Control:IQ vs 670G – Which is better? The pivotal trial results are in.

With news that the Tandem T:Slim X2 Control:IQ pivotal study results have been published, I thought it would be worth comparing the data to that from the 670G. 

There are some subtle differences in the way the studies were conducted, and therefore, the data that’s available for comparison.

Put briefly, the results suggest that Control:IQ produces very similar outcomes to the 670G.

Tandem with Control:IQ

First up, a few of the takeaways from the 26-week study for Control:IQ were:

  • 71% (+/- 12%) Time in Range (70-180mg/dl) for the study compared to 59% (+/-14%) for the control group
  • Average Hba1C after closed loop use was 7.06%, which was a reduction of 0.33 percentage points (control adjusted)
  • Average glucose level was 156mg/dl (+/-19)
  • The system was in use for 90% of the study period

The breakdown and images from the report can be seen below:

What stands out about the pre-study data is that the mean Hba1C is substantially lower than the US median, suggesting that those participating in the trial were more actively engaged in their diabetes management before getting involved. The trial team recognised this in the results paper.


What stands out particularly is the overnight Time in Range that is attributed to the more aggressive overnight target that Control:IQ is able to deploy. 

The second image shows a clear reduction in time in hypoglycaemia, which was an important secondary outcome.

So the results aren’t bad. But how do they compare to those from the Medtronic 670G Pivotal Trial?

Medtronic 670G

You have to dig back through the archives a bit to find data about the 670 trial as it’s been that long since the first APS came onto the market. Fortunately, diaTribe helpfully summarised them here.

The key points were for this three month study were:

  • A 0.5% reduction in A1c, bringing patients from a low initial A1c of 7.4% to 6.9%;

  • Time-in-range (71-180 mg/dl) improved from 67% during the two-week run-in to 72% during the study
  • A 44% reduction in time spent with low blood glucose (under 70 mg/dl);

  • A 40% decline in time spent in dangerous hypoglycemia (under 50 mg/dl);

  • An 11% decline in time spent over 180 mg/dl and an 8% improvement in time-in-range (71-180 mg/dl).

  • The system was in use for 87% of the study period
670G Adult daily mean time in range.
670G Adult daily mean time in range.

It’s worth noting that there was no control group in this study, and that the study period was three months instead of six months. 

What do we make of this?

There are a few things that stand out. Probably the biggest is the similarity in outcomes of the two systems.

Both resulted in similar average Hba1C reductions. Both had similar time in range outcomes. The T:Slim seems to have managed better on reducing time spent hypo and in severe hypo.

Another item of note is that the run in period for each of the two systems appears to be different, in that someone going on to the Medtronic system has a two week period in which the system learns about them, whereas Control:IQ had a two to eight week baselining period where it appears that settings were set up for the pump, and then a period of training on the system itself.

The results seem to suggest that Control:IQ being dropped onto the underlying pump settings worked very effectively, however, given the population that was involved in the study; 70% of the patients were using a continuous glucose monitor and 79% were using an insulin pump at the time of enrollment; it’s likely that they were a) engaged and b) starting from a reasonably well set-up point.

It would be interesting to see how intensively those not using a pump prior to enrollment were coached to get their settings correct in the baseline period.  

And the takeaway is?

Very simply, Control:IQ is no better, but importantly, no worse than the 670G (or Diabeloop’s DBGL1) when it comes to clinical outcomes. So don’t expect to change systems and get substantially different results. 

The key difference? A nicer package in the T:Slim X2, Tandem’s legendary usability and CGM that doesn’t require calibration and constant attention. And an update process that allows for you not to have to change pump system as future iterations come out.

On that basis alone, I’d expect more people to choose this over the 670G.



  1. Thanks for the interesting comparison! I do not use either of the systems, but I know people who use them. In my personal experience, people do not like the 670G or its sensors. I have not seen any data on the percentage of real-world users who continue to use the sensors after the initial training. On the other hand, people I know who have switched to Tandem from Medtronic have a learning curve, but they are very happy with the pump and with Dexcom sensors. They are seeing much better time in range than with the 670G.

    Do you know if there are any plans to study the effects of 670G vs Tandem on larger, more typical populations? I have read that the Medtronic/UHC deal included performance guarantees, do you know anything about how that data will be collected and made available?

    Keep up the great work!

    • It’s an interesting one Mike. Medtronic have published a lot of data based on 120,000 users talking about Real World use, showing 85% plus use of Auto mode, TIR in the 70s and citing good satisfaction.

      The changes they made to the transmitter around 10 months ago to remove excessive safety checking seemed to solve a lot of the CGM issues that existed early in its life, and make people less likely to give it up.

      I’m not sure about plans for real world comparisons though. It would be something that would be unlikely to be funded by a manufacturer.

    • Average age was 33, which ranged from 14-71. In each side of the trial, there were around 70% over 18 and 30% under 18.

  2. I’ve had my daughter on both systems – 670G and the currently available Tandem pump with Dexcom G6. We moved away from the 670G because the targets were preset and could not be manually changed so her a1c went up on the Medtronic system. It sounds like the Control IQ will be customizable in terms of the settings and targets?

  3. I purchased the 670g in September of 2016. It was quite possibly the WORST diabetes decision ever made. I had never had problems with Medtonic pumps and wanted the integrated CGM. I did have a chance to wear the older Dexcom and the 670g CGM side by side when I first bought the 670g and found very little differences in their accuracy.
    The problem for me with the 670g was that it could never keep up with my changing insulin sensitivity. The pump was constantly freaking out. Every time my blood sugar went up more than the pump thought it should it would go into this “calibration error, wait to calibrate, update” mode that would have me checking my blood sugar constantly. I would go through a box of 100 test strips in less than a week. Medtronic came out with a protocol that was supposed to resolve calibration issues, but it was a voodoo dance at best. The calibration errors led to sensor failures, and the resulting HOURS that I would spend on the phone with tech support were a complete waste of time, humiliating, actually because they would find any excuse they could to tell me that it was my fault. They told me that I was calibrating at the wrong time, (I calibrated when the pump asked me to) They told me that I was calibrating too often (again, I was calibrating when the pump asked me to, because the pump (or at least “auto mode” doesn’t work when you don’t calibrate. They told me I couldn’t use alkaline batteries, that I should be using lithium batteries, then they told me I couldn’t use lithium batteries and that I should be using alkaline batteries, they told me I couldn’t use generic batteries, they told me that my 170 blood sugar was too high, they even told me that I was having problems because I am overweight (which by the way, I gained 15 pounds on the 670g – but to be fair, I was 15 pounds overweight when I started the study). On that same rant, customer service started asking about a year ago what my weight was every time I called in for technical assistance, and informed me that they were “required by the FDA” to ask. They told me that I was using the wrong kind of soap, the wrong kind of tape, that I was wearing the sensor incorrectly, then they told me that I didn’t call soon enough after I was having the issue, and on, and on. Diabetes takes up a LOT of time, but the 670g and the insanely rude and malicious tech support at Medtronic took that to a level of abuse that I am not even coming close to describing.

    My husband and I stopped sleeping in the same room, I couldn’t stay at hotels (at three a.m., the alerts are very LOUD, I actually had a complaint called in to the front desk) , I couldn’t go to movies without taking the pump completely out of auto mode (which causes the sensor to fail more that 50% of the time) forget weddings, forget church, work meetings, my co-workers were constantly complaining, and my family had tuned out to the point that even if I did have a legitimate alert, they wouldn’t have recognized it because they were numb.
    One day about two months ago, I calibrated, was in range, so I decided to skip breakfast. My blood sugar seemed to be stable and I didn’t want to deal with it any more. Lunch came and I was still in range and it was nice, so I skipped lunch. No calibration errors, no alerts, just quiet and a 112-ish mg/dl flatline that was so pretty I decided not to eat dinner, and then, wow, I slept through the night. (First time in nearly three years) So I went on, and for four days I didn’t eat. Four nights of blissful, uninterrupted sleep. I hacked the 670g. But sadly, not a long term solution. So, I gave up the Enlite sensor, started eating again and started on the Dexcom 6.
    I have one word to describe the Dexcom 6
    OK more words-
    Insanely accurate. As far as CGM goes, this is IT.
    NO fingersticks. NO FINGERSTICKS!!! I also have absolute confidence that the number on my screen is correct, and I really think that this part of the loop is pretty much nailed. (I DO NOT like the 10 day time window b/c I really cant carve out a schedule for it, It’s on Monday, then Wednesday, then Friday, its frustrating, but really, I am NOT complaining. Dexcom is the Maytag of CGMs and Sensors and their customer service is fantastic) So now, I am trying to decide what I want to do next. I know someone that is in the Control IQ study who was previously on MDI. My friend had been pretty resistant to insulin pumps, CGMs, and really ANY advanced diabetes technology, and somehow managed to get in on this trial (I know, some people, right) Watching my friend, Control IQ does not have ANY of the calibration error insanity that goes on with Medtronic. There is no “time spent in auto mode” drama. The pump seems to stay in Control IQ mode most, if not all, of the time, which allows more time to be spent working on “time in range” which is waaaaay more important. My friend’s control still isn’t the greatest, but Control IQ did increase time in range, it did lower A1c’s– nearly 3 FULL POINTS, but my friend was a “whatever” diabetic at best, and even now sports a barely acceptable A1c) but Control IQ has allowed my friend to see a pretty nice improvement in diabetes management with not much extra effort. Lucky jerk.

    I don’t think that Control IQ will let you pick your b/s target, I don’t think you can set temp basal rates, but I am not sure. I don’t think that you can do some of the really helpful advanced pump hacks that the Medtronic 640 (not the 670g) would allow you to do with the basal rates. That is why I have not committed to another four years of closed loop, because I am not nearly as confident in the TSlim and I am hoping an even better pump will be available soon. I don’t like the TSlim has a built in rechargeable battery option. I spend a lot of time outdoors and hot swap-able AA batteries are easy to manage. I can buy them anywhere, I can carry three to keep me up and running for a month long back country hike. With blackouts becoming more of a reality, I like having batteries that I can store in a drawer to keep me up and running. A Sams club bulk pack can keep me running for a looooong time. I don’t like TSlims infusion set options. Again, I spend a lot of time hiking and camping, and all the extra trash that I have to carry with the infusion sets is horrible (Have to mention here that Dexcom 6 near watermelon sized disposable applicator is HORRIBLE) The pump bladder is dubious, and there is a LOT of wasted insulin if you lose an infusion set. I am stilll looking for a better pump option.

    To sum it up, I really feel like the last three years of my life have been extremely harmful to my health. Since I have been diagnosed the only negative effect that I have experienced has been a gradual weight gain, (I am getting old) but since I have started on 670g I have started to see some pretty scary issues with some of the major players in diabetes complications. This could be a coincidence. I don’t want to chance it. I am jumping off the 670g closed loop, and am looking for a better answer.

  4. Thanks for taking the time to compare the two pumps and the results of the studies.

    I just have one qualm, which is that you mention “Tandem’s legendary usability”. My experience of using the T:slim X2 after using the Animas Vibe for many years is that, for experienced pumpers, the usability factor of the T:slim is absolutely trash.

    The user interface is awful, here are some examples: It’s way too easy to confuse entering BG and carbs for bolusing. You can dismiss the maximum bolus exceeded alert without thinking since it’s the exact same sequence of screen taps as normal boluses. There are way too many alarms (incomplete bolus???) and the lowest volume setting is too loud. Having to wait for so many confirmation screens to pop up and go away. Having to change a single I:C ratio for every change in basal rate throughout the day. And the ‘three wrong taps turns the screen off’ both turns the screen off while you are using it and also doesn’t even prevent pocket touches, so what’s the point? I could go on much longer, but you get the idea.

    It really feels like the UI was what they were using for their working prototype and they never bothered to actually go in and find out how to make it more efficient and less prone to user mistakes.

    Sooo yeah having ‘upgraded’ to the T:slim over the Animas Vibe (RIP Animas), I’m really not a fan of their software design decisions. It’s certainly a nice looking pump, but the beauty is sort of only surface deep.

    • All very fair feedback. Having not used the T;Slim pumps, I can’t comment. In general though, the ratio of good to bad comments that I hear from users in the UK is about 5:1 (which is, of course, likely to be biased somewhere).

  5. I started using the T:slim X2 late last year after being a Medtronic user for over a decade. The last pump I had was a 630G, and I hated it. I sent it back before the 3 months deadline my provincial coverage (Ontario) gave me. I also hated their CGM which made me crazy! I used an older pump I had kicking around and decided to try the Tandem pump and Dexcom G5. It was a life-changing decision. I’ve since upgraded to Basal IQ and G6 set up a few weeks ago. Although it means no more extending sensors (not sure it’s a good idea with the G6 and Basal IQ anyway) and that’s expensive since I have limited insurance coverage for sensors and transmitters. I’m experiencing more time in range and less time in hypoglycemic episodes, especially at night. I know that the pump has some limitations, but is it better than the Medtronic system I used? Absolutely! Plus, my experience with Medtronic’s support was similar to the previous post that talked about putting the blame on the user for the shortcomings of their systems. I’ve had nothing but great support from Tandem and Dexcom. I can’t wait to use Control IQ. Plus, I’m an actor and having a pump that’s smaller and lighter makes using it while working onstage or onset a lot easier.

    • Hi John, Thanks for sharing your experience. I am a 20 year Medtronic user who is FINALLY eligible for a new pump and moving to Tandem X2. For the last 16 months I’ve been running DIY Loop with an old Medtronic. Most of that time I was self-funding the Dexcom G6, and cut my costs by extending them. What is it about switching to Tandem that stops a person from extending the sensors? I agree it is not best, but I find that with up front calibrations and occasional checks, they are usually very accurate for another 8-10 days. I use the “pop-out” method demonstrated on numerous YouTube videos.

Leave a Reply to admin Cancel reply

Your email address will not be published.