Analyze this: data shines within and without

“Now we are able to capture in real time, on an almost daily basis, what was the defect profile handed to us. Now we’re going back to targeted, high-frequency sites to work with them to improve what they’re sending us.”

One major preanalytical defect already spotted using this approach came when Henry Ford switched to Epic in 2014. The system asked ordering clinicians to specify whether the sample would be collected in the clinic or by the laboratory at a later time.

“When the patient showed up at the phlebotomy site with the clinic-collect order, the order would not cross the interface,” Copeland says. “Someone had to make a call to the ordering provider to rehabilitate the order. If you didn’t have this system of tracking and trending, we couldn’t focus on the select few clinics that make up a huge gamut of nonvalue-added work for which we were interrupting operations.”

A laboratory manager met with targeted clinics to explain what the problem was and to train clinicians on how to place the orders and choose the right type of collection, resulting in a greater than 90 percent reduction in this type of mistake.

Analytics also offers laboratories the promise of quantifying how improvements in their performance can cut overall health care costs, Dr. Sharma says. For example, when the microbiology laboratory at Henry Ford brought in MALDI-TOF, the time it took to identify Candida glabrata and other organisms fell from 4.5 days to 2.5 days.

“For the leadership in our microbiology laboratory, that two days was significant. So they sat down with the ICU folks to find out the downstream impact on length of stay. Just for Candida patients, the length of stay went from 14 days to 10,” he says. That cut in length of stay saved nearly $20,000 for every such patient. Now, the microbiology laboratory is looking at what downstream effect the use of MALDI-TOF to identify other microorganisms is having. While such calculations must be done manually, Dr. Sharma says analytics products could help laboratories prospectively model the cost impact of new technology and the changes it brings to lab and hospital operations.

“Business intelligence companies can come up with software to help us do this,” Dr. Sharma says. “Then laboratorians will have a far better and more data-driven justification for new investments—or not to do investments that may not pay off.”

NorDx, which operates 11 laboratories and 23 patient service centers and is owned by the Maine Health system, started using Visiun in February to help measure turnaround times and test utilization.

“We’re using that information and working with clinical management to look at, for example, whether people are ordering TSH and T4 at the same time when it’s not medically indicated,” says NorDx president Stan Schofield. “We’re working on that and developing utilization guidelines for medical staff as part of ACO preparation and participation.”

Visiun’s offering does have a learning curve and is “not a universal cure-all,” Schofield says. Visiun was awarded a sole-source agreement for the 26 health systems that are part of the Compass Group, a 501(c)(6) business league that Schofield cofounded and of which he is managing principal.

“It’s kind of a bolt-on to LISs, and it offered a low-cost option for some kind of data extraction,” Schofield says. “And it seems to be gaining adoption and credibility within the Compass Group.” A handful of the health systems that are members of Compass have adopted Visiun.

Schofield says that once the tool is mastered, it can offer a good deal of value.

“It’s pretty easy for people to write the rules. If I were to say, ‘How many patients have hemoglobin A1cs done more frequently than 60 days?’ it’ll pop up. They can write the rule in half an hour and run it,” he says.

Schofield says that many of the options in the emerging laboratory analytics space are “terribly overpriced” and that “a couple of them look and feel really well, but most of the labs couldn’t afford it.”

Another player in this market is hc1.com, one of the options the Compass Group considered. Castle Medical, a toxicology laboratory in Smyrna, Ga., that focuses on serving clinicians who specialize in pain management, uses the hc1.com cloud-based offering to manage its relationships with its physician customers. One aspect of the hc1.com offering that especially appealed to Castle Medical was its capability to deliver specialized reports to physician clients.

Castle Medical wants to help its physicians gain an overall understanding of how compliant their patients are in taking the opioid painkilling medications they are prescribed, as opposed to diverting them or otherwise misusing them.

“hc1 is helping us lead the way in this,” says Robert Mitchell, the laboratory’s president of sales. “We can report back to the practice: Hey, your patient panel is 50 percent compliant with the meds you’re prescribing, and here’s how that compares to the data pool in your region where it seems that you’re average or out of range. Maybe you need to test more frequently, or perhaps you’re below range and don’t need to test as often. That way, we provide the analytics so physicians can make a better decision on how to treat their patients.”

The plan is to provide the reporting to all clients beginning in the first quarter of 2016, along with a physician access portal, Mitchell says.

“What the analytics gives is a personalized approach to medicine. It’s not just risk assessment. It’s constant monitoring of your patient base. There’s always an inflow and outflow of patients, and your practice is always changing. What works for you in Q1 of 2016 may not work in Q4. Your patients may be superbly compliant in the first quarter. But what if you’re now dealing with a different group of patients with a higher risk of diversion or abuse? That’s what hc1 provides—real-time analytics so you’re never letting that pendulum swing too far one way or the other.”
Mitchell acknowledges the hc1.com offering is “definitely in the upper end of the [price] range, but the bottom line is an old adage: you get what you pay for.” He says the analytics tools, combined with customer relationship management features that go well beyond a “glorified Rolodex,” make the product worth the price.

“You don’t step over dollars to get to pennies,” he says. “You’ve got to invest in the right stuff.”

The consensus among those diving headfirst into the world of laboratory analytics is that these tools can reveal useful and sometimes surprising nuggets of gold otherwise buried in mountains of undifferentiated data points. Another point of agreement is that sucking up the data and analyzing it, while altogether necessary, is far from sufficient.

“In terms of analytics software, there are a lot of products out there, but they are not really being used super effectively,” Dr. Brimhall says. “They are being used to write simple reports in a lot of cases. But we don’t know the extent to which those reports are driving projects that lead to cost savings or quality improvement.”

In October, Dr. Brimhall led the first in a series of intensive, multiday training sessions dubbed the Orchard School of Medical Analytics. The training session will be offered again in April and June of 2016. The idea is to help laboratory professionals expand and integrate their analytical work and put into action what they may already be doing with analytics.

“People are just not getting the help they need to make meaningful analytics projects work in the organizations where this software has been purchased,” Dr. Brimhall says.

Dr. Sharma, of the Henry Ford Health System, strikes a similar chord.

“It’s great to buy these solutions and get some data, but where’s your governance mechanism to do something about that information?” he says. “No one can sell you that. You have to do that hard work yourself, and create it in collaboration with your other clinical departments.”
[hr]

Kevin B. O’Reilly is CAP TODAY senior editor.