Two steps forward, one step back… (EHR designs)

HIMSS level 7 and Most Wired organizations don’t have any EHR problems, do they? We don’t have any internal work-arounds, do we? No, of course not …

Working class informatics. This is what I call it.

Recently in one of our many informatics meetings, we discussed our integrated Electronic Health Record (EHR) that already communicates with our amazing lab computer and system (must see the animated gifs for our lab robot).

We are trying to reduce the number of unnecessary lab draws on our inpatients each day. Imagine this: we draw a lab test on a patient first thing in the morning: a TSH (thyroid stimulating hormone). Result comes back hours later: 27! meaning the patient seems to have hypothyroidism, and now we may want to “add on” a free T4 level and maybe thyroid antibodies. Normally, you would: go enter another lab draw into Epic for later today or tomorrow, get another blood sample and run the additional labs, that will help design your thyroid therapy.

Add to existing specimen?

HOWEVER, we can NOW use the smarts of our EHR system, and set up a button called “add to existing specimen?”, which will INTELLIGENTLY look at the existing tube of blood in the lab for this patient, see if there is sufficient blood in the existing sample, inform the provider, and give the ordering provider a CHOICE to use existing blood or request a new sample. Sometimes you WANT a new sample to be sure of the accuracy of the test, but OFTEN, you don’t care (maybe 90% of the time) and running the additional test off the original is just fine.

Win-win-win-win-win, right? Patients are saved an additional needle-stick, phlebotomists can focus on other priorities, providers and patients get the next result more quickly, the organization spends less on duplicate lab draws for that patient for that hospital stay, insurer pays less for that episode of care.

And yet. The devil is in the details. Yes, about 80% of the time, the function works correctly, the button shows up on the right patient at the right time when the right blood is available when the provider wants additional tests. This has already resulted in a 30,000 fewer lab draws in our system PER MONTH! BUT, a significant fraction of the time (and we are still investigating why), lab specimens can be ordered, the button shows up “add to existing specimen?” the provider says “Cool! Yes” and when the lab tech sees that order, pulls up their computer, turns out this particular lab test order CANNOT BE ADDED TO EXISTING SPECIMEN.


So, now comes the hard work. Take apart the machine, see where the “bug” is (you know that a ‘real bug‘ in the machine is a real thing, right?) and then put it back together. We theorize: some orders are being placed incorrectly and our lab can NEVER run this test (it is only run at other hospitals in our system). That is one theoretical problem. Some orders CAN be placed, but truly, we have the patient’s blood in a type of tube (heparinized vs not) that worked for the original lab order, but not for the add-on. Some orders CAN be placed, but they are send-out labs that we mail elsewhere and the window to mail has passed. There are probably other reasons. Now we have to go chase them down, see where the software code needs amended to avoid that situation.

Paper – the ultimate workaround

I spotted this fraying piece of paper taped inside our laboratory processing area. Turns out, just “walking around” is an eye-opening experience for a CMIO or informaticist.

The photo above indicates those labs that the lab techs see over and over and are trying NOT to make the same mistake, shift after shift. Kind of ironic, given how amazing is the rest of our much-automated lab with spotless, stainless steel machines, blood-tube railroads, auto-balancing analyzers. And here is this, dog-eared, fraying piece of paper at the computer station. Work-around.

CMIO’s take? Despite all our kudo’s: Most Wired, HIMSS stage 7, Magnet, we still struggle with taking 2 steps forward, one step back. I suspect we are not alone.

Author: CT Lin

CMIO, UCHealth (Colorado); Professor, University of Colorado School of Medicine

Leave a Reply

%d bloggers like this: