
Case study: How clinical analytics helps with meaningful use
Whether a healthcare organization is attesting to Stage 1 or Stage 2 of meaningful use this year, the process may not be entirely intuitive for physicians and other clinicians responsible for documenting patient care. At Massachusetts General Hospital (MGH), physicians have a little bit of help from the hospital’s Queriable Patient Information Dossier (QPID), a clinical analytics engine developed at the hospital that provides actionable insights drawn from patient information stored in EHRs and data repositories, makingmeaningful use attestation just a little bit easier for harried clinicians.
“You can imagine that with all the components of meaningful use that our physicians are responsible for, they are begging the hospital leadership to help them get the data that they need to put into the chart and do it in a way that doesn’t require them to take their energies away from seeing the patient,” says Dr. David Ting, Associate Medical Director for Information Systems at the MGH Physicians Organization.
“One example of this was smoking status, which you have to document in a coded field,” he explained toEHRintelligence. “What we found is that most of the time our doctors do document it, but in a text record. They’ll say, ‘Here’s a 28-year-old woman with a history of smoking for the past two years.’ So, clearly that is a woman who smokes, but it’s not entered in a coded field, so it doesn’t count for meaningful us even though it’s in the record.”
Dr. Ting realized that instead of browbeating physicians into changing the way they operate, QPID could be used to identify data in the free-form text fields and automatically populate the correctly coded field. “We said, ‘Let’s create a service where QPID scans all of the notes for all of the patients seen at Mass General within a certain timeframe, and do an analysis based on what you find in the chart about who was not a smoker.’ And the reason I said not a smoker was that I needed to make sure with high specificity that we can tell that a patient definitely does not smoke,” Ting said.
“After doing the study, we found that QPID was 99% specific. If QPID says you’re not a smoker, then you are not a smoker. In fact, we did a head-to-head study using human nurses doing chart reviews, and QPID came out better than the human nurses because the humans were missing things. We took that data from the text fields and we automatically put it in the EHR in a coded field, so we managed to fulfill that measure without bothering the doctors. They could document it where they were used to and still meet the requirement for meaningful use.”
Massachusetts General is in the process of attesting to Stage 2 of the EHR Incentive Programs, and is also in the middle of switching out its homegrown EHR in favor of Epic Systems, Ting said, while exploring ways to hold on to the rich functionalities that QPID provides. “Certainly there are search tools within Epic, but they don’t have the whole galaxy of functionalities that we’re used to,” Ting points out. “When you go to a vendor product, you lose a lot of the flexibility of having control over your own system design.”
The organization will, however, gain the ability to standardize its technology across multiple care sites, as well as reduce the amount of time and effort being spent to develop and maintain health IT applications in-house. “We realized that we can’t sustain being a software company,” Ting says. “We are really an integrated health delivery system, and so our main initiative is to take care of patients. That’s where the Epic decision came in. Our challenge right now is figuring out how to bolt QPID onto our future EHR as much as possible. We need to be aware of where that integration can happen.”
Source: www.qpidhealth.com