Sunday, June 3, 2007

Future -The Data Mining Race: Making the Unknown Known at the Point of Care

In a Healthleader’s feature story, “Data Mining and Innovation: Keys to U.S. Reform” (June 27, 2006), and in a chapter “Employers Push to Release Medicare Claims Data” in Innovation-Driven Health Care (Jones and Bartlett, 2007), I predicted extracting and distilling clinical information for use at the point of care would be a major factor in improving health care. I observed data mining was already rampant in Medicare and pharmaceutical related industries.

Shift to Data Availability at Point of Care

Now the focus has shifted to applying IT extracted patient claims data in the private sector at the point of care, especially in emergency room and hospital settings. The idea is to simplify data access for doctors and other health professionals and make it immediately available when it’s most needed – when doctors are confronted with patients in emergency or hospital settings. Many of these patients are in critical condition or have complex illnesses.

Two organizations are leading the race to supply data extracted from existing data bases at the point of care: 1) Fusion from Carefx, Inc., a Scottsdale, Arizona, firm; 2) Medecision, Inc, based in Wayne, Pennsylvania. Their data extraction systems are complementary rather than competitive.

Carefx clients include more than 200 hospitals while Medecision clients consist of 21 Blues plans and other health plans. The marketing conduits for the two companies may differ, but their aim is similar – to make the unknown known by providing medical professionals with timely and critical information extracted from existing patients’ databases.

The Carefx idea is to funnel data from everything taking place within the hospital and to make it available at the doctor wherever he or she is within the hospital. The hospital is a very complex organization, and communication between departments and other specialists may fall through the cracks. As Peter F, Drucker observed in 1975, “Even small health care institutions are complex, barely manageable places. Large health care organizations may be the most complex organizations in human history.”

Carefx extracts its information from the hospitals’ information systems, unlike Medecision which gets its data from health plans’ databases. Both Carefx, like Medecision, simplify physician access to patient information. Their systems seek to satisfy physicians and hospitals clients by increasing work flow efficiency.

Interview with Henry DePhillips, MD

To give healthleaders readers insight into the race to simplify and speed access to patient data, I interviewed Henry DePhillips, MD, executive vice president and chief medical officer of Medecision.

The interview focuses on IT technologies that create a patient clinical summary and transmits it to the point care. This summary is designed to cut costs per patient episode for health plans, help doctors make more informed decisions, and, in some cases, raise doctor reimbursement by documenting complexity of care.

Reece: Is it safe to say you champion better use of technologies for physicians at the point of care?

DePhillips: Yes, there’s much technology can bring to physicians. Technology has brought leaps of efficiency and productivity to other industries, but not to practicing physicians at the point of care.

I’ve been working with Medecision for the last three years, and we have come up with something called the patient clinical summary. It’s an innovation that is and will be a market-mover and market-changer for health plans.

Reece: Why so?

First, I’ll give you a high level view, and I’ll then go further if you like. The bottom-line is that Medecision is in its 19th year, and all that time we have done care management systems for payers – for health plans across the country. We have expertise on payer data sets and interfacing payer data systems.

We’ve leveraged that experience to a capacity to extract useful clinical information from payer data sets – claims, pharmacy, lab results – and use our analytics capability so summarize, clinically validate, and distill that information into a concise individual electronic health record called the payer-based clinical summary.

Using our web tool, we transmit that information to the point of care in less than a minute. We’ve deployed clinical summaries to emergency rooms and other care settings across the country. When a patient shows up who’s insured by an organization that runs our software, the emergency room staff can pull down a patient-based health record and get a complete history what payers know about the patient.

This is invaluable in a high-intensity-illness- low-availability-of- information setting. Doctors who are using it are happy with it. It contains a medication list, a medical problem list, procedures that have been done, studies that have been done, doctors patients have seen, hospitalizations patients have had, and even some predictive modeling data identifying gaps in care.

That sums up what we’ve been doing. We have a third party study validating what the medical cost savings are for payer and patients associated with the use of this information, which amount to $545 savings for each transaction in the emergency room setting.

How long have you been offering this condensed payer-based clinical summary at the point of care?

: We’ve had in production for two years.

So it’s relatively new.

DePhillips: Yes, but it’s production-hardened. It’s certainly not vaporware. It’s an evolutionary document. We continue to do market research, as recently as today. We continue to improve the record and processing capabilities, and we’ll add more data sets over time. The care management data are relatively new additions.

Reece: Who are your clients – doctors, hospitals, health plans?

Our client is the payer – the insurance company. Companies using our software deploy it to their provider networks. Different payers are using it in different ways. Some are focused on emergency rooms and cost savings. Some are concentrating on doctors, some in their own private offices.

By payers, you mean health plans – large and small?

That’s correct. Our clients include 21 Blue Cross Blue Shield health plans. About 2/3s of our clients are non-Blues plans.

So you’re preaching the gospel of cost savings to payers? What is the evidence for that?

We commissioned a study by company called HealthCore to do an independent study on medical cost savings. The bottom-line was $545 in cost savings every time one of these documents was transmitted from Blue Cross Blue Shield of Delaware to the level 1 Trauma Center at Christiana Health System in Delaware.

Give me a feel of how that $545 savings is achieved.

There are four subcategories that showed statistical significant differences in savings.

First was laboratory costs. You don’t need to repeat lab tests that have already been done.

Second was a reduction in cardiac catheterizations. As you know, if someone comes into the emergency room with what looks like unstable angina, you have almost no choice but to send them to the cath lab. But if you have access to the workup that’s been done previously, you may not have to send them.

Third was medical supplies. If you have a patient that comes into the emergency room, you’re likely to want to get access to their blood stream right away, with an IV for venous access or central line. But if you have their medical history in front of you, you may not be in such a rush.

Four, there was significant increase in reimbursement for cognitive care for physicians in the emergency room, about $35 on average.

Reece Why is that?

DePhillips: What we believe happens, and this is a consensus opinion, not scientifically based, is that in using the patient clinical summary we basically make the unknown known.

For example, if the patient presents with a fractured femur, and the clinical summary shows the patient has hypertension or diabetes, the medical complexity is higher. Doctor can legitimately code more appropriately for the medical complexity of the case.

Reece: In other words, the complexity is documented. I’ve had some experience with that. A friend of mine, Allen Wenner, a family physician in South Carolina, has developed software called the Instant Medical History. The patient creates their own medical history using a “yes’ or “no” algorithm based on their age, sex, and gender. When the patient enters the exam room, the doctor has a complete narrative history. The end result is that the doctor can code at a higher level, with an average increase of $50, because of documented complexity.

Right, Wenner’s system is making the unknown known.

Reece: Or putting the obvious into complex context. In any event, you’ve been transmitting these clinical summaries for two years, and you have a clinically focused record based on payer data that’s available at the point of care. What are applications outside of the ER?

DePhillips: We’re seeing different deployments based on different business needs. We’ve proven cost reductions in the emergency room, but we haven’t yet proven it elsewhere. Health plans focusing on slashing costs use it most in the emergency room.

But we have other plans who using clinical summary reports for marketing purposes. Plans want to develop the reputation of being a doctor friendly organization. Health plans go out to employer groups and say, “We think your employees will receive higher quality care, certainly more informed care, because we supply our physicians with the previous clinical summary” We have many plans deploying it without regard to the health savings component.

It’s interesting to hear you say plans are changing their strategy so they will be perceived to be more doctor-friendly. I have a chapter in my book, Innovation-Driven Health Care, which is just out, entitled, “Health Plans and Banks Move to Ally with Doctors.” Health plans will need to ally with consumers in a market-driven environment if they hope to retain them as customers. Plans will have to be “doctor-friendly” in the consumer driven era. Among doctors, plans are in the process of converting from being perceived as the “enemy” to being an “ally.”

: When you think about it, giving payer-based clinical summaries at the point of care helps doctors make more informed decisions.

It’s important here to mention our business model, how we charge for these clinical summaries. The payer, the insurance company, pays for the transaction. It’s free of charge for the physician, and there’s no installed software necessary to receive it. We’ve created a situation to make it simple for a doctor to receive the information.

Reece: Do you think the health plans are beginning to realize that, as we move into the consumer-driven era, they will need to pay more attention to cultivating not only physicians but consumers, too.

Yes, I see health plans beginning to stutter about what to do about consumerism. Many plans have created initiatives to engage consumers in the clinical and financial process. Many of these initiatives have failed. You see a lot of stopping and starting, failing, falling back, and trying again. Payers are feeling their way as to what works and what doesn’t.

If you think about how the payer can best contribute to the entire health system, the biggest asset payers have is data. I’ll grant you claims data was not originally designed to be used in clinical care. But because of the analytics, new technology, clinical validation, and predictive modeling, you can now derive quite a bit of useful clinical data.

Reece: United Health Care pioneered exploiting claims data. From claims data, United learned, for example, who was using insulin or oral diabetic agents, They knew who the diabetics were, and they could then track their costs and outcomes.

DePhillips: That’s a good example. You get my point.

No comments: