Friday, August 9, 2013

It Needs to Print Well

Yesterday I had the annual review with my doctor. Besides the standard protocol I got a chance to discuss my views and plans for possible enhancements to the technology we use. I got really valuable feedback, encouraging enthusiasm, and plenty of new ideas as well.

There is one thing I really need to pay attention to. While it is really useful to be able to explore the data interactively on a computer screen, the IT systems and also the practices are still based on handling stacks of paper. However nice and fancy a visualization is on the screen, it also needs to print well.

This time, we spent quite a bit of time with my doctor, writing down values from my pump and meter, with a pen, on paper. The Accu-Chek Spirit Combo system has a software system to download log data from the devices to a computer, not all information is accessible.

For instance, the basal rates are only presented as graphical presentations. The adjustments to treatment often involve fine tuning the rates, and it is important to know whether the basal rate is 0.70 or 0.75 units per hour. It's really hard to tell that distinction from the graphics currently available. Only the total amount of insulin per day is printed as a number, not the values for the individual steps.

A capture of the report produced by the Accu-Chek SmartPix software.
It's hard to tell the exact levels of basal insulin from the graphic.

Another subject to modifications are the parameters used by the bolus calculator, an assistant program in the glucose meter that helps setting the correct bolus for a meal. These include target range for glucose level, carb ratio (amount of insulin required for a gram of carbohydrates) and sensitivity to insulin, and have different values depending on the time of day. These values are not included in any of the reports produced by the software. Instead, I must manually access them via the meter and tell each value to my doctor who then writes them down. This process alone easily ends up taking a few minutes. I'd much rather spend that time actually discussing the values and possible changes to them. Fortunately my doctor was not in a hurry.

I have had similar problems with a continuous glucose monitor in the past. There are two different models available, the Medtronic Guardian REAL-Time and the Dexcom Seven Plus. I've had the feeling that the nurses are a bit reluctant to give out the Dexcom one. I've used it once, and think I now know the reason.

The results of the continuous glucose monitoring need to be printed as well, as there is no suitable place for them in the electronic health records. And it is also easier to annotate the printed versions.

The problem is that the software used to visualize the data from the Dexcom device does not print the essential information of amounts of carbohydrates, insulin doses, or exercise. On screen you can access the information of each event by bringing the mouse cursor on top of a respective icon. But when you print the information, it's just a Word document containing the screen captures as such, and you just get the icons. So, having printed all the data we then spent many many minutes with the nurse going through each meal, insulin dose, and exercise for the whole week. I'd read the values from the program and she'd write them down on the prints, with a pen. We certainly did not feel productive at all.

On screen you can interactively explore each value independently.

So I'm just marking it down here as a design principle. Perhaps not the most important one, but clearly one that must not be forgotten. However fancy a visualization is on the screen, it also needs to print well.


  1. Scott Johnson from mySugr just blogged about the same topic, with a different angle.

  2. Great post, Mikael, you're right that we need to keep it relevant to those who are helping us at the doctor's office.