CGM: What Is It Good For???
by Gary Scheiner MS, CDE
Every time I hear that “What is it good for?” phrase, I think of the Motown hit from the late 60s rallying against the war in Vietnam (and war in general).
Unlike war, there is lot that CGM is good for. Users benefit from high/low alerts and the ability to react to the direction of glucose levels rather than just moment-in-time readings. But what about healthcare professionals and their use of CGM data? Unfortunately, very few of us have the time or expertise to break down the data and come up with insightful and clinically relevant recommendations. That’s one of the major reasons I’m working on a book (to be published by the ADA in 2015) on sensor data interpretation.
The fact is, sensor data charts and trend graphs can be downright scary to look at. There is SO MUCH STUFF staring you in the face that it is hard to know where to begin. My recommendation is to avoid going into it with an open mind. That’s right. Have an agenda – a set of specific objectives -- before looking at a single data point. That way, you’ll be able to glean value from the information without suffering a nervous breakdown.
Here are some of the items I like to focus on:
1. Is the basal insulin dose correct? Any time a patient goes more than 4 hours without eating or taking rapid-acting insulin, the basal insulin should be holding their glucose steady.
2. Are the meal insulin doses correct? Seeing where the BG is 3-4 hours post-meal will tell you if meals are being covered properly.
3. Are the correction dosage formulas correct? Sensors will show how the BG responds to correction doses – does the glucose come down too much or not enough?
4. How long does the insulin last? This is an important concept for anyone who takes multiple doses of insulin throughout the day in order to avoid “stacking” and over-correction for highs. The sensor trend graphs show very clearly how long a bolus of insulin takes to finish working.
5. How much post-meal spiking is taking place? It is difficult to track this using fingersticks, since we never really know when the spike will take place. Because CGMs update every 5 minutes, the post-meal “peaks” become easily detectible on trend graphs.
6. Is it time to escalate therapy? A view of the “big picture” via multiple-day trend graphs and summary statistics let us see if the current therapy is doing the job.
7. Are asymptomatic lows occurring, and are lows being treated properly? CGMs are very good at catching even mild lows (including during the night) that may produce rebound highs. You can also determine if the patient is over- or under-treating their lows based on the glucose levels that follow over the next several hours.
8. What affect is exercise having? Most CGM download programs allow you to look at specific days of the week so that peri- and post-exercise patterns can be detected.
9. Are there periodic events influencing glucose levels? The effects of episodic occurrences such as menstrual cycles, infusion set changes, work/school schedules and weekly activities can be seen on intermediate-term trend graph reports.
10. How are we progressing? Statistical reports from CGM are much more valid than reports from BG meters, since data is being collected just about all the time. Averages, standard deviations, # of high/low events, and percent of time spent above/below/within one’s target range provide excellent indices of the quality of one’s control.