Imagine your drinking water supply was safe some days, and highly contaminated on others. The only way to know—and respond by either treating it or not—would be to test it. And if the testing supplies weren’t widely available and very affordable, you wouldn’t be surprised when people frequently got sick.
Unfortunately this is the case of diabetes testing supplies. Glucose meters, testing strips, finger sticks and continuous monitoring devices are among the overpriced essential technologies for people living with diabetes.
Insulin is the vital hormone that a diabetic’s pancreas either makes too little of (type 2 diabetes) or fails to make at all (type 1 diabetes). Unlike many other hormone medicines such as cortisol or levothyroxine, insulin has a fast action time. A person's blood glucose levels are ideally kept within about 72-140 mg/dL (4.0-7.8 mmol/L). Too much insulin and you could quickly slip into a coma, too little and you could slowly slip into a coma. The amount of insulin a person needs varies enormously depending on body mass, activity, stress, and nutrition, and the amount needed by the same individual can vary widely day to day, even hour to hour. For this reason, insulin-dependent diabetics administer anywhere from 4-10+ injections per day to maintain optimal blood sugar levels, and test their blood sugar at least 2 or 3 times. Insulin aggressively corrects high blood sugar. And you can’t correct what you can’t measure.
A 2012 study found that testing supplies account for a quarter of the pharmacy costs for a person with diabetes. Four companies account for half the market share: Roche, Johnson & Johnson, Ascensia, and Abbott. In 2017, the global blood glucose monitoring market was valued at over $10 billion, and it is projected to continue growing. A person with good insurance could face a $15 copay for each item, but someone without it would pay a minimum of $100 a month, if they bought their supplies from CVS:
The structure of a test strip is quite simple: a chamber for a small blood sample, an enzyme that turns glucose concentration into free ions, and conductive materials that measure the resulting voltage. A screen displays the result. Test strips are printed in enormous rolls and each strip costs less than 10 cents to make. Test strips are proprietary: they can only be used with a specific meter produced by the same brand. And most are coded, with expiration dates beyond which they are no longer recognized by the machine.
Since about 2010 the market has been rapidly expanding with new products, including continuous monitoring devices - wearable meters with a small subcutaneous catheter that takes tiny blood samples and contains chemicals and electrodes. These are extremely useful from a disease-management perspective, because they deliver a blood glucose reading to an external device or a smartphone every five or so minutes. Patients can make smaller and more frequent insulin adjustments and obtain better glucose control. But these devices are expensive. Dexcom, the market leader, charges $1027 for a three-month supply: $327 for a transmitter, $700 for sensors. And an added one-off cost of $365 for the receiver that displays the results. We need an alternative. Diabetes should not be an opportunity for a shareholder to buy a second home.
In March 2020 I produced a prototype for an open-source glucose meter. The two-chambered test strip is constructed using thin aluminum sheet, plastic, indium tin oxide, and adhesive, with copper contacts. The working electrode chamber is saturated with glucose oxidase. PVCA (an old credit card) separates the working electrode chamber from the reference electrode chamber.
The electronics for the Open Glucometer are a Raspberry Pi running a C program that measures differential voltages across the two 1.5 μl blood samples using an analog-to-digital converter 64 times per second. It displays voltage over time on an OLED screen.
Next, I’ll be gathering more data on voltage difference over time compared to externally verified glucose mg/dl, formulating this for addition to the software and OLED display, and experimenting with chamber washing techniques (exploring re-usable test strip design). The cost to build this prototype was ~$50; my goal is to reduce this to closer to $10, with each test strip lasting two weeks and costing less than 10 cents. Open source glucose oxidase could be produced using the same bioreactor currently being used to produce insulin.