I want to start with a statement: “I am not a developer”. I learned Pascal (on an early Apple Computer with 2 (sic !) 180k (?) floppies) way back in the early 80s, learned a little Fortrun (did I even write this correctly?) in University in 1990 and wrote about 10 lines of ABAP in 1998. That’s it. Oh, I know a little SQL and XML. …so much for my vast development skills!
I am working on specific topics in the pharma and life sciences space and one issue that emerged is the US Sunshine Act, intended to shed some light on the amounts of money that pharmaceutical and medical device companies spend with healthcare professionals on free meals, educational material, seminars and so on. The whole is intended to make the connection between a physician and the pharma company a bit more transparent for the consumers. Most companies have tackled this and have already published these data…typically in a flat list that (maybe) you can filter by physician name like this Payment Report. The problem is mainly to accurately collect the data from all the different IT systems that companies typically have. To build a report on these data and to share them in a controlled fashion should be fairly easy… so I thought.
Now I am sitting here and was looking for a way to prototype a simple report based on a table with about 10 columns and a couple 100.000 entries. As a non-developer (see above) this normally meant to look for a developer who understands how to do this, get the money to pay them and all this in no time. …. of course! And then I found Lumira! And to say one thing…. creating the sample data in EXCEL was the hardest part. Once I had a simple spreadsheet with sample data, the loading and creation of charts and a dashboard was rather easy.
Just a couple of things you should look at:
- The desktop application freezes every now and then. So save early and save often. Saving a visualization or a dashboard doesn’t save the entire file.
- I used the fomulas in Lumira quite a lot. This way I can refresh the source data (in the first few columns) and the calculated values are recalculated. Nice! I can control the calcualtions while somebody can suipply the raw data. So don’t do calculations in the source, do them in Lumira.
- I finally managed to get my data onto a map as well. As all data are US-based all physicians reside in a US town in some state. So I didn’t had a “Country” column in my data. However you have to, otherwise the geo-plotting doesn’t work… the appliation actually freezes. And unfortunately the visualization always starts with a world map and you have to zoom to see the US only. I didn’t find a way yet to fix and store the zoom.
So within one day after actually understanding what Lumira is actually doing, I registered myself, I downloaded the application, I loaded a medium size data set (35.000 lines), I figured out how to report on the data, added formulas and defined several visualizations incl. a map; I put the visualizations into a Board, combined several Boards into a Story and shared the visualizations. … and this is one result (the legend is in the visualization but doesn’t show up in the board …):