Two years on...

Two years ago today, Erick, Catelyn Eileen, and this ol' Dawg came on board to commence a month's worth of intensive study (mostly MU Stage 1 IFR) and internal training requisite for the ensuing full-bore, full-contact deployment of our Nevada-Utah REC effort. On the following May 10th I launched this blog. It was not universally loved at HQ at the outset (our REC Executive Director at the time: "you're exceeding your scope").

Well, notwithstanding, I was neither to be intimidated nor deterred. Nor shall I stop now. There's just too much good stuff going on. Too much to continue to learn. Stuff with stakes of real consequence for the benefit of society.

___

UPDATES

I just bought and downloaded this series.


Click the thumbnail graphic above for the link. Looks great thus far. 60 bucks for 12 hours of A/V lecture material from a respected subject matter heavy hitter? Nice. I'm gonna watch it all before saying much more, and I've got other balls in the air.
___

BACK TO REC-RELATED STUFF
 
A HIMSS12 press contact from Price Waterhouse Coopers sent me this. Some interesting stuff:

Outcomes-based reimbursement is the future of the heath industry. To improve patient outcomes, proactively identify chronic and high-risk patients in this new environment, and effectively manage their financial performance, healthcare organizations must be able to provide analytics at the point of service and rely on historical and longitudinal data to manage patient populations.

One constant in all of the new care and reimbursement models is data. With the digitization of healthcare, new opportunities are rising from a marked increase in the channels, volume, and complexity of information available. Organizations will compete on how effectively and affordably they manage patient care and identify patients who need preventive care. Healthcare organizations need strategies for mining data, conducting and integrating evidence-based research, translating findings into practice, and driving the behavior changes required for patient compliance.

The United States is in the midst of its largest health information technology (IT) investment and transformation. So far, the federal government has paid $2.5 billion in incentives to 800 hospitals and 33,000 physicians for using electronic health records (EHRs). Thousands more are in the pipeline to receive the bonuses, which could total $28 billion by 2015. Paper records are fading away. IDC Health Insights predicts that the majority of U.S. providers will use an EHR by the end of 2012.

Now, they’re moving quickly to capitalize on all of this data. It is no longer good enough to know what happened and why it happened six months ago; organizations need to know what is happening in real time, what is likely to happen next, and what should be done now. This new focus on informatics applies to the provider, payer, and pharmaceutical sectors, each of which has unique expectations, needs, and challenges.

Last but not least are the patients. In this changing landscape of new technologies, regulatory requirements, and healthcare enterprise strategies, all health sectors view patient engagement as a way to drive profit, either through cost reduction or revenue increases. Now, they must find ways to engage patients in their clinical informatics efforts...

They are frank to note some persistent barriers:

Technology

Data integration and lack of standards are the biggest challenges:
Not long ago, health professionals’ principal complaint was the lack of funds to invest in health IT. The world has changed. The HIMSS Nursing Informatics Workforce Survey historically showed financial resources as the top barrier to their work. In 2011, it shifted to lack of integration/ interoperability. Indeed, all health sector professionals HRI surveyed cited data integration and interoperability struggles. Among providers and health insurers, 71% of respondents said integrating data from multiple sources was the top technical goal of their organization over the next two years. Next in the line of technical goals was data standardization (rather than documentation standardization), with 56% of the respondents indicating this as a goal and 86% classifying it as challenging to accomplish. Two of the greatest technology challenges that emerged in the research were lack of standards and lack of confidence in vendors. Lack of standards: As the Office of the National Coordinator’s Health IT Standards Committee works on standards, implementation specifications, and certification criteria for the electronic exchange and use of health information, lack of industry standards was a common theme during interviews across the sectors, even from organizations that are leaders in informatics. Nearly 84% of survey respondents indicated standardizing data would be either very challenging or challenging to accomplish in meeting their technical goals over the next two years. “If you look at the ‘meaningful use’ standards, there’s not a single data quality standard among them,” said Texas State’s Fenton. “There are organizations that have ‘pancake people’ entered in their systems — records of patients with a height of 5 inches and weight of 300 pounds, for example. That’s a problem if the medicine you’re giving is based on body mass index. There needs to be consideration for dynamic versus static data. Date of birth and ethnicity shouldn’t have to be entered more than once,” she said. Another challenge is the conflict between providing rich data and creating an easily adoptable process. For example, free form narratives in physicians’ notes contain critical details, but they require more sophisticated data mining techniques, such as natural language processing or other “listen and interpret” technology. Nearly all interviewees remarked about the importance of this non-discrete data, but they struggle to find a balance between creating too few and too many forced data fields within their clinical systems for making data discrete, and therefore, easier to mine...

That last paragraph brings to mind Praxis EMR. I have but one solo doc EP on that platform, with whom I've honestly not had much interaction (I stay pretty much in "Squeaky-Wheel World" with respect to my EP caseload), so I really can't say whether the Praxis "Template-Free" approach might get us a two'fer: "nuanced" impressionistic free-text notes and structured data for statistical mining.

Dubious, in a word.

With respect to "pancake people," I've long been a crank regarding "data quality" -- the GIGO thing that won't be addressed by HIE "authentication" protocols and Stage 2 "interoperability" requirements.“If you look at the ‘meaningful use’ standards, there’s not a single data quality standard among them.” Yeah, ouch. That has to change. My wife and I both cut our professional teeth in the contractual and regulatory "DQO" world -- Data Quality Objectives, often hewing to forensic QA standards. (we frequently worked on stuff for radiation dose/exposure litigation).

This is as close as you'll get within the proposed 2014 CEHRT standard (PDF):

The International Organization for Standardization (ISO) defines usability as “[t]he extent to which a product can be used by specified users to achieve specified goals with effectiveness, efficiency, and satisfaction in a specified context of use.” Many industry stakeholders have acknowledged that a gap exists between optimal usability and the usability offered by some current EHR technologies. However, to date, little consensus has been reached on what might help close this gap and what role, if any, the Federal government should play related to the usability of EHR technology. In June 2011, the HITPC issued a report to ONC that explored the challenges associated with EHR technology usability and user-centered design (UCD). In its report, the HITPC identified certain “desired outcomes of improved usability” including improved safety and reduced cost, clinician frustration, training time, and cognitive load for clinical and non-clinical users alike. In November 2011, the Institute of Medicine (IOM) released a report titled “Health IT and Patient Safety: Building Safe Systems for Better Care,” in which the usability of EHR technology and quality management was often referenced. The IOM noted that “[w]hile many vendors already have some types of quality management principles and processes in place, not all vendors do and to what standard they are held is unknown.” Moreover, given this concern, the IOM recommended that “[t]he Secretary of HHS should specify the quality and risk management process requirements that health IT vendors must adopt, with a particular focus on human factors, safety culture, and usability.”... [pp 38-39]

OK, but, absent data quality standards, making it easier to input garbage data (nominal "usability" willy-nilly) will have its consequences as well.

The upshot of which should be obvious.
___

OK, AS IF WE DIDN'T ENOUGH TO FRET ABOUT
 
I downloaded this today into the Cognitive Crack Pipe that is my Kindle. Apropos of all the sexy "Personalized Medicine" infatuation of late:

9. We Confuse DNA with Disease How Genetic Testing Will Give You Almost Anything
I really like the science of genetics. In high school, I enjoyed calculating the probabilities of various genotypes using the simple genetics Gregor Mendel discovered cultivating pea plants. In college, I was fascinated to learn how the selective pressures exerted by one very common infectious disease (malaria) actually favored the persistence of particular genetic diseases in human populations (sickle cell disease, glucose-6-phosphate dehydrogenase deficiency). And in medical school, I was intrigued by the mechanics of DNA: how the double helix is replicated, how it gets transcribed into RNA to make proteins, how it gets recombined so we can pass on some of our mothers and some of our fathers to our children, and how it can get usurped by other life forms (viruses) so that our cells work for them. Genetics is a wonderful mix of mathematics, evolutionary biology, and biochemistry. It’s good stuff. But I am much less enamored of the idea of testing healthy people’s genes. Some think that genetic testing will provide a road map to optimal health. Genetic testing is already useful in helping us tailor therapy to individual cancers and is likely to become more useful in predicting how well patients will respond to various drugs. And gene therapy—treatment for a specific disease that involves altering DNA itself—could, in certain settings, prove to be a genuine medical cure. But genetic testing could just as easily be a road map to widespread ill health. Already, numerous commercial enterprises exist that will take your DNA (and your money) and tell you about your future. One such company, 23andMe, promises to “unlock the secrets of your own DNA,” while Navigenics wants you to be tested “do everything you can to stay healthy.” And deCODEme hopes that genetic testing will “prompt people to do the right thing.” This commercialization of genetic testing appears to be selling health, but from my standpoint at least, it’s selling overdiagnosis. Genetic testing of healthy people is the most extreme manifestation of early diagnosis. Here the diagnosis being sought is not a disease but rather the underlying genetic predisposition for a disease. In short, genetic testing is looking for genetic risk factors. Because everybody is at risk for something, it’s a strategy that will make literally all of us sick. We already have genetic tests to screen for the predispositions to a lot of diseases—more than I could possibly cover here. And because we are in the midst of an explosion of genetic research, we will undoubtedly have even more tests by the time you are reading this. But the fundamental questions about genetic testing will not change. They are the same ones that should be asked about any early diagnosis effort: How many people will needlessly be told that they are somehow abnormal? What will we do to them?...

Genetics is not destiny
The information contained in genes is often described as the blueprint for the human body, although some scientists feel that the more appropriate analogy is to a recipe, which turns out a little different each time. There is a gene with a set of instructions responsible for eye color, another with instructions for how to make insulin, and another with instructions that may or may not enable you to roll your tongue. And you have about twenty-five thousand others. Genes are composed of only four building blocks, whose names are abbreviated with the letters A, C, G, and T. The code is formed by stringing these building blocks together; the average gene has three thousand building blocks (the range is from 252 to 2.4 million). The vast majority—over 99 percent—of this genetic information is identical in all of us. This makes sense, since we all have so much in common: we each have two eyes and one heart (with the same four chambers), we each walk erect, and so on. But the small amount of genetic information that does differ from person to person really matters. It’s a big part of what makes people different from one another. In the simplest case, genetics would be completely deterministic. Genes alone would be solely responsible for individual characteristics. But genetic variation is not the whole story. Even identical twins, who have exactly the same DNA, or genotype, are not exactly the same. Environmental factors, particularly in early life, also matter. Things like nutrition and harmful exposures to toxins or radiation affect human characteristics, even before birth, as does physical and intellectual activity in childhood. There is a broad scientific consensus that virtually all variation is the result of the interaction between genes and environmental factors. And then there is luck, or the random play of chance. The same genotypes in the same environment may still yield quite different people. This leads to a key distinction that is relevant to genetic testing: the distinction genotype and phenotype. The complete set of genetic instructions contained in your DNA is your genotype. The human that others can observe—your physical, biochemical, and behavioral characteristics—is your phenotype. You don’t experience your genotype; you experience your phenotype. And it is the combination of your genotype, your environment, and luck that determines your phenotype. Genetic testing attempts to predict your phenotype based solely on your genotype. While there’s really no reason to have a genetic test to predict an aspect of your phenotype you already know about—you wouldn’t, for example, do a genetic test to see if you had blue eyes—some genetic-testing companies are in fact promoting tests just like this. They claim they can test your genes to see whether you have trouble tolerating milk products, or whether you have problems with ear wax, or even whether you like Brussels sprouts...

10: Get the Facts
A lot of messages about health screening are simply variations on the same theme—in one form or another, they all push the idea that the best way to stay healthy is to look hard for things that might be wrong. Sometimes the messages reflect the best of intentions: disease advocacy groups and some doctors advise people to be screened because they believe it is right thing to do. Others times they reflect more self-serving motives: health-care companies, hospitals, and some doctors advise people to be screened because they are in the business of selling the service. But regardless of the underlying motivation, what you really need to know is whether these messages are supported by good hard facts.

I should start by telling you the unfortunate reality: all too often, there won’t be any good hard facts to find. There is a reason for this. Most healthy people will not soon (or ever) develop the particular disease we are trying to diagnose early. So getting reliable information about the value of early detection for the few who will get the disease requires studying a lot of healthy people for a long time. And a big, long study is a very expensive study. The numbers are impressive: a typical randomized trial of mammography, for example, enrolled around fifty thousand women, followed them over a decade, and cost tens of millions of dollars. Not surprisingly, there are not a lot of these studies, although there should be. The millions we would pay to study the value of early detection pales in light of the billions we spend putting it into practice without knowing if it helps.

But since there aren’t a lot of good hard facts out there, it is important to recognize when you are being led to believe that people know more than they do. Many messages about early detection—advertisements, public service announcements, health Web sites, and even news reports—are plainly misleading. They typically exaggerate the risks you face as a way to scare you into taking action.

Yeah, geez, we're Overdiagnosed, and then subsequently typically "Overdo$ed." And, HIT/HIE unconstrained by DQO stds, along with e-Rx are gonna make it easier than ever, Bayesian considerations and lack of software QA rigor notwithstanding.

SUNDAY UPDATE

Nice comment on The Health Care Blog:

Blackcoat says: March 4, 2012 at 11:30 am

To take a page from the field of finance, the problem here is about Black Swans and Complexity. Understanding risk, data, probability, statistics, and choice – whether in personal health or investing – is challenging on a good day. It’s hard for people to take in a constant shifting mass of information and to weigh risks and benefits.

When it comes to money or health, there’s so much emotion and important variables at stake, it’s even harder. Most people don’t want lots of information or deep statistical analysis. They just want to be told what to do, by the experts that they trust...

'eh?
___

OPINION


MARCH 6th QUICK UPDATE

Health Information Exchange in the U.S. Today

View more presentations from SuccessEHS

Interesting. 3rd slide tabulates the top HIE vendors. More thoughts about this shortly...
___

March 6th, GOP Stupor Tuesday errata

"HIE"? DOA?

Be There From Anywhere: The AirStrip Story"AirStrip Technologies built its revolutionary AppPoint™ software development platform with a vision of securely sending critical patient information directly from hospital monitoring systems, bedside devices, and electronic health records [emphasis mine] to a clinician's mobile device. AppPoint was also designed to solve core challenges in mobile software development, such as developing native applications that provide the requirements of a rich user experience while at the same time being able to scale and adapt to an ever changing world of mobile operating systems and devices. FDA cleared and HIPAA compliant, AirStrip applications are powered over wired and wireless networks, delivering virtual real-time waveform and other relevant clinical data - anytime, anywhere..."

OK, I saw this news item today that HCA Sunrise (Vegas) is now deploying some of this technology. Interesting.

HCA West (Sunrise Health System) has a seat on our HealtHIE Nevada Board. Hmmm...
___

More to come...