Getting Out of the Box
Walter "Buzz" Stewart
CEO, solving big & small healthcare problems with real-time data
Do you know the feeling when you’ve worn a cast for an injured leg or arm? At first, it's so restrictive you wonder how long you can stand it. Then, you gradually adjust. The initial discomfort and confinement fade over time, creating a new reality with a feeling that “it’s not so bad.” But once the cast is removed, you genuinely sense relief and an appreciation for newfound freedom.
We can all get into a mental box the same way when stuck with something we cannot change. We gradually adjust, become de-sensitized to the constraints, and lose sight of being stuck. We accept this version of reality for what it is until something happens that helps us recognize the self-imposed constraints.
This mental box is a phenomenon common to healthcare when HIT vendors develop innovative solutions that use EHR data to make it easier to get work done better and faster. Invariably, they are hit with severe constraints to the EHR data they can get in real-time. As a consequence, their solutions are dummied down from using data to actually do work for customers to just displaying data where customers only know what work needs to be done. This is a common scenario, and it’s at the heart of many failed HIT vendor solutions.
I recently spoke with a colleague who is an executive of a well-known healthcare data analytics company. Like many others, his company offers dashboards that describe what happened. His customers are continually aware of problems that need to be solved for things like care utilization, quality, and risk management. While his “real-time” dashboards use HL7 v2 messages, vendor APIs, and FHIR API(s) to access snippets of EHR data, for the most part, they rely on nightly data batches to deliver a rearview picture of how things have been working.
My colleague has deep knowledge and experience using metrics to understand how things are working in healthcare. I wondered why he wasn’t putting that knowledge to better use by using EHR data to simplify or automate work, instead of just pointing to problems. I mentioned it’s now easy to get any EHR data in real-time, and he could supercharge dashboards. His response: “I already get the data I need in real-time.” ?
领英推荐
The mental box strikes again! He meant, without recognizing it, that he had adjusted to making compromises in what his solutions could do because of the constraints of the data he could get. His thinking and his product were constrained without him knowing it. He needs his “cast” removed.
Healthcare systems have been settling for looking in the rearview for decades. The mental box gets in the way of seeing solutions that get them everything they need. But, once healthcare systems are liberated, they act with their newfound freedom to scale strategic advantages and deliver patient experience.
?
?
?
Investor
1 年Outstanding piece, Buzz. I applaud the focus on real time proacting with the data rather than focusing on looking in the "real view mirror" with outdated large chunks of often noisy legacy DBs. Keep up the good work.
3/3 (p.s., I still refer to the Brownson reference you pointed me to. Thanks.)
2/3 “with what” and “when” being pre-specified to all working along the fulfillment process; principals know what’s needed prior to the start of the job. For healthcare, yes, it’s engineer-to-order, but without that prior knowledge. All principles do not know what to do next to advance the next priority to-be discharged or admitted ED patient. Lab does its work, but based on what communicated priority? Based on its objective function to maximize its own productivity? Or based on the next to-be discharged or admitted patient, as clinically determined? In efficient automotive flow manufacturing, all parts to be assembled are sequenced in order of the next to-be assembled car. I need—we need—real-time information integrated across all suppliers and resources (i.e., lab, imagining, etc.) that tells them, in real-time, what they must work on right then in order to facilitate the progress of the next-to-be discharged or admitted ED patient. Everyone and everything work in concert. Real-time is essential because patient condition can and does change so the sequencing will change, and ought to be allowed to to a point. No more mis-synchronization, no more waste of many partial fulfillments. Do you have that?
1/3 Okay, Buzz, I’ve a challenge for you. Let me set context. In discrete manufacturing there are various fulfillment models: make-to-stock, make-to-order, engineer/design-to-order. Make-to-stock requires a forecast of quantities of what might be sold, and storing that inventory until sold. That is, what needs to be done is prescribed well ahead of fulfilling the order. Make-to-order is manufactured once a order is made, and so requires some lead time and there’s no finished goods inventory. What needs to be done is known, but when it’s to be done isn’t known until an order is placed. Engineer-to-order needs some time determining what needs to be done, and then the how of fulfilling or execution can commence. This is healthcare. When a patient presents at, say, ED, clinical staff must, among other things, engineer a treatment. And, yet, there’s an additional complexity in healthcare scenarios: the “engineering” happens as the treatment is being administered or shortly before. In the case of “discrete manufacturing engineer-to-order,” once the design of product and process are completed, then (ideally) fulfillment begins with prescribed processes steps where with the “how,”
Associate General Counsel at McGraw-Hill Education
1 年That shift of perspective of what you can expect from what you have versus what you really want or need is what transformative thinking is all about!