Articles & Experts

Diabetes Patient Management: A Simpler, Better, and Safer Way to Dose Insulin

Thursday, November 14, 2019

by David Kliff of the Diabetic Investor


Over the years there have been several notable advancements when it comes to insulin therapy. The introduction of analogue insulin’s, “smart” insulin pumps and continuous glucose monitoring systems (CGMS) are just some that quickly come to mind. Coming in the very near future will be insulin pens which talk to a patient’s smartphone and the most promising advancement since the discovery of insulin, insulin dosing algorithms. Algorithms that not only help the patient more effectively dose their insulin, but algorithms which learn allowing them to get even better over time.

Let’s face facts, for the majority of patients dosing is basically guess work, especially those who follow multiple daily injection (MDI) therapy. Typically their physician assigns a set amount of basal insulin to be taken each night before bed or each morning before the patient starts their day. The basal insulin is then supplemented with short-acting insulin taken when the patient has a meal or snack, otherwise known as the bolus. In theory the patient would calculate the amount of insulin to take with a meal based on a multiplicity of factors, such as;


Carb intake

Target Glucose Range

Current Glucose Level

Time to Action

Duration of Insulin Action

Insulin to Carb Ratio’s

Insulin on Board


This is not a complete list but it does explain why so many patients use the WAG (wild ass guess) method dosing their insulin. Yes there are several apps available which can perform the calculation for the patient. However, the apps require the patient enter all the above-mentioned factors so it can make an accurate calculation. Quite frankly the majority of patients either don’t know what to enter or don’t take the time to enter it.


For these reasons and to keep insulin dosing simple, many physicians dumb down insulin dosing; telling the patient to dose a set amount with each meal. This may not be the most effective method but as we all know while insulin is a life sustaining drug it is also can be a lethal drug when dosed incorrectly.


Thanks to advancements in diabetes device technology we are quickly approaching the day when all the guess work will be taken out of insulin dosing. A day when the system will do all the heavy lifting for the patient who will have just one job, injecting the recommended dose. The cornerstones of these soon to be available systems are an insulin pen which communicates with the patients smartphone, a CGM which also communicates and an insulin dosing algorithm which takes all the data generated by these devices and calculates the amount of insulin a patient should dose.


Although no one company has begun marketing a “Tyler,” our term for these connected pen/app/CGM systems, several are under development. Lilly, Novo Nordisk and Sanofi are developing their own version of a Tyler. Systems which will communicate with the two most popular CGM systems the G6 from Dexcom and the Libre from Abbott. Thankfully the CGM component of Tyler are becoming more patient friendly, easier to use and most importantly cheaper.


One company, Companion Medical already has an FDA approved connected pen called the inPen which communicates with a smartphone app which also can collect glucose data from a Dexcom CGM. The inPen does not yet come with an insulin dosing algorithm but does offer the patient and their physician a plethora of data which can then be used as a learning tool. Given that Dexcom also owns TypeZero an insulin algorithm, it would not come as shock if one day this algorithm becomes part of the inPen app.


No matter which system a patient chooses, one thing is certain: when used as designed all the guess work will be taken out of insulin dosing. Even without these new sophisticated insulin dosing algorithms, several studies have noted that patients using a CGM achieve better outcomes no matter how the insulin is delivered, insulin pen or insulin pump. It stands to given that when insulin dosing algorithms are added these results will get even better.


Quite frankly with these advancements we are not going out on a limb when we state that in the very near future it will be easier than ever to be an insulin using patient. Thanks to these learning algorithms all the guess work will be gone. Providing a simpler, better, and safer way to dose insulin.