knowing changes everything
Think of the last decision you made (other than what two toppings to get on your large pizza). What information did you need to make a decision? What information could you live without? What made you uneasy?
How did the decision turn out? How did your dataset affect the outcome? What information did you wish you had when you made the decision?
Did you even know what you didn’t know?
I’m not entirely sure why this picture relates, but I feel like it does, and I really love this parking spot.
It’s been an exhausting week of clinical calls. Wednesday alone, I was in 6 hours of one-on-one IDIs with anesthesiologists, intensivists, and trauma surgeons. The topic was clinical monitoring and management - how certain data influences the clinical care given. Early in the week, we asked an anesthesiologist what the value of additional monitoring in the ICU would be - if I gave you the ability to measure oxygen saturation through the bladder, would you want it?
He told me, “A lot of people are reluctant to increase monitoring because they’re not sure of the value of the data.”
I’m an engineer. I thrive on data. Sometimes inconsequential data. Sometimes to a fault. It’s even part of my own personal mission statement (stolen from the admissions office at Georgia Tech): knowing changes everything.
Doctors are used to a certain dataset when they make a decision. Certain things are known - elevated core body temperature is a marker of infection, low urine output can be an early indicator for kidney failure, hypothermia impairs oxygen perfusion. The data corresponds to a current standard-of-care decision. All of these micro data points in the hospital paint a picture of how a patient is recovering.
Imagine, though, that tomorrow I gave you the ability to noninvasively measure tissue perfusion. That is, the ability to tell what percentage of nutrients that enter your body (or bloodstream) were actually being pulled into your cells. How would that change things?
What would that enable?
This is the difficulty of designing diagnostics. We have to imagine what clinical care might look like after the data becomes available. What can we enable that we weren’t able to do before? Do we even know what we don’t know yet? We’re designing for data and we don’t yet know how that data will be used. Once we know, what will knowing change?
Nowhere is this more profound than in the developing world. The same doctor told us shortly after, “The more parameters you’re monitoring, the less important [a single one] becomes.” The information we take for granted in a standard post-op ICU - heart rate, sats, respiratory rate, blood pressure - aren’t always available in a rural clinic, much less diagnostics that can detect infectious disease. Imagine how we might transform the face of a malaria outbreak if we could detect the disease in real-time using proven methods. Imagine what we could do for dengue. For cancer.
The value of data is what it enables. What treatment decision it can set in motion. What behavior it can change. Knowing changes everything - what is it that we can come to know?