What goes up…. Must come down. Unless you’re a #PWT1D using #Fiasp.

What goes up…. Must come down. Unless you’re a #PWT1D using #Fiasp.
What goes up…. Must come down. Unless you’re a #PWT1D using #Fiasp.

So then. 10 days after switching back to Novorapid because I was having a devil of a time with the Fiasp, I received a new prescription of the liquid that, in the US at least, is more expensive than Printer Ink (and therefore makes Gold look cheap). Time to start off with the whole shebang all over again.

Let’s remind ourselves of the marvellous qualities of this version of Aspart once again. Roughly 50% more insulin available in the first 30 mins post bolus leading to a much quicker effect, which was great while it lasted.

After the upset of last time, I’d had to determine for my own sanity, whether there was simply an off batch. I hadn’t suspected so, but still, one can never tell, so we did one more prescription. If Alliance were supplying the Novo insulin in a bad way, the first time and that was what has caused issues, then surely that couldn’t happen a second time?

Just to confirm, the insulin was received whilst at work from a fridge. It went into a fridge at work, and then was retrieved before leaving. It spent maybe an hour out of the fridge before returning home. The vial used for the pump refill came straight from the pharmacy.

And in it went.

On first use, it seemed to work well, with the rapid action that’s expected the first time I bolused. Overnight, there was a slow increase, then this morning, I had a cup of coffee. Normally I tell OpenAPS that that is 10g of carbs, and the subsequent insulin delivery manages the Ephidrine linked blood glucose rise.

This is best reviewed by taking a look at the below picture:

So let’s look at the blue box. That’s Fiasp as I expect it to work. No pre-bolus, very quick action and allows me to run just off Super-Micro Bolusing. And a welcome return too.

However, the orange box. That’s where we see things misbehaving. You can see in the IOB pill that there are 3u of IOB. There was a 2u bolus at 9.30. That means that following the 10g carb equivalent entered for coffee, OpenAPS and I had given me 4.8u of insulin as SMBs and a bolus. With an overall change in glucose level of about +2mmol/l. Now by my reckoning, that’s suggesting an IC ratio of something in the order of 4u per 10g of carbs, or 1:2.5. And we’re not down yet.

It looks as though the issues I was seeing at the end of the last period have not been resolved with a ten day use of Novorapid.

The final test in this process is that 2u bolus at 0930. If the Fiasp is working as it should, I’d expect to see a steep drop commencing within the first hour after the delivery of the insulin, and a drop of around 4mmol/l that will need intervention, as I should see glucose levels drop below 4.0mmol.

Instead, we have this:

As the purple box area shows, whilst still fasted, a 2u bolus (plus whatever else was onboard) has lowered glucose levels by what amounts to not a lot – 7.5ish to 5.8. A drop of only 1.7 mmol/l. This equates to an ISF of around 0.85mmol/l/1u, which would also equate to a Carb Ratio of 3.5u/10g carbs.

I think that it’s fair to say at this point that this is not acting quite as the original hype, or my own for six months, suggested it should. As a matter of fact, it’s not even acting like regular Insulin Aspart (Novorapid).

But anyway, leaving it for a further two hours results in a further drop, some of which has to be attributed to the decision from OpenAPS to SMB when the curve flattened out.

Now by this point, I’m hungry, and lunch is looking interesting, but I’m in a quandary. Do I try and bolus with these ridiculously insensitive ratios that seem to be generated, or do I syringe Novorapid?

Well it depends on the hypothesis I’m following. If it’s that my liver isn’t responding to Fiasp, or that I have localised sensitivity, then, yes, okay, that makes sense, but what if it’s something else?

Let’s suggest that it’s an immune response or insulin pooling, which the above graph could also be suggesting. If it’s that, then doing a bolus at these ratios isn’t like to stop a lunch rise, and more importantly, is likely to cause a significant drop well after I’m expecting it.

So, in the name of “Science” and with some careful observation going on here, that bolus was done with the ratios calculated earlier. A 1u/4g of carbs, with the knowledge that there is a 60g protein load on at the same time. What we see seems much more akin to the effects of Fiasp that I’ve previously seen. Both the ISF and Carb Ratio in my pump have been adjusted and OpenAPS is using those to calculate its adjustments.

The updated values, for now are least, seem to be working.

This does lead me to suggest there is some form of insulin resistance going on, whether that’s local to the cannula site or generally, and that potentially what I previously saw when I thought things were going horribly awry was this having an effect and not enough insulin covering both glucose absorption and some level of incretin effect generating glucose from the liver.

In fact, the last 24 hours aren’t too bad. With the exception of the over announcement of a meal in the evening, resulting in a low that needed to be treated (which would have been the same if I had manually bolused), what I’m seeing is that the Fiasp seems to be acting as expected in terms of its action time.

The latest dialled in settings, with a day of observation, i.e. not really long enough, seem to suggest that there’s been a quantum change in mealtime insulin need (I note there is less of an issue on basal, although a slight one) more than anything else.

So the question is, do I continue in this new land of much higher insulin need, or do I give up on it and go back to Novorapid? I still like the fast action of the Fiasp, but is it healthy to be reaching a TDD amount which is starting to look like the doses given to someone who is suffering the early identifiers of double diabetes?

If we look at how my insulin sensitivity factor has changed, it looks like this:

And what about the costs? Such an increase in insulin usage results in 10ml phials being consumed every 11-12 days instead of every 21-22 days. That may mean the per volume unit cost is the same as Novorapid, but it also means the required use cost is double.

Whilst I am sure there are people out there who are happily using Fiasp with no issues, there’s been a lot of noise amongst social media sources that raises questions. It’s clearly not easy to switch to and changes happen irregularly and inconsistently. I’ve got a ton of data, a set of tools and a loop to help me manage this. Without them, it would be a nightmare. As others have found.

So with Fiasp, what goes up does eventually come down. The only question is whether you can predict when that is likely to be.


  1. This is almost exacttly 🙂 the same what happened to me when I tried out Fiasp in the Omnipod. First day sort of okay., but needed more insulin immediately (20%) and then chaos. At first I thought it to be my fault but second Omnipod gave exactly the same. I can almost put both curves on top of each other and woildn’t notice the difference. Was my first time with Fiasp and not some time after the start. I took of my Omnipod still working and watched how basal and bolus came out: not good but that is not hard science. I had the feeling the problem is the Omnipod with Fiasp. I don’t use much insulin and don’t do large bolusses and maybe that has something to do with it. Drove me crazy. Back to Novorapid now and a happy flatline and no problems. Maybe I’ll try another vial next week out of curiosity…

  2. Very worrying. I like many others am following your experiences with great interest. I’m MDI but have also seen big variation at times with Fiasp. As a result of your experiences and my own I have now reverted to Novorapid Nd just use Fiasp for corrections which seems to be working pretty well.
    Look forward to following your excellent reported progress.

  3. It would be interesting to see the results of a different brand of insulin, Apidra is another slightly faster insulin. I am using Fiasp with similar results in that needing more insulin overall. I have some Apidra on hand and will be using it for my set change this morning.

  4. I’m 10 weeks in using Fiasp in a pump with CGM and the most noticeable thing I have found is the change in the amount of basal changes needed. 1 week all is fine then a tweak here and there is needed, leave it a week then it’s either an up or down adjustment again and so on. Carb ratio doesn’t seem to be affected since the initial change which entailed less insulin per carb and the same for correction.

    It’s a very time consuming way to deliver insulin more so than using animal insulin in my pump and only a .1 point drop in A1c down to 6.3

  5. Thanks for this, Tim. I appreciated your post back in June where you gathered the experiences of others using Fiasp. Any plans for a followup to that? It would be really interesting to see if the behavior you’re seeing is common among those who have been using Fiasp for a while.

    • Hi Eric, based on the comments on this post, some of the responses to the survey and the conversations I’ve had both in person and on Twitter and Facebook, it seems that these type of things are happening to larger number of people who have used this in pumps than I expected to see. I’ve continued to ask people to add those bad experiences to my survey form to try and gain some idea, but six months down the line, I’m not sure there are that many Fiasp users still to give a good cross section of feedback.

  6. After 3 months on Fiasp in Omnipod pump my HBa1c hasn’t changed noticeably, 39 rather than the previous 3 years, 37. My TDD is possibly about 10% higher and according to Libre my time in target is roughly 60% rather than 52% on Humalog so nothing too awful so will stick with it for another few months unless there is significant change. I’m now aware that there are others having problems and they would be the ones that post but I’d be interested to hear of anyone else who is similar to myself. Although this is positive I’m not yet convinced I’ll stay on it as there do seem some inconsistencies that I can’t quite explain, especially when there are so many other factors to take into account in general diabetic/pump life if this makes any sense?

    • Interesting Jane. Would you mind detailing the inconsistencies as that may help others identify if they are also seeing them. FWIW, over my six months before the latest change, it was working well.

      • Really it’s the processing time of higher carbed/ fatty foods, the extended bolus doesn’t stop the night highs and I wake after this sort of meal the night before, to find my BG much higher than I want or need, even after say 40g of rice, creamy pud, small amount of chocolate. An earlier increased bolus leaves me low within an hour of eating. When high it takes much more of a bolus to correct than it would with Humalog/ NovoRapid, it’s all a bit hit and miss but still early days after 3 months. Low carb is fine and I stay between 4-7.5 all night which I’m more than happy with.

  7. After 5 months of use I’m experiencing exactly the same issue with Fiasp, suddenly I need a lot more than usual (around double to bolus or correct), my Basal requirement is up 1/3. Very very odd.

  8. I have tried #fiasp in the #omnipod for the past month, switched from #novolog and similar to others have seen inconsistencies, specifically like my basal rates no longer work and I need to bolus every 3 hours to avoid running high. For meals I typically only bolused 2-4 units with novolog but need to do 1.5-2x to get the same effect. I’ve also rage bolused and had more lows. It does get on faster with a bolus but I appear more insulin resistant and feel it. I was so excited for the faster acting hoping to fix the postprandial rise but nah…Another mentioned this requires more insulin to cover for coffee and higher protein or fatty foods and I have found the same. I thought maybe it was an Omnipod thing? I wonder about the stability using it in the pod and for low bolusers if it becomes less stable I.e 2 units may work differently than 6 units. My BG levels have been higher with peppered extreme lows and I am switching back to Novolog today with my site change.

    • I think the only thing it’s fair to say about Fiasp is that people seem to have fairly varied experiences of it. I’m not sure that’s specifically Omnipod related.

  9. I’ve had similar issues on FIASP, been using it for over 9 months now. I use it in a Medtronic 723 pump (open loop, using G6 CGM with xdrip+/nightscout,I wish I could find a decently priced pump for openAPS). Initially I loved it, initial onset is quick! However as time has passed, I’m starting to see more drawbacks than advantages.

    First major drawback:
    As a very active outdoors person, I’ve found that a normal meal bolus combined with ANY amount of activity (walking, cycling, rock climbing) would immediately PLUMMET my blood sugar. The only way I’ve been able to have any stability in being active was to either not have any insulin on board and suspending basal when starting my activity, or have a TON of low supplies at the ready for whatever my activity was. This was a huge inconvenience. It’s as if I got my heart rate up at all, almost all the insulin I used was onset immediately, not the initial hit and then 3-4 hour normal delivery like I saw using FIASP if I was just at my desk at work.

    Second major drawback:
    Lately, I have seen a HEAVY rise in insulin use recently that led me to think I was getting major insulin resistance problem. Went from about 11g/1u boluses initially, all the way down to 8g/1u lately. I’ve also had the same low BG after dinner phenomenon, and then I’d treat like normal for the low, and then end up going high all night. I even tried dual-wave bolusing with almost no initial insulin delivery and relying on a 2 hour delivery window, with similar results. I steadily kept increasing overnight basals to try and catch this, to the point where I went from ~.7u/hour to the point where I got to almost ~1.2u/hour for the first half of the night and was still having overnight high problems.

    At the beginning of this year I started a keto diet to try and just cut out the carbs and need for insulin, and got wonderful results for the first two weeks, using less than 20u/day, average BG of 112, really only using microbolusing when I saw the rise in BG to correct for the <10g meals I was eating. Now, two weeks later, I'm starting to see the overnight highs coming back, need for more meal insulin (even with extremely low carb, high fat meals).

    I think I've done about all I can in trying to make FIASP work. My next pump swap I'm going to use some of the leftover Novolog stock I have to see if that will help, I remember being much more stable on that insulin before.

    All in all, I don't think the fast inital onset of FIASP is worth the instability. It might be OK on closed loop, but I can never adjust my basals fast enough manually to make this insulin work for me, and I've determined it's just too unstable for my lifestyle.

  10. Like others here, this was my experience on Fiasp a week after using with Omnipod dash. I didn’t go down the path of adjusting my ratios as I could tell they were becoming insane. I was at 4.7 ISF on novorapid and on Fiasp, I was injecting it like water at one point. I changed sites multiple times to no avail. Switched back to novorapid and things have settled down again, though I am slightly more resistant than before the Fiasp.

Leave a Reply

Your email address will not be published.