Editor's note: This blog is part of Fiscal Fitness, a regular H&HN series exploring the cost containment strategies hospitals are employing in response to reimbursement pressures and an uncertain economic climate. Read more at our Fiscal Fitness page.

Avid readers of Fiscal Fitness know that the multimedia series is about more than merely cutting costs. During the past several months, we've tried to highlight organizations that are aggressively attacking the significant financial and operational challenges that lie ahead. They are using Lean, Six Sigma and other performance improvement techniques to reinvent themselves, change how they deliver care and ensure that they remain viable even as the landscape around them is continually shifting.


For all of the hospitals and health systems we've profiled thus far, there's been one major constant: understanding the power of data. Whether we've focused on the supply chain or energy costs, or utilizing Lean, data is king.

At Danbury (Conn.) Hospital, the power of data is helping clinicians — namely doctors — dramatically change their practice patterns. Since 2009, they have cut charges by $6 million. I know, I know, much of what we've written about in Fiscal Fitness is how hospitals are becoming more efficient and improving their bottom line. Trimming charges isn't exactly the same thing, but it isn't a square peg into a round hole either.

As Matthew Miller, M.D., chief medical officer at Danbury told me during an interview last week, "The imperative to manage costs without sacrificing quality and safety is just getting more urgent" as reimbursements are continually squeezed.

So, in 2009, Danbury started a project to take a deep dive into specific DRGs and assess utilization. One thing became abundantly clear: there was pretty wide variation and doctors were ordering more tests than seemed clinically necessary. Like so many other institutions, Danbury had evidence-based order sets for physicians to follow. However, Miller said, those order sets were essentially a menu of everything that the doc may want to order. Once CPOE was introduced, well, ordering was made that much easier with the simple "click" of a button.

Now, to be fair, Miller noted that there are some strong philosophical reasons why docs order extra tests and labs. "We have a culture of curiosity; doctors want to know answers. Defensive medicine also drives it." But ultimately, he says, it wasn't clear why certain tests were being done on certain patients. For instance, looking at heart failure patients, echocardiography utilization was around 60-65 percent. After crunching the numbers and studying clinical best practices, clinicians dropped that down to 25 percent.

Miller told this more recent anecdote: "I did a study a few weeks ago on CT angiograms, which are typically done to rule out a patient getting a blood clot in their lungs. I looked at all of the CT angiograms being done by department, looked at what percentage of patients with a certain diagnosis had a CT angiogram and what percentage were positive. When we looked at all cancer diagnosis, we found that around 5 percent had a CT angiogram. Now, that's not a lot, maybe 100 of them, but what percent were positive? Zero. Zero. You show that data to the oncologists and it makes the doctor wake up."

Danbury is now evolving the project from just looking at DRGs to taking a more horizontal view of utilization. So, for instance, head CTs. A typical example, Miller said, is the confused patient who is on a lot of medications, but has not fallen or hit their head. The gut reaction from most clinicians, he said, is to get a head CT. It is not his goal to eliminate scans for potential stroke patients, but rather those that got confused because they have pneumonia, for example. "Those are the ones I want to get at and get those cross diagnoses. So you have to partner with radiology and say, 'You are seeing all of these requests. You know perfectly well that the indications are soft on many instances. Help us educate the clinicians to get better on the indications" so they know when to order or not to order the test.

Ultimately, Miller said, it comes down to data. Show docs the data and they'll change. At Danbury, that seems to have resulted in more efficient — but still high quality — care.