Step up your diabetes game! Learn the ins and outs of continuous glucose monitors (CGM) including key features, what and how to order, interpreting CGM data (time in range, time above range, etc.), and how to make adjustments to a patient’s insulin regimen, aka “insulin pattern matching”. We discuss common scenarios including overnight hypoglycemia, exercise-induced hypoglycemia, post-prandial hyperglycemia, what to do when the patient runs high all the time, and how to incorporate GLP1 agonists and SGLT2 inhibitors into a patient’s regimen. Returning guest and Paul’s favorite frenemy, Dr. Jeff Colburn (VCU Health) returns!
Claim CME for this episode at curbsiders.vcuhealth.org!
Patreon | Episodes | Subscribe | Spotify | YouTube | Newsletter | Contact | Swag! | CME


Recall that an average blood sugar of 126 mg/dL is equivalent to an A1C of 6% and the A1C increases by 1% for every ~30 mg/dL increase in blood glucose (i.e. 154 mg/dL = 7%, 183 mg/dL = 8% and so forth) —ADA eAG/A1C Conversion Calculator accessed 17 March 2024.
Since April 2023, CMS says any patient with 1) diabetes taking insulin at least once daily or 2) problematic hypoglycemia is eligible for a continuous glucose monitor (CGM). A follow-up with the treating clinician every six months is required to “document adherence to their CGM regimen and diabetes treatment plan” (CMS.gov L33822 accessed 17 Mar 2024).
Studies suggest that patients wearing a CGM see an improvement in A1C values even without any changes in therapy (Vigersky, 2012; Jancev, 2024). Dr. Colburn hypothesizes this occurs due to immediate feedback on eating behaviors.
CGM pro model: A sensor is placed in the clinic and worn for 14 days. The patient cannot see the readings to prevent them from acting on the blood glucose readings.
CGMs for personal use consist of a sensor (to detect glucose values), a transmitter (to send blood glucose data), and a receiver (to display the data). Dr. Colburn notes that most people use their smartphone as a receiver and in newer models, the sensor and transmitter are combined. Dr. Colburn instructs patients to apply a sensor to the back of the upper arm while holding their arm parallel to the floor with the elbow flexed at 90 degrees.
The two common models of CGMs available in the United States, Dexcom and Freestyle, have sensors that can be worn for 10 days or 14 days respectively. Thus, prescribers should write for 2 or 3 sensors per month. The Dexcom G6 requires a transmitter, which has a 3 month battery life, but the Dexcom G7 has a transmitter built into the sensor “all-in-one“. For Dexcom patients should download the Dexcom app and sign up for the patient portal, Clarity. For Freestyle patients only need to download the app. Dr. Colburn and The Curbsiders do not have a preference or any personal interest in either brand.
Smartphones can upload blood glucose data to the cloud which can be monitored by the treating clinician, caregivers, or family members. Dr. Colburn notes that most practices set up a group account rather than individual accounts for each clinician to allow for cross-coverage.
Below are standardized CGM metrics (Szmuilowicz, 2022)
Dr. Colburn encourages patients to keep a record of eating behaviors and activity patterns to help make sense of glucose trends.
While the A1C gives a three-month look back, the TIR is more “real-time” since it provides a 10 to 14 day look back based on the most recent sensor data. Further, the A1C gives an average glucose, but it might miss wide variations in blood glucose readings captured by a CGM. Finally, CGM data is not affected by conditions like anemia, which can falsely raise or lower the A1C.
On the contrary, CGM readings are susceptible to interference from certain substances. For instance, the earlier models of Dexcom were subject to interference from acetaminophen and the Dexcom G6 may have falsely elevated readings in patients taking hydroxyurea (Heinemann, 2022). The Freestyle device appears to be prone to interference by ascorbic acid (Heinemann, 2022).
Dr. Colburn points out that CGMs have an alarm for lows and highs that can be set to customized values (e.g. under 70 mg/dL or above 250 mg/dL). He cautions that CGMs measure glucose in interstitial fluid (NOT in the blood), which is about 15 minutes behind the actual blood glucose value with current models (Zaharieva, 2019). This is of particular concern if a patient has blood sugar that is rapidly rising or falling and could cause a delay in the treatment of hypoglycemia.
When evaluating blood glucose patterns, first, address any hypoglycemia and then, address hyperglycemia.
Broadly, metformin*, GLP1 agonists, SGLT2 inhibitors, and DPP4 inhibitors can lower both fasting and postprandial glucose (Szmuilowicz, 2022). These agents should not cause significant hypoglycemia in patients unless they are also taking insulin or an insulin secretagogue (i.e. sulfonylurea).
*Metformin and basal insulin mainly target fasting blood glucose.
Below are some common blood glucose patterns and some potential approaches to treatment. Remember that there is often more than one right answer in diabetes care.
Nocturnal hypoglycemic events often wake patients due to counterregulatory hormones (glucagon, growth hormone, cortisol, epi/norepinephrine) changes aimed at blood glucose normalization. Dr. Colburn cautions that patients should not, “just let the body fix” their hypoglycemia because recurrent hypoglycemia can be damaging and has been associated with increased mortality, poor cognitive function (in children), and possibly dementia in adults (Amiel, 2021)! If hypoglycemia occurs, follow the rule of 15, take 15 grams of carbs, and repeat the blood glucose in 15 minutes. See Curbsiders episode #397 Insulin, T2DM, Fanny Packs, and Hypoglycemia.
Reduce the dose of basal insulin for patients with nocturnal hypoglycemia. This change might paradoxically improve the A1C because as Dr. Colburn notes, patients can have reduced insulin sensitivity for nearly 24 hours after a hypoglycemic event (Heller, 1991; Lucidi, 2010). Further, patients often eat high-sugar foods to self-treat hypoglycemia. Thus, lowering the basal insulin dose and avoiding hypoglycemia might improve the A1C by maintaining insulin sensitivity and avoiding the need to treat with high-sugar foods.
Omit the evening dose for patients on a sulfonylurea who are experiencing nocturnal hypoglycemia. Patients on a long-acting sulfonylurea might need to reduce the dose or stop it altogether. Dr. Colburn often switches to another class of agent since sulfonylureas do not have added organ benefits outside of surrogate endpoints like the need for retinal photocoagulation or microalbuminuria (UKPDS, 1998; ADVANCE, 2008).
Patients with type 1 diabetes and type 2 diabetes treated with insulin or insulin secretagogues are at risk for exercise-induced hypoglycemia (Younk, 2011; Cockcroft, 2020) due to impaired counterregulatory response to exercise mitigated by diabetes itself and medications (Younk, 2011). Further, insulin sensitivity improves for hours (maybe up to 72 hours) after exercise (Bird, 2017). Rather than discouraging exercise, Dr. Colburn suggested a few strategies to mitigate the risk of hypoglycemia including reducing the basal insulin dose, reducing the preceding prandial insulin dose, or taking a snack with complex carbohydrates, protein, and fat (e.g. half a banana and peanut butter) prior to exercise –expert opinion.
Patients using conventional fingersticks to check blood glucose often check before meals and miss episodes of postprandial hyperglycemia. As CGM use increases, patients and clinicians will likely be more aware of post-meal spikes in blood glucose.
GLP1 agonists (including GIP/GLP1 agents), SGLT2 inhibitors, DPP4 inhibitors, or mealtime insulin can help with postprandial hyperglycemia. As discussed with Dr. Colburn, adding a GLP1 agonist or SGLT2 inhibitor (or both) to basal insulin and/or metformin can prevent the need for multiple daily insulin injections. However, some patients may require mealtime insulin. In those cases, patients can start by taking short-acting insulin with the largest meal of the day rather than injecting insulin at each meal (Samson, 2023), the so-called “basal plus one” approach.
DPP4 inhibitors have a limited role due to weak effects on A1C, cost, and lack of added organ benefit (Dungan, 2023 UpToDate accessed 17 March 2024).
Kashlak Pearl: Dr. Colburn cautions that rapid-acting insulin should not be given between meals to avoid “stacking” doses. Patients should be instructed to take insulin at the prescribed times (e.g. once-daily basal insulin and rapid-acting insulin three times daily with meals). Patients should work with the prescribing clinician to make adjustments based on blood glucose data.
Dr. Colburn finds this to be the least challenging pattern since it’s clear that medication needs to be added or increased. However, don’t forget to address lifestyle modifications!
Kashlak Pearl: Basal insulin has a ceiling effect at about 0.5 units/kg/day so further increases have diminishing returns and lead to “overbasalization”.
Signs of overbasalization and the need to intensify therapy beyond basal insulin include (Cowart, 2020):
Dr. Colburn recommends addressing fasting glucose first and then, addressing mealtime needs. GLP1 agonists and SGLT2 inhibitors have added organ benefits so the choice between them can be dictated by a patient’s comorbid conditions (e.g. obesity, CKD, heart failure, etc.). Remember, while insulin helps lower blood glucose, it lacks added organ benefits in type 2 diabetes and puts the patient at risk for hypoglycemia and weight gain (ACCORD, 2008). That said, Dr. Colburn notes that clinical inertia often delays the initiation of insulin. Remember, you can always back off on insulin as glucose control improves, especially after the addition of adjuvant meds like GLP1 agonists and/or SGLT2 inhibitors.
Listeners will learn how to recognize and respond to patterns observed in blood glucose readings
After listening to this episode listeners will…
Dr. Colburn has no relevant financial disclosures. The Curbsiders report no relevant financial disclosures.
Watto MF, Colburn J, Emi Okamoto, Williams PN. “#433 CGMs, Insulin, and How to Adjust Diabetes Meds to Match Blood Glucose Patterns”. The Curbsiders Internal Medicine Podcast. thecurbsiders.com/category/curbsiders-podcast Final publishing date April 1, 2024.
You can try Freed for free right now by going to freed.ai. And listeners of Curbsiders can use code CURB50 for $50 off their first month.
Download the 2024 Residents & Fellows Survey Report for free at panaceafinancial.com/report.
Writer and Producer: Matthew Watto MD, FACP
Show Notes: Matthew Watto MD, FACP
Cover Art and Infographic: Matthew Watto MD, FACP
Hosts: Matthew Watto MD, FACP; Paul Williams MD, FACP
Reviewer: Emi Okamoto MD
Showrunners: Matthew Watto MD, FACP; Paul Williams MD, FACP
Technical Production: PodPaste
Guest: Jeff Colburn MD
The Curbsiders are partnering with VCU Health Continuing Education to offer continuing education credits for physicians and other healthcare professionals. Visit curbsiders.vcuhealth.org and search for this episode to claim credit.
Got feedback? Suggest a Curbsiders topic. Recommend a guest. Tell us what you think.
We love hearing from you.

Yes, you can now join our exclusive community of core faculty at Kashlak Memorial Hospital along with all the perks:
Notice
We and selected third parties use cookies or similar technologies for technical purposes and, with your consent, for other purposes as specified in the cookie policy. Denying consent may make related features unavailable.
Close this notice to consent.
Comments
Great episode. Keep the clinical pearls and dad jokes/medical puns coming please. As a clinical pharmacist working in an outpatient family medicine clinic, I greatly appreciate when a CGM helps a patient be more compliant and take their diabetes more seriously. It`s a great non-pharmacological intervention when approved and covered. I would like however to point out two facts about CGM`s that were missed in this episode: 1) A1c reductions, as noted in the posted meta-analysis, is only -0.3% more than placebo, which is statistically significant but not clinically significant (would have to be more than 0.5%). 2) Trial patients are usually followed very closely by multidisciplinary teams (often q2-4 weeks) and are often monitoring their glucose over 10 times per day. In the limited observational studies we have, there is a 'wearing off' effect often after the 6 month mark, monitoring becomes less frequent much like many lifestyle interventions. Therefore, it is likely that in the real world, less patients would have the A1c drop as those seen in the studied population. This is important to note as CGM`s are quite costly, even in Canada.