Posted by on Monday 12 March 2012

iBGStar Review - blood glucose meter for iPhone / iPod Touch

I was delighted to be invited in January 2012 to be one of the first people in the UK to use and review the iBGStar - a blood glucose meter by Sanofi Aventis designed to work with an iPhone or iPod Touch. The agency that sent me the meter and asked me to provide feedback prior to the UK launch and swore me to secrecy until today, when it becomes available for purchase in Boots. I was provided with an iBGStar and a supply of strips free of charge, and I've not been paid to write this review or say anything, good, bad or otherwise - what follows is just my opinion after a few weeks' use. To save my fingers I'll lazily just be saying iPhone from now on. If you've got an iPod Touch you'll need to do a little mental inline editing.

I should start by saying that I reeeeeally wanted to like this meter right from the outset. The designer in me has long had a love for that fine fruit-based technology company. I've been logging all my BG results on Apple products for years and the idea of a meter which plugs straight into an iPhone or iPod Touch has immediate appeal.

First impressions
The first thing you notice is that it is absolutely tiny! There's also a pleasing 'Apple' influence to the general design. The shiny curved edges on the front, the smooth nicely shaped matt-finished reverse side and, perhaps nicest of all, the fact that the on-off 'button' is seamlessly built into the front fascia of the device. The designers have also bothered to build-in a nicer-than-expected screen on the front of the meter which displays results and a small animated prompt during BG tests which means that it is perfectly possible to carry out a test whether the meter is connected to your iPhone or not (handy if, for example your phone ran out of charge when you were out for the day). First impressions? Even before I'd got a strip out of the pot - this is going to be amazing.

How small is small?
By my reckoning it's about 55mm by 24mm (including the dock connector). Everything about this meter is smaller. The strips are smaller, the pot they come in is smaller. The supplied finger pricker (more about that later) is small too. The supplied black zip-up case is perhaps 50%-60% the size of a standard Roche Accu-Chek one and maybe 70% the size of the case for the Contour-Link meter that works with 'Artoo' my pump. If you took the strips, fingerpricker and iBGStar loose in a coat pocket they would take up very little room. The iBGStar aligns well with the width of my iPhone4 and is just slightly narrower than my 1st Gen iPod Touch. Of course you don't keep it connected to your mobile device the whole time, and with another nice touch they provide a small cap to cover the dock connector which doubles as a little clip to secure the meter and stop it getting lost in the bottom of a pocket or bag if you take it 'loose'. The advantage of this tininess, of course, is that it makes it much less hassle to take your meter with you. And you don't have to worry about all the careful logging etc when you are on the run, unless you want to. The iBGStar can be used perfectly well on its own and will remember and allow you to refer back through several readings on the device itself (it stores 300). When you have a bit more time you can sync with your iPhone and add any insulin doses/carbs/notes to keep your records up to date. If you were having a period of lazy no-logging the 'unplugged' iBGStar helpfully displays an average figure when first turned on allowing you to keep a general eye on things. To get the most of the meter though, you'd want to download and use the iBGStar App.

The iBGStar App
Downloadable free of charge from the App store, compatible with iPhone, iPod Touch and iPad. Requires iOS3 or later. I haven't tested any iCloud-based shenanigans for people with multiple devices, so I cannot confirm or deny whether the iBGStar App would allow you to use several devices concurrently. Having taken a while to find a logging app that suited me I was a little worried that the iBGStar app might be a bit clunky, but was pleasantly surprised. Data entry is slick and easy and individual records consist of tabbed 'cards' which can be flicked through on the home screen. Easy enough to add and edit things like insulin dose and carbs eaten, while a series of 'notes' can be added from a prepopulated menu of food/exercise/dose/meds/general. There are some pretty sensible entries already there 'ate fatty food', 'light exercise' and so on, but you can edit and amend any of the pre-defined choices to make the app work for you. You can also add any extra ones that you might find helpful. There are some other sweet personalisation touches too. For example you can choose the 'background theme' to the home screen from six choices (cityscape, beach, forest...) and as you flick through the result cards watch the sun rise, set and the stars come out depending on the time that the BG reading was taken. I liked the zoomable 'scatter graph' in the data analysis section too, which colour-codes readings above and below the target range you specify. You can also view averages, numbers of tests and even Standard Deviation (how raggedly spread out or tightly packed your results are) over 7, 14, 30, 90 days or a custom range. Critically it is also very easy to email-out the results data, including all notes, carbs and doses for printing and waving in front of your Doctor or DSN. The only tiny quibbles I had with the app were that it did not allow fully pump-friendly dose recording (Artoo does meal doses to two decimal places, the iBGStar App only allows one) and the 'logbook' view is a bit limited showing BG results only - with the 'expanded view' being particularly useless and simply showing the same BG data a bit more spaced out rather than revealing data for doses, carbs etc.

Taking a test
The supplied finger pricker is er... a bit flimsy. It's servicable enough and at least has a variable depth setting, but lacks the smooth solid action of my Accu-Chek MultiClix and has a slightly worrying, plasticky, 'cracker novelty' feel to it. It stayed in the box. The sample size is a modest 0.5 microlitres, strips don't need coding and results take around 6 seconds. One slightly surprising irritation was that the strips felt like they were the 'wrong way round' in the pot. Unlike any other meter I've had the 'sample' end of the strips for the iBGStar are not the end that you grab hold of, so each time you get a strip out you need to turn it around before pushing it into the meter. The strips themselves are also a fairly uniform black so it's not always immediately obvious which end is which. There is a tiny notch in the 'meter' end, but in low light and for overnight tests this was a full 'lights-on' meter. I also found it slightly tricky to apply the sample in that the finger-to-strip alignment seemed to need to be a bit more precise, and the strips themselves a little less 'slurpy' than other meters I've used. If you use the meter plugged into your iPhone and with the iBGStar app running you get a little swooshy animation on the phone screen after the 'apply sample' prompts. I preferred to use the meter separately and then plug it in to sync data and log meal and/or other details. The iBGStar has a rechargeable battery and is supplied with a mains charger or can be charged via computer USB port. I have a fairly intensive testing schedule but a full charge still lasted my me a good two or three weeks. Helpfully the iBGStar begins to show remaining charge as a percentage once you have dipped below 50%.

Accuracy
So far, so good. If I were finishing the review here it would be a straight 4.5 stars. Unfortunately though I have a nagging worry. Quite by chance, when I was asked to review the iBGStar I had recently switched from an Accu-Chek Expert to a ContourLink (the latter sends results to Artoo via bluetooth). When I switched between those two meters, out of interest I did a couple of tests between the two. Now I had three meters on the go, and strips for all of them, so I did a few random tests across them all. Now at this point I must absolutely stress that these tests are in no way 'properly' scientific. The sample size is one - me. I do not have access to any lab results for any of the blood samples so I can never know which of the meters, if any, was ever giving a truly accurate result. All meters are permitted to have a 20% variation from a lab result (there's a whole blog post right there) so any result could be a bit high on one, a bit low on another. All I can say is that I got a larger-than-average drop of blood, stuck a strip into all three meters, tested and compared the results. To begin with it was all hunky-dory. But then I noticed a test where the iBGStar read quite a bit higher than the Contour. A few more tests followed and by and large the Contour and Expert were closer together, the iBGStar almost always higher. And the higher the level, the more extreme the variation. As a pattern I actually liked that predictability. I know people who find their correction doses less and less effective as their BG rises. A meter that reads a little higher at higher levels could be ideal - set the correction factor appropriately and you would automatically get a more effective correction when you needed it. Sadly though I soon found the results that broke the pattern. Results with larger differences at lower levels, and results will almost no difference at higher levels. I mentioned this in my 'feedback' to the agency who had sent me the iBGStar and had fantastic customer service experience from Sanofi Aventis who sent me a new meter along with some control solution and asked for the old one back so that they could have a look at it. I've not heard whether they found anything, but the results with the new meter (which passed the control solution test) are very similar to the first.

Once I had the replacement I spent a couple of days cross-testing the three meters just to double check before I mentioned this experience here. There were 22 tests taken over a 2 day period including before and after meals, before and after exercise and overnight. For what little they are worth here's what I found when I looked at the results:

  • In general terms the iBGStar read the highest averaging 8.9mmol/L (160mg/dl) and the Contour the lowest averaging 7.7mmol/L (139). The Expert averaged 8.0mmol/L (144).
  • The average difference between the iBGStar and the Contour was +1.29mmol/L (+23mg/dl)
  • The average difference between the Expert and the Contour was +0.35mmol/L (+6mg/dl)
  • Notwithstanding the official 20% 'acceptable' variation, I consider +/-0.5mmol/L (9mg/dl) to be a better indication for my own purposes of whether or not readings are 'more or less the same'
  • Only 18% of results between the IBGStar and Contour were within 0.5mmol/L
  • Between the Contour and Expert a much more respectable 73% came within 0.5mmol/L (with more than a quarter of results coming within 0.2)
  • As soon as differences between meters reach +/-1.5mmol/L (27mg/dl) it begins to make a difference of almost half a unit on an insulin dose or correction for me. An extra 0.5u when I don't need it could even risk dropping me into a low level hypo
  • Over the 2 days 37% of readings on the iBGStar were 1.5mmol/L or more higher than the Contour
  • There were no readings that showed that much variation between the Contour and Expert. None.
  • The single worst difference was 10.5mmol/L (189) on the Contour vs 15.8mmol/L (284) on the iBGStar - a difference of 5.3mmol/L (95), that's more than 50% of the reading - the Expert read 11.2 (202).

Again I must stress that these are just my own not-very-scientific results. I have no idea which results, if any, were the 'real' ones, but the variation in the iBGStar and the lack of variation between Contour and Expert cast a shadow of doubt over the iBGStar for me. It may just be something to do with my blood chemistry and the iBGStar's strips - I have no idea. I may be the only person in the UK who would find this variation.

Conclusion
I really like this meter. It's small, nicely designed. The app is great, very easy to use and ticks almost all the boxes for me. I successfully used it on its own for a week and a half, even after I had noticed its tendency to read a little high. I simply reduced my 'usual' correction factor to take into account the possibility that the meter might think I'm higher than others I have used before. At the end of the day BG meters for use at home are only ever going to give you a general guide as to your levels. If you read the small print in the handbooks they never pretend to do anything else. Back in the real world though, I do need my meter to perform as consistently as possible. If it's going to have a quirk, that's fine - but it has to be a reliable quirk. My life 'playing at being a pancreas' is full of enough variables without adding another into the mix. I'm sure the iBGStar will work well for many people - unfortunately I'm not one of them.

Final score: 3/5

UPDATE: Sanofi checked the first meter and gave this response.

18 comments:

  1. Interesting review. I managed to get hold of one these devices early through my diabetic clinic in the UK. I was really looking forward to using it as in theory the ipod/pad/phone should allow for a really good, fully functional logbook. It should also be relatively easy to code a bolus calculator in there too.

    I used it for one and a bit weeks of testing.

    I was disappointed - I don't care how small the tester is if I still need to carry a case with finger pricker, strips (spare lancets? somewhere to get rid of used strips?). I don't want a finger pricker sitting loose in my pocket and filling up with dust through the hole in the bottom... or rattling around in my bag! anyway, rant over on that front!

    I was most disappointed by the app - you can record data but it doesn't do enough with it. As the reviewer states, the logbook is a waste of time even in "advance" mode (which just added a blank line of data to each day in my case!). The trend graph was ok, but I can't imaging looking at 90 days of data on it (450 points on that screen?!). A simple averaging algorithm to provide averaged data when looking at longer timescales would have been an idea.

    For me though, its the lack of bolus calculator with insulin tracking (like I get from my Accu-Check aviva expert) that was the killer. Yes, I can do the simple carb counting numbers in my head, but the tracking of earlier insulin on days where lunch and late breakfast are well within 4 hours of each other... its a really useful feature. And I get much better data analysis from the Expert both on the device and with the 360 software supplied by Accu-Check.

    If the app had been as good as it should really have been, then I would have picked up an ipod touch (I don't have one already or an iphone I borrowed one for the testing) and used it exclusively as blood glucose tester. It wasn't, I didn't, back to using tech that works ok but feels like its way way out of date.

    My final rating 2/5 but I'm not easy to please with blood glucose testers - the expert could be much better as well - I'd probably only rate that 2.5 to 3/5!

    Just my tuppence worth...

    Additionally, I did contact sanofi about their plans to add a bolus calculator, they said they had none "in the near future", but didn't say that they were even looking beyond that (why bother spending the £development?)

    Paul

    ReplyDelete
  2. Have you seen the Medisana GlucoDock http://bit.ly/xLEpYU available on the Apple Store (but more expensive) - I would be interested to see if you thought it was any better.

    Cheers,

    Joel

    ReplyDelete
  3. Think part of why the iBGstar is missing a lot is because of FDA restrictions and they had a difficult time getting it through the FDA.. since the FDA pretty much requires the ipod be a viewer device, and not process data (interpreting data) otherwise the ipod needs to be certified as a medical device as WELL..

    ReplyDelete
  4. Joel - I've not heard of the GlucoDock, it looks like it is not currently available in the UK. A compare-and-contrast would be interesting!

    ReplyDelete
  5. Good old FDA... I agree that there needs to be some level of control in what is available, but seriously! If I use a calculator and a bit of paper to work out the numbers do they need to be certified as medical devices? Just certify the code surely?!

    Anyway, that glucodoc looks pretty much the same as the ibgstar in terms of functionality (again no bolus advice or active insulin tracking).

    You know what I would really want from the "ideal" glucose meter? In no particular order a combination of:

    Built in lancing device and test "strips" (as per the accu-chek mobile)

    Bolus advice calculator and active insulin tracking (as per the accu-check aviva expert)

    Decent logbook functionality including logging of meals, insulin, activity (sports etc including intensity/duration), "freehand notes" (something like the old OneTouch UltraSmart but even better!). - I don't want to carry a paper and pen to do all of this!

    Decent physical user interface for all of the above that doesn't feel like its out of the 80s (i.e. smartphone style touch screen interface would be ideal).

    Clear simple but detailed user interface that gives me the information I need (average results (including pre and post specific meals), trends (both continuous and discontinuous but linked to activity).

    Anyone want to step up and make me one?! Do I ask too much?!

    ho hum...

    ReplyDelete
  6. Just goes to show how impossible it is for the companies trying to develop these machines... My list would have some things you mention above, but others from your wishlist would be on my 'really would rather not' list.

    Without question at the very top of my list would be a meter that read to +/-5% accuracy or better 99.9% of the time - the rest I can work around :)

    ReplyDelete
  7. The accuracy thing is a really good point.

    Luckily for me I work in a scientific environment where the +/- 10% I think most offer (they say within 20%, I can't imagine its really +/- 20...) is actually considered pretty accurate. So I kind of just "deal" with it automatically! If I have a test that says 4.2 but I feel hypo I treat it (4.3 - 10% = 3.8 ish) and if I feel fine at 3.9 I don't (well, depending on what I've eaten or done lately I might have something carby but not glucose tabs - don't tell my doctor that though!)

    The companies do have a difficult job, there's no doubt. But I never get the impression that they really want to spend the money to develop the systems as well as they could... They get their money from the test strips, not the meters. So as long as the meter works ok and the larger diabetic demographics are catered for what's the point (from a financial/business standpoint)?

    The things that bug me most are still the "look how small our meter is" lines when you still need the rest of the paraphernalia to go with it (which is actually one of things I've got used to, I don't want a tiny device I want a functional one which gives me useful data! I'm quite happy carrying around the little case for most of the time). The other thing is the "look how quick the results are". They've been less than 10 seconds for a good 10 years I'd say... where's the real innovation?

    Anyway I could rant and moan about my frustrations with blood testers for a long time! But it doesn't get me anywhere, if only I had the time/expertise/funding to design and build my own "perfect" device!

    ReplyDelete
  8. I really love the idea of these devices either plugged into my iPhone of synced later on, as mentioned though the let-down seems to be the software, and each company is unfortunately consistent in its approach not to seek out existing software but to create its own proprietary app from scratch and miss out features that a simple survey of diabetics would tell them they needed..! Whilst I haven't yet purchased either the ibgstar or glucodock, I do use SiDiary on my iPhone which includes bolus advice and logbook, if I move to one of these units I'm going to lose that functionality!

    ReplyDelete
  9. I have used a Freestyle tester for many years and was VERY excited at the thought of the iBG tester as I love my iPhone. Very disappointed after doing tests on both (plus an accu-chek I have) to find the iBG is always much higher (around 2.5). When I try and keep my levels between 4 - 7 a 2.5 fluctuation is unacceptable and can lead to bad things.

    With the Freestyle if I was reading 9 I may take a corrective dose but on the iBG I may actually really be 6.5... hypo time.

    Now abandoning the iBG and sticking to my trusty Freestyle.

    ReplyDelete
  10. I have also been trying out the iBG star and checking results with my Freestyle mini. I have had a difference of 5mmol/L with the iBG star reading higher then my Freestyle. I big disappointment. I hope they manage to sort this out.

    ReplyDelete
  11. Hey, I've finished reading this review and I do say, the iBGstar seems fantastic! I'm only 14 and I was diagnosed around 2/3 months ago with type 1, It's a pain in the butt having to write down my BG into a booklet, and all that. Sometimes I forget my meter at school or, I'm just too busy or too lazy to poke my finger, wait for the results, change the lancets, dispose the testing strip and write down the results. I love having an accurate number, but I've been wanting to check out this iBGstar because it seems, very handy and petite. I just wanted to know, if it actually worked, and how much were the testing strips? I'm from Canada, so I'm not sure if they have the type of testing strips here, or if they have the iBGstar meter.
    I hope they correct the accuracy on this!

    ReplyDelete
  12. So sorry to hear about your diagnosis SL. Not sure about strip availability in your country, but the strips are the standard BGStar ones if that helps. Yes the super-small size is great, and even with the strips and lancing device it still makes a smallish thing to carry round. I know you are supposed to change lancets every time, but mine do me fine for weeks which helps with reducing how much of a faff testing is :) hope it goes well for you I you use the iBGStar - you seem to be exactly the target market for it!

    ReplyDelete
  13. Mike you made a great report!
    I have been using the iBGStar for nearly six month and must say it is too dangerous as a Medical measurement instrument. 20% Tolerance is way out Dangerous! and sometimes it is much more than 20% as you experienced.

    Told the support about this incident but she could not understand the problem, typical sales person. To make it simple I told her an example with a Tyre pressure. Ask her how would she react if I gave her a Tyre pressure instrument that indicates 8.8 instead of 6.8

    Last night I was feeling shivering and felt like having a Hypo. Continued cooking my diner and checked my blood before taking my diner.
    iBGStar indicates 5.8 which I consider perfect!
    But my condition said No while my hands and body feels shaky. 30 minutes later after my diner things returns to normal.
    I skip my Metformin last night.

    This morning I took another blood sample on both iBGStar which meassures 6.9
    Accucheck meassures 5.0 !
    a near 2.0 difference !!
    The biggest difference I ever had was 2.2
    That is abusive tolerance !
    Medical instruments that has that tolerance should be banned.
    If I have a blood pressure instrument that shows 100/150 and actually is 80/130, I would ask for a better replacement or throw in the rubbish bin.

    Ok, all Negative been said!
    As an Apple & gadget lover, I say they have won the Design and especially their App !
    I will finish the remaining stiff and stop using this beautiful design gadget!
    I'm not risking my Health to such an idiot medical instrument.
    Again The App remains on my iPhone and will keep using it. It's a near perfect App.

    I like to see an improvement on iBGStar product and they need to get this Feedback.

    Extra opinion; the AccuCheck is not even the most accurate instrument either. The most reliable is LifeScan product from Johnson & Johnson.

    Again my advice is, Don't get this iBGStar unless you like to collect design gadgets for $50

    Grant

    ReplyDelete
  14. I have seen favourable reports for lifescan and freestyle meters (small research studies comparing both, presumably funded by the manufacturers). As far as I'm aware only the newer Bayer/Contour meters 'officially' promise more accuracy (+-15% rather than the ISO 20%) though I think almost all meters are more accurate than that most of the time. It does come as a bit of a shock though when a meter makes full use of the +-20% window as the results become largely unusable!

    ReplyDelete
  15. Your is one of the best post on accuracy I have read on the net. Thank you for your post. I wonder if you did further testing with, or continued using the iBGStar. I have used the IBGStar to compare with lab results during two separate Oral Glucose Tolerance Tests, by reading the IBGStar immediately prior to drawing blood sample at the lab. During the first OGTT done in October 31st this year, the IBGStar read 7%, 9% and 13% higher than lab values at start of test, 1h and 2h intervals, respectively. During the second OGTT done in September 20, 2012, the IBGStar read 7%, 6%, 12%, 13%, 32% and 46% higher than lab values at varios time intervals. The number of data may be few but from this data I conclude this device to be consistently higher than lab results, with a couple of readings out of the +/-20% FDA range. I would like to switch to a more accurate reader and would appreciate your recomendation. Thanks again.

    ReplyDelete
  16. Thanks for you kind words Manoel. I bemoan the 20% accuracy here: http://www.everydayupsanddowns.co.uk/2012/04/meter-accuracy-narrow-window.html As to recommendations... there are so many factors to consider for each individual I would hesitate to make suggestions for anyone but myself.

    I've used relatively few over my D 'career' and usually run a few tests to compare before the switch so that I can get a feel for any likely change in general trend (whether a little higher or a little lower than I'm used to). I also try not to run more than one meter at a time so that those general differences can be accounted for with target range/correction factor etc. Good luck with finding one that suits you.

    ReplyDelete
  17. I recently purchased two iBGstar meters through a reduced price offer in the US ($5 at Walgreens pharmacy). I have compared with Free Style meter. Both of the iBGstar meters read 30 to 40 mg/ml higher (about 6 to 8 mmol/L). I did a number of blood tests as well as control solution tests. There may be an increasing error for higher blood sugars as mentioned in this post. For me this makes the meter unusable.

    ReplyDelete
  18. Further information on iBGstar.

    There is a table of reading comparison in the FDA approval for the iBGstar meter. http://www.accessdata.fda.gov/cdrh_docs/reviews/K103544.pdf

    Summary of the FDA study on 15 meters and many tests:

    Actual iBGstar
    50 56 mg/ml
    108 136 mg/ml (consistent with most observations in this post)
    270 360 mg/ml

    So there is a sliding scale in correcting the iBGstar reading that seems to be reproducible. The relation between the two values is a straight line following:
    ActualBG=9.53+IBGStarReading*0.724 (in mg/ml)
    I'm not sure why they can't program this linear correction from the FDA study into the device. Again this is valid for 15 iBGStar meters so probably is appropriate for general use with a grain of salt.

    Roughly, take 3/4 of the iBGstar reading and add 10 to get the actual BG.

    ReplyDelete