Diabetes Technologies that never took off. Whatever happened to the “Sprinkler Needle”?

Diabetes technology that never took off. Whatever happened to the "Sprinkler Needle"?

One of the biggest difficulties people with Diabetes have with MDI is timing insulin delivery. This isn’t helped by advice received from HCPs which tends to err on the side of caution in terms of delivery timing.

In the past (30 odd years ago in fact) research has looked into ways to improve insulin absorption, and effectively try to improve the issues with this. One thing that really stands out for me is the “Sprinkler Needle”.

The Sprinkler Needle

This is as straightforward as it seems in its name:

It’s a needle with a load of holes down the length and around the circumference, and this configuration was shown to have a noticeable improvement in insulin absorption. How noticeable?

Well, the data from a study run in 1987, using Human Actrapid insulin showed some pretty impressive results. Bearing in mind this is a set of n=10, so only a small sample, for me the points that stand out are:

  1. The increase in plasma insulin concentration earlier after injecting. The image below shows that the sprinkler needle has an almost 50% greater absorption at 30-40 minutes after injection. In terms of insulin, that’s the equivalent to the difference between Fiasp and Novorapid.
  2. The reduction in postprandial glycaemic level. The data seems to show that with the same injection strategy, using Human Actrapid insulin, the postprandial rise using a sprinkler needle at worst was roughly 75%-80% of that using a conventional needle.

Both these observations suggest that sprinkler needles are, in the treatment of Type 1 Diabetes, a good thing. Add to those the research done on site dispersion for large doses, this time using Insulin Aspart, and the benefits of both needles and cannulas that provided better dispersion of insulin become very apparent.

The research using Aspart suggests that for an 18u dose, delivering it in nine 2u doses instead of one 18u dose provided a third faster absorption up to 50% of max insulin concentration point. That, of its own count, should encourage anyone injecting insulin to use multiple sites for large doses.

If you spend more time digging through the literature, it also becomes apparent that delivering insulin through multiple apertures reduces the pressure required for delivery. This is likely to benefit pump manufacturers as well as PWD in terms of insulin delivery, in both a reduction in pressure needed and better resilience in the presence of blockages.

Given this experimentation was undertaken with ActRapid, which we all accept was not the fastest of absorbing insulins, imagine the benefits of this type of technology with the current batch of fast and ultrafast acting insulins. I suspect that we’d be hugely surprised and very pleased.

With all these benefits appearing obvious, and research dating back 30 years with follow ups in the interim, why is this not a standard of diabetes care in the 21st Century?

Issues with Sprinkler Needles

Whilst this idea has been around for a long time, and there have been a fairly significant number of patents submitted, nothing seems to have made it in to production. One of the most well known names to hold a patent is Unomedical, who make cannulas for Medtronic, Animas and Tandem. With all these patents out there, what’s caused these not to become mainstream?

There is little data available to determine the answer to this, however, relatively recently, Becton Dickinson worked with Medtronic and the JDRF to introduce a new, better cannula that was intended to reduce occlusion risk. In two years of pumping I’ve had no cannula blockages that might show up as occlusions, however this was felt to be necessary. This was known as FlowSmart and added an additional side opening to the cannula which was supposed to make insulin flow better.

FlowSmart ran into problems early on though, as users reported unexpected high glucose levels and when removing the sets, seeing kinked cannulas. This occurred to such an extent that BD issued a voluntary recall. Reading between the lines, the addition of a hole in the side of the teflon cannula was causing enough of a reduction in structural strength to cause the cannula to kink on insertion.

This may, of itself, provide a leading indicator as to why this type of needle/cannula has not made it into wide use. Considering the idea:

  1. Manufacture of something with a blocked end and holes up the sides is likely to be harder and more expensive than a simple, single, tube. The BD products mentioned earlier were ~25% more expensive than traditional cannulas.
  2. Structural integrity of a needle/cannula using the gauge that insulin needles and cannulas are now presents something of a challenge. The very nature of adding additional holes to a tube reduces its longitudinal strength and increases the flexibility. By making it more efficient, there is potentially a higher risk of kinking/concertinaing on insertion.

These points alone would make the investment in this technology tricky at best, especially when you consider the overall population that will benefit from it and the potential return, and probably demonstrates why FlowSmart had charitable involvement in its development.

That’s not to say that no-one is interested though.

Is someone looking at better insulin delivery through these techniques?

The answer to this one is yes. A start-up in California, Capillary Biomedical Inc. They describe themselves as

a medical device startup developing technologies for diabetes management. We aim to dramatically simplify glucose monitoring and insulin delivery, reducing the burden on both patients and the overall healthcare system.

The key point here is that they are working on a sprinkler-type cannula for use with pumps, and as ever, the JDRF is supporting this research. Based on limited interaction with the company, or in other words, I’ve read their website, they appear to have prototypes that work and have been operating. The below image shows one of their cannulas infusing across multiple layers:

The details the company presents so far suggests that they are looking to deliver something of interest to those of us living with T1D, and the outcomes of other experimentation suggest that it should be beneficial.

What can we take from this?

It’s obvious that treating diabetes is not something that is limited to artificial pancreases, faster insulins and stem cell transplants. In the here and now, what appear to be very simple solutions could and potentially would improve lives for anyone with Type 1. The simple fact that it’s possible to repetitively and deterministically improve insulin performance by a simple mechanical means shouldn’t be underestimated.

At the very least, with the knowledge of the research, as a user on MDI, if you are planning on eating something that needs a large insulin dose, split it into more than one part in different locations. Then you’re at least getting the benefit of a sprinkler approach.

How long will products like this take to come to market? I can’t answer that one. But the overriding principle of “Keep It Simple, Stupid” may provide a surprisingly large benefit to many of those living with T1D.


  1. Makes a lot of sense. Was taught to split doses over 4 by Dr Bernstein

    Thanks for your articles. I follow them all

  2. Very interesting. I’m not a pump user but was wondering whether you’ve tried running two cannulas at the same time just to see how much this helps absorption. Is it even feasable to T off one cannula with a short link pipe to the second?

    Keep up the good work!

    • I’ve never tried splitting the tubing either. I have seen good results with applying boluses over multiple sites.

  3. Actually Tim it was 9 x 2u doses (not 2x 9u). Bernstein always advises splitting any dose above 7u which seems to be good advice for those on MDI.

Leave a Reply

Your email address will not be published.