Is Dexcom G7 or G6 better? - Review

After reading several posts about the G7 on this forum with my G6 backup supply shrinking, and with great trepidation, I have finally taken the plunge and decided to simultaneously use both a G7 and a G6 for several weeks, and learn and document the differences.

Both my G6 sensors and my G7 are for distribution to patients with government payors only (IE Medicare devices). The G7 is a Rev 10 LBL 017963 with no white underline, predating the latest approved tandem integrated Rev 001 devices. The G7 sensor was produced on 2023-02-01 and expires on 2024-07-31, so there are about 7 months left before the expiry date—lot 1523048822.

My G6 presoaked 24 hrs., is placed inline center to center 1 ¼” above my belly button, and the G7, no presoak, inline 1 ¾” center to center above the G6. I understand those are not recommended placement sites in the US but are in other parts of the world. I am a side sleeper, both sides, so all recommended sites give me compression lows, and I never get compression lows at this placement. Plus, I don’t have to be a contortionist to apply devices to my stomach easily. I guess that in the US, placement was more about bragging rights and marketing than performance. The MARD rating on the G7 stomach trials was a tiny fraction of a % worse than the G6. However, the MARD rating on the arm was a tiny fraction better on the G7 than on the G6, hence being able to market the G7 device, as a better product. Sort of like a 60MPH speed limit and one car’s cruise control holds the vehicle within the speed limit at 59.996MPH while the other car’s cruise control violates the speed limit holding 60.001 MPH.

Day one BG readings between 100-135 were impressively dead on between the two devices. The difference between the two devices rarely exceeded 1 mg/dL and mostly read the same. Below BG 100, the G6 dropped farther, faster than the G7. When the G7 got to 80, the G6 was reading 62. Greater than BG 130, the trend was reversed, and by the time the G6 was at BG 150, the G7 was reading 190. Fingerstick on both the low end and high end averaged somewhere between the two.

Within less than 24 hours, both devices had stabilized. The G7 on the low side was aggressively reading conservative lows. The G6 was showing BG 81, the G7 was at 74, and the fingerstick was at 87. The G7 algorithm was reading more like the G5 used to rather than the G6. I like the G7 in that it makes you act because when going high, the highs display much faster and keep climbing unless action is taken and the low end does the same. Being MDI acting early helps as the high and low trends and alarms start sooner.

By day five, the G6 and G7 were reading neck and neck (within 7 points and when there was a delta, the G7 generally read higher than the G6 up to 190, which was as high as I wanted to push my BG. On the low end of the scale, G6 and G7 were neck and neck (within 4 points, and when there was a delta the G7 generally read higher than the G6) down to 75 which was as low as I wanted to go as I was outside on an extended walk and don’t feel comfortable exercising outdoors for extended periods of time below 75. I occasionally continued to lose the odd 1-3 dots over 24 hours on the G7, but during that same time, the G6 had dots for those same periods, so I just appeared to have smoothed those dots into data and the graph. I am not crazy about the G6 smoothing, as it gives a false sense of security. Sometimes, the receiver displays smoothed data while I suspect not receiving any data from the transmitter.

Day seven the “Brief Sensor Issue Don’t remove sensor—temporary Issue. Wait up to 3 hours.” I also got the “keep sensor within 20 ft of the receiver.” The Bluetooth range had seriously deteriorated. I keep my phone under thick pillows so that repeat low alarms don’t keep me awake at night. The G7 has quiet modes available to vibrate or even silence all alarms for up to 6 hours, and that looks like a significant improvement over the G6 and its predecessors. Alas, I am still stuck with the G6 alarms while running the G6 and G7 simultaneously, and under the thick pillows no longer works as it can cause Bluetooth failures on the G7 while the G6 still performs fine. Day seven had a few 15–25-minute data drops in the morning, leading me to believe that my sensor could fail at any time, so I kept my phone within 2 ft of my sensor, but as the day progressed, I went back to having steady, solid readings.

Day eight, much to my surprise, after the day seven dropouts, I fully expected the G7 sensor to fail on day 8. However, I was back to getting all readings even when my sensor and phone were 20 ft+ apart. On day eight, I only appear to have lost two dots in the entire 24-hour period. For the first time, I questioned my heretofore hypothesis that losing Bluetooth was due to an accelerated transmitter battery life degradation. With the G6, once sensor dropouts repeated with increased frequency, the sensor would fail and need to be replaced.

On day nine, the “Brief Sensor Issue Don’t remove the sensor. Temporary Issue. Wait up to 3 hours” resurfaced. The biggest issue was that, like the G6, when Bluetooth does not work for a short time, most of the missing dots are backfilled when it works again. Again, this is a reminder to dose the trend, not individual data points.

On day ten, everything worked perfectly without even one dot dropout, and changing to a new sensor was a breeze. I love the 30-minute warmup.

Since currently there is no way to restart a G7, I am only looking at both devices in 1 standard 10-day period. I never calibrate my devices, as my history has been that calibration is a nuisance and makes it harder to manage my data. Occasional signal losses and the occasional dot spatter pattern don’t bother me either. In comparing the G6 and G7 graphs for the same period, it is easy to see that the G6 just takes the same data (and data losses) and smooths the data, so you normally don’t notice losses without getting a major extended loss. The G7, starting Day 3, missed the odd dot here and there and shows a scatter pattern about once every 24 hours. That pattern and loss represented 30-42 minutes per 24-hour period and can easily be mitigated by watching the trend before the loss. Given the body and technology variances, it never makes sense to micromanage data points. Look at data points, sure, but BG management needs to be macro-managed based on trends, not micromanaged using a few data points. I suspect that integrated pump manufacturers use this same philosophy.

This first 10-day test, I am managing my insulin based on the G6 device, however, closely monitoring the G7 results. For my second 10-day test, I will manage my insulin based on the G7 while closely monitoring the G6 results.

My expectations must have been low as I entered this exercise, expecting the G7 to be a miserable experience. So far, I have been highly impressed with the G7 results. Is the G7 perfect? No, it still requires substantial improvements to be perfect. Many features are a significant improvement over the G6, and the only setback is that Bluetooth is or appears to be more unreliable. With the smoothing and backfilling of the 5-minute dots on the G6 and the backfilling on the G7, it is hard to pinpoint how much worse the G7 Bluetooth is versus how much better the G6 was at covering up its lost data points. Is the G7 better than the G6? For me, yes, but YMMV. It is indeed a different experience, and experience is often an emotional rather than technical difference, especially within a group of people that may be inherently reluctant to change. I also can’t overlook the G6 “smoothing”, which was hiding data losses and giving me a false sense of security. For some people, like a white lie, that is a benefit, while for others, not so much. Either way, the G7 sure beats frequent finger sticks, and diabetes management is all about minor incremental improvements until a cure is found.

Update: After running two consecutive parallel 10-day tests, I am ready to go full-time exclusively on the G7 before I go nuts. Segal’s law came at me full force "A man with a watch knows what time it is. A man with two watches is never quite sure. " In the second 10 days, the G7 never dropped and never even missed one dot in ten days. Every so often, there would be up to a rare 50-point BG difference between the G6 and the G7, and every time that happened, I did a fingerstick and found the G7 to be within 20 points of the fingerstick, and the G6 was way off. I am sold on the G7. YMMV, but I would recommend, if you are upgrading from the G6 to the G7, to run them parallel for 5-6 days to get a good feel for how the algorithms and results differ.


Good review and comparison. There is a few good improvements to the G7 but are minor in my opinion and they missed out on two big ones, which is a Libre “competitive” 14 day sensor life, especially because it can’t be restarted like the G6, and direct to watch. I hear they are coming to the G7 but in my experience Dexcom keeps promising and not delivering because it is all about marketing and maximizing profits.

1 Like

I believe direct to watch is available. The problem I see is that the Bluetooth can only connect to one medical device at a time and I rather have that be my phone and then use either the Blose or Dexcom App on my watch to poll the phone every 5 minutes. I think it is just habit or an emotional tie I am not quite ready to break, There is really no technical reason I see not to pair my G7 directly to my Google Pixel 2 watch. Right now I am running G6 and G7 parallel so don’t want to add any additional challenges, but when my test is over, I may do the direct to watch on the G7. I am not currently using any third party apps such as xDrip to bring the data to my watch. The polling option is an Android app that can be directly downloaded from the Google Store.

No simple direct to watch that is endorsed or provided by Dexcom yet. See my post at:

I am using my last G6 sensor now, so I am very glad that you shared your good experience with the G7. I always choose to calibrate but maybe I will try going without the constant need to calibrate. Thanks for sharing your experiences CJ.

1 Like

Super helpful summary - thank you !


Thanks for the very thorough review. For me, moving to a 30 minute warm-up from a TWO HOUR WARM-UP was enough. G7 is still not approved for Omnipod, however.


For me, I am hoping to be able to use G6 until the G7 will work with BIQ. I don’t want CIQ - target is too high, the algorithm won’t do the “automatic” insulin dosing until you reach 180 (I never allow myself to get that high) and I use temp basal function often. CIQ doesn’t allow for that. I highly value the BIQ suspension when going low.

1 Like

@CJ114 Thanks for the report as I’m thinking of G7 too. What I want to comment on, though, is you’ve blown my mind! I have NEVER thought of putting my pump or CGM in the middle, ie above the belly button! Guess I’m just a right/left kind of guy! I have a rotation for my pump (Omni Dash): left side of stomach bellow rib cage w/ rounded side to left, then same area w/ round to right (moves the cannula), then the right side, then the left arm inside bicep round up/down, then left tricep (back of arm) round up/down, then repeat. I’ve never used the right side, go figure. Gives me about a 24 day rotation. Now I have to give some thought to the center position! BTW, no problem with compression lows and I’m a side sleeper (both sides) too!

My CGM positions are from the belly button up to the xiphoid process. Sometimes, I remain on the centerline and sometimes 1/4 inch or so to the left or right of the centerline, giving me many options. I am currently running concurrently G6 and G7 for a few weeks to clearly understand the difference in performance between the two devices and the devices and fingerstick results. Here is a picture of my last G6 about to run out, the new G7 in place, and a G6 at the end of its soaking period ready ready to accept the G6 Transmitter.

I asked my doctor in December to update my prescription from G6 to G7. When I received my 90 day supply of G7 I was surprised of 9 sensors 6 are “rev3” and 3 are rev 12.

On purpose I waited many months to switch to G7 hoping to wait for the bugs to be worked out. At least my supplier seems to still have a lot of early rev’s that likely are not going to provide the best experience.

If I had known I’d get a bunch of rev3 I wouldn’t have switched yet.

@artwoman, I too thought the CIQ target was too high but after reading other experiences, I have joined the 100% sleep mode crowd and am very happy with the results (TIR 95+, A1C 5.5).

1 Like

What is rev3, what is rev 12? thanks!

These are the G7 revision numbers and they keep going up as Dexcom tweaks their Sensor/Transmitter device. I am not on the latest revisions and they work fine for me so far. They seem to have started their Rev numbers again with the devices that have been approved for pump integration, so the Rev 3 is actually possible to be the latest current model. The latest Rev’s appear to have a white underlining the Rev number while the older ones with higher Rev numbers were not underlined.

I tried a G7 6 months ago and it was an early rev. The only issue I had was poor signal from the sensor to my phone.

Everyday I went for a walk my phone was in my right front pocket and I would get alerts for signal loss because the Dexcom was on my left arm.

Got it, thanks! Didn’t know where to look, now I do. This forum is terrific, important information generously shared, and the caring and knowledge of its users when there is is a question, is invaluable.

1 Like

I agree, the bluetooth is worse than prior versions but hopefully that is being dealt with in the more recent versions. Other than the bluetooth issue, everything else seems a vast impreovement.

“Rev” on the green box flap indicates the label (box and included information) has changed. It does not indicate the device inside has changed.

CIQ - even with sleep mode - doesn’t allow for me to be as aggressive in my T1D mgt as I need to bee (I was dx’d at stage 2 kidney disease at 15 years olf. Dx’d at first with T2D since I was 30 years old, and Dx’d T1D 6 months later. Due to my poor kidney function it behooves me to have very very tight and aggressive mgt. Having access to temp basal helps me to essentially stay between 60-140 (although like all of us there are rogue bg’s). After 40 years with T1D, my kidney disease is at about 3.75 (so to speak). And I am still alive. I don’t mind the extra effort it takes, and I WILL NOT do dialysis. That would impact my ability to work and I am not eligible for Medicare so it would be outrageously expensive. Avoiding high (or high-ish) bg’s has kep my kidney function fairly stable.


thanks for this! helps me get ready to make the change!