by Muz Ahmed
Muz Ahmed is one of Timmy’s long-term volunteers working abroad. After traveling with Timmy’s Microsoft Medical Team to Ecuador in November 2012, he decided to put his career as a software engineer in Seattle on hold in order to build an EMR (electronic medical record) to improve Timmy’s ability to gather and maintain patient records in its developing world clinics. For the past 8 months he has been combining his passions for health and technology to help revolutionize Timmy’s work via this EMR – TimmyCare. This blog post is Part 2 of a 3 part series. Click here to check out Part 1: Introducing TimmyCare.
For those who have read Part 1, you’ll remember the post ended with a cliffhanger – TimmyCare had declared open war on Paper and was gaining ground, but the victory seemed small and Paper barely noticed. To win would require legitimate approval from Timmy’s medical volunteers – many of whom were less that secretive about their limited enthusiasm for this non-paper solution. Ironically, many of their complaints stemmed from the look and feel of TimmyCare, which had been based on the original (and vile) paper forms. (Did you know that evil and vile are anagrams? So are paper and destroyer of rainforests)
The Paper Clinic
To appease the medical volunteers, I set out to discover why our foe, Paper, set such a high standard for maintaining patient records. Here are a few points that I uncovered about Paper:
1. It’s cheap (once you ignore the whole destroying-the-earth fact)
2. It’s cheap lightweight, easily transported, and easy to replace
3. It’s cheap highly available
4. People are used to paper – they don’t have to be taught how to use it. Therefore, seeing more patients is as simple as handing out more paper to another doctor.
Each of these factors was contributing to a high barrier of entry for TimmyCare v1 – which wasn’t cheap, lightweight, easily available, or easy to use. But one of TimmyCare v1’s most limiting qualities was its lack of access to historic patient data (anything previously written down on paper). Prior to any introduction of TimmyCare, Timmy’s staff and volunteers would haul all of the paper records to each site to ensure that when a patient was seen in one of Timmy’s clinics, the doctor had access to all of their previous histories (a tantamount point for quality of care.) TimmyCare v1 couldn’t provide this – it lacked access to the original paper data and provided no way to view past information. This simply wouldn’t fly with the doctors (and rightfully so).
Speaking of Doctors
Doctors are a great bunch – they save lives, they volunteer to come on medical trips, they like peanut butter, and so on. But a majority of doctors write like they’re being electrocuted while holding a pen between the 3rd and 4th digits of their non-dominant foot. And despite their limited penmanship skills, they seem to enjoy writing (on paper). This poses a 3-fold problem when using paper records in Timmy’s clinics:
1. During clinic, non-medical professionals struggle to read the doctors’ scribbled patient notes.
2. Post clinic, when trying to gather metrics, the reader has difficulty differentiating numbers from full out paragraphs.
3. Most doctors will fight to the near death to keep using their precious paper. This poses quite a titanic trial because they are capable of saving themselves!
To make headway in convincing them that TimmyCare would be an improved alternative to Paper, I realized that we would need to 1) Surface the previous paper histories in a user friendly fashion and 2) Ensure that the entering and viewing of data wouldnâ€™t slow them down. It’s true that with TimmyCare, Timmy’s clinic run faster. However, the doctors on average are slower, which clearly doesn’t scale. Enter TimmyCare v2, to a Daft Punk set: Harder, Better, Faster, Stronger (on repeat).
As Part 1 of this blog post described, version 1.0 was simply a digitization of the paper forms used in clinic. It was about as user friendly as Ikeaâ€™s furniture assembly instructions. It came with a steep learning curve, and did little else beyond disparaging users with red text when some arcane error occurred.
Version 2.0 was slated to offer far more information on the screen in a non-intrusive way. Doctors would be able to view summarized information from previous visits on the same screen, offer alerts for medical allergies, and present important information like pregnancies and chronic conditions in an obvious manner. Additionally, TimmyCare v2 aimed to offer more than just the same information as the paper versions of the histories – it would make data entry efficient, more aligned with the approach doctors normally take, and flexible enough to allow them to enter data as they pleased. And, unlike the rather strict v1, v2 came equipped with snazzy animations and nifty features that (boring, old) v1 lacked. You can see a few of these updates in the hypothetical patient preview below.
We recognized that the new version required some extensive changes to the underlying software to ensure it remained efficient – many of which were untested. Nevertheless, TimmyCare v2 made its first appearance during the Cincinnati Med School Public Health Trip in April. It entered confidently and seemed to work well, until it faced some heavy load. The new version began crashing on certain machines, it was losing data on certain records, it was rendering incorrectly on others, and most importantly was annoying the doctors.
Two things you should not do (in software development):
1. Deploy untested code
2. Fix that code with more untested code on the spot
The 1st rule was broken due to time constraints and under-consumption of caffeine. The 2nd, however, was unforgivable. Fixing one thing while breaking another does not keep you on even ground.
Nail in the Coffin
Eventually v2 stabilized. Doctors were seeing the value – they were finding it intuitive, useful, even co-operative. They were writing vitals, notes, assessment plans, some Haikus. But the celebration was short-lived when we discovered another unsightly flaw related to gathering data for pediatric care. In order to diagnose proper physical development in children, we needed growth charts and they were only on…Paper! So, Paper struck again, and this one was deep.
That following night, while the rest of the world slept, and more rainforests fell, diligently, TimmyCare v2 plodded along. On the dawn of the last brigade night, it delivered. Growth charts? No, dear readers, not growth charts. It delivered the ability to plot any vital or lab result for a patient. Doctors could see summaries of vitals in a single click. And that finally tipped the scales in our favor!
You see, Paper got complacent. It reigned supreme for thousands of years, forging alliances with graphite and ink, smugly dominating the world. Though the new development to TimmyCare v2 was no surprise attack, Paper still didn’t see it coming.
The next stage, naturally, is world domination over Paper at least. TimmyCare is slated to run trials in Timmy’s other clinic sites (most recently we expanded the EMR to Timmy’s clinics in the DR) to see if it can increase clinic efficiency and quality. However, there is much more to this project than saving rainforests, and charting patient vitals. That riveting discussion, however, is left for Part 3 – so stay tuned!