Tim Mullaney

It pains me to say it, but Minnesota’s done it again.

As I’ve mentioned in the past, I praise Minnesota grudgingly because of my loyalties to my home state of Wisconsin. But the numbers don’t lie, and they routinely show the Gopher State at or near the top of the heap for high quality long-term care. (Please note that Wisconsin’s no slouch in this area.)

Now, the Land of 10,000 Lakes has shown how long-term care providers — and their pharmacy and acute care partners — can improve resident outcomes and streamline operations by harnessing technology. The state’s Medicare Quality Improvement Organization, Stratis Health, and a number of providers recently wrapped up a successful health information exchange demonstration project funded by the Centers for Medicare & Medicaid Services.

An official report for CMS is forthcoming, but I spoke with a few of the project leaders, who shared some of the lessons learned and described notable achievements of the initiative. There’s a lot for long-term care operators to glean from their story and from the resources they’ve put together.

Defining health information exchange

For one thing, they found that the very definition of health information exchange is murky, so I’ll start by explaining the term. The phrase might seem self-explanatory, but it doesn’t refer to just any old exchange of information among different healthcare settings — for example, sending a fax is not true HIE, nor is providing view-only access to an electronic medical record.

Interoperability is a key feature of true HIE, setting it apart from these more basic types of sharing. When data is shared among interoperable computer programs, it flows seamlessly from one setting’s system to another’s — workers aren’t stuck with receiving data in static format.

“Interoperability actually makes that data meaningful between two different EHRs or other products that have the patient’s information within it,” explained Paul Kleeberg, M.D., chief medical informatics officer for Stratis Health.

So, an initial step in piloting information exchange was finding long-term care providers using EHRs that could support interoperability. As you probably can guess, this was not an easy task, given that long-term care has not been on the receiving end of the government money that has supported EHR adoption in hospitals and physician offices.

In fact, the task was even more difficult than the project leaders expected. I first spoke with them back in September, when the initiative initially was scheduled to end. However, due to a longer than anticipated ramp-up, they had been given an extension to continue their work through March.

No money, but no time to wait

At least from where I sit, this initiative really reinforces that long-term care deserves some government largess to defray the significant expense of implementing a robust EHR. I say “deserves” because the government is expecting its Meaningful Use dollars to support health information exchange, and nursing facilities are one of the most important players in HIE. Unfortunately, there’s still foot-dragging on incentives for LTC.

And yet, for as pricey as an EHR investment might be, what nursing homes really can’t afford to do is wait to get on this bandwagon. That’s because, as the Minnesota project itself illustrates, there’s already momentum behind HIE, and long-term care facilities that are laggards likely will be at a competitive disadvantage.

The leaders of the Minnesota project appreciate that EHR costs might seem — or be — prohibitive. But, “at a minimum,” long-term care facilities should “look at what their EHR culture looks like” and “what kind of leadership they need to carry out the change management,” said Candy Hanson, program manager for health information technology for post-acute care at Stratis Health.

Furthermore, many providers have purchased a system but “have a long way to go to take advantage of all the capabilities,” said Stratis Health’s Deb McKinley, director of communications. For example, they might be gumming up their workflow by not taking simple steps such as changing default settings, she noted. Stratis Health has created a toolkit to help long-term care providers implement and maximize the value of an EHR system.

‘Staggering’ documentation issues

If the slow start to the initiative illustrated some of the more disheartening aspects of the technology landscape in long-term care, the last few months have been a different story. Once the pieces were in place, the project moved very quickly and exceeded expectations, Hanson told me.

“It took us sixteen and a half months to get to testing exchange, and then within three weeks we were testing, including with the pharmacy, who we weren’t expecting to be part of the exchange,” she said.

Thanks to the pharmacy participation, the demonstration included a robust prospective medication review pilot. The results here really distilled the value of HIE, Hanson and McKinley explained.

The pilot moved medication reconciliation upstream, so that once a nursing home informed a hospital that a patient was accepted for admission, the pharmacy dispensing for the SNF could look at the incoming resident’s medications and flag issues. This alleviated the substantial burden caused when SNF staff would reconcile post-admission, and then have to phone or fax the hospital to get questions answered.

The information exchange also revealed some glaring issues stemming from differences in how hospitals and SNFs document around medication. For example, in Minnesota, if a patient on a medication enters a hospital and continues taking the pills, the hospital is not required to document why. A clinician might write an explanation and put it in a “doctor’s note” field of the electronic medical record, but information in a doc note field is not easily transferred, resulting in a delay in getting that information to the nursing home, McKinley explained.

Of the roughly 440 medications screened through the PMR since December, about 65% were without an indication or diagnosis in the hospital medical record, Hanson said. The figure is “staggering,” but it does not mean that 65% of the medications are likely to cause patient harm, she was quick to note. However, “you can’t ignore 65%,” she said.

Boiling it down

Beyond the PMR, the demonstration confirmed the superiority of having a standardized, concise medical record for each patient, Hanson and McKinley emphasized. Having 40 or 60 pages of hard copy information soon may not be seen as acceptable, particularly in a fast-paced setting like the emergency department.

There are 15 standard fields in an electronic continuity of care document (CCD), Hanson said, and the “primary value is that the data is structured so you know what to expect when you’re looking at it.” In a test, an emergency department nurse compared a number of formats, and said that the CCD is just what she needs to care for her clients. Only two software products used in the demonstration could produce a CCD, but Hanson and her colleagues said they are speaking with companies about the imperative need for this function — and the vendors are responsive, they said.

Some of the participants are going to continue to utilize the information exchange they’ve been testing in the demonstration project. They’ll focus on improving workflow and identifying meaningful metrics. These might “inform and influence” clinical decisions by revealing acute and post-acute discrepancies. For instance, the drug Ambien has different prescribing parameters in the two settings. Achieving better alignment could improve resident outcomes, Hanson stressed.

With the completion of this demonstration project, Minnesota appears poised to maintain its reputation as a long-term care leader. Yet, to be fair, the push for HIE is a nationwide effort, and many states are making progress. So I doff my cheesehead to Minnesota … and encourage the good people at the Wisconsin Statewide Health Information Network to give me a call and tell me why their work is equally impressive.

Tim Mullaney is Senior Staff Writer at McKnight’s. Follow him @TimMullaneyLTC.