) website that launched Oct 1, by bringing in new expertise to fix it.'s problems include site availability due to excessive loads, incorrect data recording among other things.

President Barack Obama said Monday that there is "no excuse" for the problems at the site.

But his IT advisors shouldn't be surprised -- the success rate for large, multi-million dollar commercial and government IT projects is very low.

The Standish Group, which has a database of some 50,000 development projects, looked at the outcomes of multimillion dollar development projects and ran the numbers for Computerworld.

Of 3,555 projects from 2003 to 2012 that had labor costs of at least $10 million, only 6.4% were successful. The Standish data showed that 52% of the large projects were "challenged," meaning they were over budget, behind schedule or didn't meet user expectations. The remaining 41.4% were failures -- they were either abandoned or started anew from scratch.

"They didn't have a chance in hell," said Jim Johnson, founder and chairman of Standish, of "There was no way they were going to get this right - they only had a 6% chance," he said...


In the United States, 2 million people are infected with drug-resistant “superbugs” every year, and at least 23,000 die as a result. Such numbers, journalist Maryn McKenna suggests, will seem trivial if we reach the point when all antibiotics are no longer effective — something that’s on track to become a reality.

Considering the full implications of a post-antibiotic era, McKenna concludes that it wouldn’t be so different from the apocalypse. And to know what we’re facing, we need only look at where we’ve come from:

Before antibiotics, five women died out of every 1,000 who gave birth. One out of nine people who got a skin infection died, even from something as simple as a scrape or an insect bite. Three out of ten people who contracted pneumonia died from it. Ear infections caused deafness; sore throats were followed by heart failure. In a post-antibiotic era, would you mess around with power tools? Let your kid climb a tree? Have another child?
To start with, McKenna writes, the loss of antibiotics will mean the end of modern medicine as we know it, impeding everything from surgery to chemotherapy to the far more prosaic:
At UCLA, [Dr. Brad] Spellberg treated a woman with what appeared to be an everyday urinary-tract infection — except that it was not quelled by the first round of antibiotics, or the second. By the time he saw her, she was in septic shock, and the infection had destroyed the bones in her spine. A last-ditch course of the only remaining antibiotic saved her life, but she lost the use of her legs. “This is what we’re in danger of,” he says. “People who are living normal lives who develop almost untreatable infections.”...
A huge potential problem for human health.

OBSERVATION is a half-billion dollar site that was unable to complete even a thousand enrollments a day at launch, and for weeks afterwards. As we now know, programmers, stakeholders, and testers all expressed reservations about’s ability to do what it was supposed to do. Yet no one who understood the problems was able to tell the President. Worse, every senior political figure—every one—who could have bridged the gap between knowledgeable employees and the President decided not to.

And so it was that, even on launch day, the President was allowed to make things worse for himself and his signature program by bragging about the already-failing site and inviting people to log in and use something that mostly wouldn’t work. Whatever happens to government procurement or hiring (and we should all hope those things get better) a culture that prefers deluding the boss over delivering bad news isn’t well equipped to try new things.
From and the Gulf Between Planning and Reality

More to come...

No comments:

Post a Comment