Always looking beyond the horizon: Lucile Salter Packard Children’s Hospital Stanford

Page 2 of 2

Free: Many believe larger health organizations should see Big Data as a viable tool in treating their patients. Have you seen any evidence to support that claim? 

Longhurst: I think clearly, Big Data is in the early part of the Gartner Hype Cycle. It’s a term that people like to throw around like “predictive analytics.” I think, has yet to really show the full benefit of  potential value. That being said, we do have some early pilot experiences with what would be consider Big Data type of projects. One of them, for example, has been around solving the problem of alert fatigue with bedside monitors and alarms. Recognizing that 95 percent of alarms that go off are not clinically relevant, the Joint Commission issued a recommendation in 2013 that all hospitals form an alarm fatigue task force to help look at this problem internally. Nobody argues that bedside monitors save lives, but it is clear now that the unintended consequences of bedside monitors, because they go off in irrelevant ways more often than not, that alert fatigue could actually mask real problems.

When we put together a task force at our hospital, I included one of our fellows in our Informatics Program because I wanted to make available to the team the incredible gold mine of data that we have been collecting here at Stanford Children’s Health for the last seven years.

Lucile Packard Children’s Hospital has been recording all bedside monitor waveforms for seven years with our vendor partner, Philips. There is really no other database like this in the nation according to our vendor partner.

Most hospitals discard their waveform data after 24 hours so using this incredible volume of data, which frankly dwarfs all of our electronic health record data, is a huge benefit. Our EMR over the last seven or eight years has accumulated about 2 TB of data. Our waveform data represents around 250 TB of data. Using the data, we can really start to hone in on where the issues are. One of the issues we saw was that children need age-stratified alarms. A two-year-old child has different needs for alarm settings than a teenager and yet our approach to doing that was based on some very, very old data. So we use our new database to generate new percentile curves for respiratory and heart rate alarms and hospitalized children. We have now implemented those in the patient care setting and we’re finding that we are actually reducing those respiratory rate alarms in a significant way. That would be one very good example, I think, of a big data enabled project in a pediatric healthcare setting.

Free: For better or worse, the government has played a large role in the evolution of healthcare IT. What are your thoughts on how the government ought to participate in healthcare IT moving forward? 

Longhurst: I think, as a bland statement, obviously regulatory requirements are critical for ensuring that patients get the best possible care in environments all over our country. Specifically, I think that there is general agreement among clinical informaticists that Meaningful Use regulation as part of the HITECH (Health Information Technology for Economic and Clinical Health) Act  has been far too prescriptive. In other words, very detailed in what it’s trying to achieve and not given freedom with incentives to innovate.

As a result, we have seen a lot of very frustrated clinicians that are punching buttons and software to meet reports and regulatory requirements that really just don’t make good clinical sense. So, I’d say, as a more nuanced statement, that it is important to have regulation, but that regulation should still allow innovation within boundaries and not be as prescriptive as our Meaningful Use legislation has been for electronic health records.

Free: Vendors also have shaped the history of healthcare IT to both greater and lesser levels of success. If you had to identify the area where vendors today are missing the plot, so to speak, where would you say that is?

Longhurst: Great question. I have been a member of several different vendor communities. I find that each company tends to have its own cultural approach to interacting with clients. That obviously stands the gambit from approaches that I don’t think are very effective as a client to those that I think that are.

At the end of the day, any good company is going to listen closely to their clients for feedback on where development priorities should be. Typically, clients are listening to who pays the bills. So within the health IT community there is a little bit of a discrepancy because the folks paying the bills are the hospital administrators, not the folks were using the software on a daily basis. So I think that’s part of where people in worlds like mine become really critically important is being embedded within hospital administration leadership gives us the opportunity to represent our colleagues with discussions with the vendors where development needs to occur, where the problems are and where the priorities should lie.