Health Interoperability Exposed

Good news or bad? Insurance companies "own" clinical data...

The big news of the week – Aetna’s acquisition of Medicity – understandably has created a buzz throughout the industry.

In theory, the notion of genuine health information exchange (HIE) between providers and payers makes sense. Unfortunately, the “all for one and one for all” rallying cry holds little credence among providers – particularly if payers hold the keys to the castle.

Providers view sharing patient information with skepticism, even when the exchange occurs with a “friendly” entity, like a local hospital. They worry about the security of the information and that the data they offer may be put to nefarious use – to plot referral patterns, for example, or to extrapolate other business intelligence. 

Consider the level of misgiving, then, that accompanies the idea of sharing information with a traditional adversary.

Unfortunate though it may be, providers become wary when payers declare an altruistic dedication to improved clinical care. Cynicism abounds when providers feel payers go to great lengths to minimize reimbursement. 

In other words, the Aetna-Medicity deal may do little to resolve the impasse that exists between payers and providers, despite the homage paid to the importance of clinically integrated HIE – unless, of course, the company commits to genuine two-way information exchange, and uses the information to improve its own processes and workflow.

There’s no question about the potential that claims data holds to enhance clinical care. Providers along the entire continuum of care would have comprehensive visibility into the diagnoses, therapies, treatments and medications their patients receive. Clinical decision-making could be improved and waste reduced.

Conversely, payers could improve their own efficiency by taking advantage of the EMR data that would suddenly be at their fingertips. Imagine how much time could be saved in the pre-authorization process if payers relied upon information already at their disposable, rather than continuing to require physicians to fill out paperwork, submit faxes and emails, and schedule conversations with reviewers?

Here’s another example: My own practice recently was penalized when a payer declared we were not meeting standards for prescribing generic medications – when, in fact, we exceeded the payer’s parameters. We use electronic prescribing through our EMR and our patients often pay cash when picking up their medications. Claims data alone, therefore, did not provide a full picture of our prescribing habits – but information contained in our EMR would have supported our practice.

Bottom line: the idea of clinically integrated health information exchange has potential. But few on the provider side of the equation believe the promise will be borne out in practice if left in the hands of payers. If their deal is to succeed, Aetna and Medicity needs to commit to genuine bidirectional data sharing and begin to regard the provider community as allies in the effort to improve quality and reduce the cost of healthcare. 

In short, dialog must trump dollars. 

 

HIMSS 2010

I am looking forward to the annual Healthcare Information and Management Systems Society (HIMSS) meeting this year. From my vantage point, interoperability is at the forefront. There are literally thousands of vendors present. The incredible display of varied and new technologies only serves to underscore the ever-deepening silos of medical information.

The HIMSS Interoperability Showcase is one venue to see Information Exchange in progress, but in my opinion, it represents only the fledgling state of what is needed to serve the day-to-day workflow of a busy practitioner.

This year will be an important milestone for healthcare IT, and interoperability in particular. There is so much buzz about “the stimulus” and “meaningful use,” and one feels as if we have finally overcome the coefficient of friction needed to push EMR adoption (finally) into widespread use.

When talking to hard-nosed, daily grind, community physicians, they remain extremely skeptical of receiving a single penny for IT adoption.  More importantly, they worry that physicians and hospitals will adopt IT to “get the checkboxes” for meaningful use, but no great improvements in actual patient care will result.

Much of this has to do with a compelling need to practice connected healthcare—a need to improve collaboration and act as a true care team without walls.  Additionally, a population oriented approach to chronic disease management is becoming an ever increasing priority.  However, to accomplish these noble goals, overworked and often underpaid physicians demand tools to markedly improve efficiencies in work-flow, communication, and “closing the loop.”I know that it seems as if I continually beat the dead horse of semantic integration on this blog, but I will again stress that a complete and clinically relevant presentation of data must be the foundation for managing the increasing complex transitions of care that we are seeing in healthcare today.

So if you are attending HIMSS this year, try not to observe these wonderful, yet disparate technology jewels, shining as individual displays; instead try to see the bigger picture and imagine them dazzling vibrantly together. Interoperability as a technology will play a role, but cooperation, and acceptance for change will ultimately lead to a sum greater than its individual parts.

The Medical Home and Interoperability (Follow-Up)

This post is directed to the comment from my previous dialogue on Medical Home and Interoperability as a deeper explanation to encourage a continuing discussion. 

In response to a real world example, diabetes mellitus serves as an excellent model for the need for semantic integration and population-based decision support.

Simple problems, such as comparing blood work from differing lab systems can be troubling. Comparing “blood sugar” to “serum glucose” to “glucometer” for instance; or “HBA1c” to “glycohemoglobin” to “glycosylated Hb” to “hemoglobin A1c” is in itself a daunting task. Let alone the myriad of other tests that need to be directly and indirectly compared.

What about the hurdles of simply identifying new patients in the diabetic population. Take for instance a patient who has a history of gestational diabetes diagnosed during a remote pregnancy. This information might only be present in the obstetrician’s office records or in the labor and delivery information system. If this patient were to have an eye exam years later, which demonstrated retinopathy, clearly we would like to be able to “connect the dots.” The problem, of course, is that the decision making is obfuscated by information silos and translational difficulties that result from differing medical vocabularies.

More difficulty arises from identifying negations of care. It is easier to use information technology to see if a patient has satisfied a particular quality measure, but providing a list of diabetics who “haven’t had a urine microalbumin level checked in 6 months” is technically challenging on many levels.

I believe that the times cry out desperately for improving diabetic care through the use of this kind of advanced technology.

The now famous UKPDS study showed that for every one point drop in a diabetic’s HbA1c level, there was upwards of a 14% decline in morbidity (i.e., kidney disease, blindness, loss of limb, stroke, etc.). It seems obvious to me that we must direct our most intensive efforts on treating those outliers with poorly-controlled disease, if we are ever to have any hope for a healthier, and thus less economically burdensome, population.

As a primary care physician, I am passionate about this subject, and I know my dedicated colleagues compassionately yearn to help this population. What are needed are the proper tools to accomplish this. I submit that a semantically organized interoperability platform, coupled with advanced population surveillance technology is a critical, yet difficult and poorly understood component.

The Medical Home and Interoperability

 

The patient-centered medical home concept has generated tremendous discussion during this dynamic period of proposed healthcare reform. Whether or not this model will finally take hold remains to be seen. It is clear, however, that a new paradigm in chronic disease management must occur. The combination of an aging population, increase in co-morbidity, and an emphasis on ambulatory management, places a heavy burden on technology to come to the rescue.

For those not familiar with the medical home model, it goes like this: a personal physician oversees a team that takes care of chronically ill patients at the practice level. Care is coordinated across the continuum (acute, subacute, ambulatory, long-term care, home care). Adherence to evidence based guidelines are the foundation for a payment model that may support $40-$51 per member per month, depending on illness severity and the physicians level of medical home designation.

The National Committee for Quality Assurance (NCQA) has emphasized the following:

• Access and communication
• Patient tracking and registry functions
• Care management
• Patient self-management support
• Electronic prescribing
• Test tracking
• Referral tracking
• Performance reporting and improvement
• Advanced electronic communications

Success of this crucial shift in healthcare delivery is incumbent on significant financial incentives.Clearly, the infrastructure and resources needed to accomplish the above tasks can seem formidable to primary care clinicians. The ability to automate much of this work can mean substantial dollars remaining in the hands of physicians. 

Electronic data is only a small piece of the solution. Aggregation of disparate information is another small piece. Semantic normalization, however, is crucial for difficult burden of advanced analytics and disease-specific dashboards.

Most important, in my opinion, is the ability to monitor specific populations in real-time, using advanced and meaningful decision support coupled with appropriate messaging and alerts. Semantic interoperability is thus the cornerstone for the long-term success of the patient-centered medical home model.

“Meaning” Full Use: How Many MS’s do you know?

           

Take Ms Magazine, Master of Science, Microsoft, just to name a few.  Now step into an EMR.

I have seen the abbreviation MS used in the medical record so many ways.  Here are just a few:

    • Morphine Sulfate   
    • Magnesium Sulfate 
    • Ms.  (as a term of address like Mr. and Mrs.)
    • Mississippi 
    • Millisecond
    • Mental Status
    • Mandibular Series (X-ray order)
    • Mitral Stenosis
    • Musculoskeletal
    • Medical Student
    • Multiple Sclerosis
    • Motor Strength

To a computer, the same 2 letters can mean vastly different concepts, including medications, demographics, diagnoses, procedures, and measurements.  Furthermore, the 4000 mg of morphine sulfate used to treat preterm labor, would be a fatal overdose of morphine sulfate in a routine post-op patient (where 10 mg is often a sufficient dose to treat severe pain).

We’ve certainly come a long way in diminishing the use of unsafe abbreviations (thanks in large part to JCHO regulations), but the fact remains that clinicians continue to mix standard and vernacular terms in their daily lives.  Even if we could successfully map disparate terms, what happens when a new instance of “MS” is used—Menkes Syndrome, motile sperm, mass spectrometry?

The point I am raising is not about abbreviations, it is about how we use terminology and symbols to represent reality in vastly different ways.

It is fascinating to realize how the human brain can actively discern truth through the use of context.  By understanding the association of terms or abbreviations to specific metadata, a semantic neighborhood can be constructed that allows machines to begin delineation of context as well.

For instance “MS” in the context of an echocardiogram report clearly represents “mitral stenosis,” and “MS” in a nerve conduction study might represent “milliseconds.”

It is vital that we begin to address the huge challenge of semantics in healthcare. 

There are over 540,000 words in the English language (5 times more than during Shakespeare’s time), and technical information is doubling every two years.  

Meanwhile, the desire to share data is rapidly becoming a mandate in healthcare.  The term “meaningful use” is becoming a buzzword.  Let’s just make sure that in this avalanche of data, that we find “meaning in full use.”