Skip to main content

You Call That an Artificial Pancreas?

Updated: 8/14/21 9:00 amPublished: 1/21/14

By Gary Scheiner MS, CDE

By now many of you have heard and commented on the latest product approved by the FDA: Medtronic’s MiniMed 530G insulin pump with Enlite CGM. As a “threshold suspend” system, the MiniMed 530G technically falls under the FDA’s new “Artificial Pancreas Device System” category. But before you go running around yelling “We’re cured! We’re cured!” and chucking your diabetes supplies into a recycling bin, there are a few things that everyone should know.

The 530G is NOT an artificial pancreas as many formally think of the term; it is a step toward the artificial pancreas. It doesn’t deliver insulin automatically. It doesn’t eliminate the need for fingerstick blood sugar measurements, though far fewer are required for many patients than in the old days! And it doesn’t alleviate the ongoing decision-making required to manage diabetes on a daily basis. But it does represent an important first step toward the development of an artificial pancreas. Why? Because this is the first time the FDA has approved a device that makes a decision, any decision, about insulin delivery without the user’s involvement. The 530G will temporarily suspend delivery of basal insulin for up to two hours any time the Enlite glucose sensor detects a blood sugar that is below a user’s set low threshold (60-90 mg/dl). That’s important progress – many don’t realize it, but even traditional blood glucose monitors are not technically approved for insulin dosing decisions.

So where do we go from here? Let’s review the steps toward an artificial pancreas, examine the pros and cons of each stage, and see what we can do until it hits the market.


This is what the MiniMed 530G currently offers. Research has shown that the two-hour suspension of basal insulin does not tend to lead to undesirably high blood sugars or ketoacidosis. It simply takes a patient from a hypoglycemic state (who is not responding to alarms) and suspends basal insulin for two hours – over the next couple of hours blood sugar rises out of hypoglycemia (more details on the pivotal study are explained here). This can be a source of comfort for loved ones (particularly at night) and a potential life-saver for anyone who experiences hypoglycemia while unconscious or is unable to respond to the alert. But there are a few issues.

The MiniMed 530G with Enlite CGM relies on a sensor to measure glucose levels accurately in a relatively low blood sugar range. That’s challenging for any sensor, since a 15% error is magnified at low blood glucose levels. Although Medtronic’s new Enlite sensors are more accurate than their first-generation Sof-Sensors, they can still vary considerably from fingerstick and lab values (according to Enlite’s product label, the sensor varies by about 14-15% from lab values, compared to about 20% with the Sof-Sensor). Despite the improvement in accuracy, there is still the chance that the Enlite will miss some low episodes and generate false alarms when the blood sugar isn’t actually low. The threshold suspend feature is intended for use when someone is unresponsive (e.g., sleeping), and it is important to note that suspending basal insulin is NOT the best way to deal with 99% of low blood sugars. Eating (or drinking) rapid-acting carbs will raise the blood sugar MUCH faster and more predictably (5-15 minutes with carbs vs. at least 30 minutes with insulin suspension). It’s good to see that Medtronic acknowledges this in the product’s labeling, as the pump is not indicated for reducing hypoglycemia.

So what can we do today that mimics what the MiniMed 530G does? Simple. Use a CGM with an alarm that alerts you to take action. If the CGM’s standard alarms don’t cause you to take notice or wake you while you’re asleep, figure out a better system. Use the vibrate feature, put the receiver in an empty glass on a bedside table, place it near a baby monitor/speaker, keep it with a partner who is more likely to hear/feel it, or choose a CGM with a more robust set of alerts (in the US, the other option is the Dexcom G4 Platinum). Animal lovers can also look to obtain a Diabetes Alert Dog that lets its owner know when blood glucose levels are dropping. And of course, there is the old fingerstick fallback: More frequent blood glucose monitoring gives you a better chance of catching lows before they hinder your ability to self-treat them.


Products are already in the R&D pipeline that will slow or curtail a pump’s basal insulin delivery when blood glucose levels are APPROACHING a low threshold (better than the aforementioned MiniMed 530G, which suspends insulin delivery when the low threshold is reached). The obvious benefit of predictive suspension is that it has the potential to greatly reduce the frequency, severity, and duration of hypoglycemic episodes. Of course, its success hinges on using a glucose sensor that is accurate and dependable, and an algorithm that knows when/how much to adjust basal insulin delivery when a low is approaching. Medtronic plans to launch a predictive low glucose suspend system outside the US in early 2014.

Until that’s available, there are several ways to prevent lows using currently available technology. All CGM systems feature a “Fall Rate” alert that lets the user know when the blood glucose is dropping quickly. Medtronic CGMs also feature a “Predictive” alert to let users know when a low is likely to occur soon. Responding to these alerts with food (when blood glucose is approaching hypoglycemia) is an effective way to prevent the low from happening in the first place.

Even those who don’t have CGM can prevent most (if not all) low blood sugars through common-sense approaches: making sure basal and bolus insulin doses are set properly; accurate carb counting (so as not to over-bolus); adjusting food/insulin for physical activity; taking precautions when consuming alcohol; delaying mealtime insulin with slowly-digesting foods; and performing an adequate number of fingerstick blood sugar measurements throughout the day. It can also be extremely helpful to look at one’s own data to determine the times and sources of lows so that preventive measures can be taken.


A pump/CGM combination that will automatically keep blood sugars within an acceptable target range overnight is also in the research pipeline. This will be accomplished through the infusion of minute boluses when blood sugars are rising, and reductions in basal insulin when blood sugars are falling. For this type of system to function properly, the user must not eat or bolus for several hours leading up to bedtime, nor should they eat or bolus during the night. When working properly, auto-pilot has the potential to make nighttime highs and lows a thing of the past. (It’s not clear if a treat-to-range system would come before or after nighttime auto-pilot, or if the two stages would be combined into a single product. The technology for both systems already exists and is in development and testing.)

For now, just about anyone with diabetes can keep their blood sugar reasonably steady through the night by making sure the doses of BASAL insulin (from a pump or injections) are set correctly. Basal insulin’s job is to offset the glucose secreted by the liver, so it should keep blood sugar from rising or falling significantly while we sleep. Of course, there will be times when the standard basal insulin doses will need adjustment – we have to be prepared to make temporary changes to basal insulin levels after high-fat meals, extended exercise, alcohol consumption, use of steroid medications, and during times of illness.


This is what I refer to as the “quantum leap.” The first three stages are well within our grasp. Heck, even at the slow rate with which the FDA approves new products, we’ll all probably see those first three stages within our lifetimes. But the really BIG hurdle – automatically controlling blood sugar following meals, during exercise, with hormonal changes and other factors such as stress and illness – is going to take a lot more time and work. In all likelihood, it will require the use of multiple hormones (perhaps adding glucagon and/or amylin to the mix), and a mechanism for making the insulin start/peak/dissipate much faster than it does now. The CGM sensors will have to be highly accurate and reliable, and the algorithm for interpreting the sensor data will have to work in an intelligent, proactive fashion.

But when it works, how sweet will that be? Blood sugars will approach those of the non-pancreatically-challenged, before and after meals. During and after exercise. Even during tax season. Not having to THINK about how every little thing affects our blood sugar will leave us enough extra RAM in our brains for more important things, like sports statistics, anniversary dates, and our kids’ shoe sizes.

Until that happens, we owe it to ourselves to do the best job possible of being our own artificial pancreas. Someday the “real deal” will become reality. And when it does, we want to be in the best shape possible so that we can really enjoy it.

And since we won’t have to worry about bizarre middle-of-the-night lows, drinks are on me.


Gary Scheiner is Owner and Clinical Director of Integrated Diabetes Services, a private consulting practice for people with diabetes who utilize intensive insulin therapy. Gary is the 2014 AADE Diabetes Educator of the Year. He has written several books on diabetes self-management, including “Think Like A Pancreas” and “Until There’s A Cure”. Gary and his team offer consultations worldwide via phone and internet for those looking to gain better control of their diabetes and enhance their self-management skills. Gary has had type 1 diabetes for 28 years and has worn and trained on every make and model of continuous glucose monitor and insulin pump. He can be reached at [email protected], or (610) 642-6055.

What do you think?