Diabetes Wise

Guardian RT vs Paradigm REAL-Time System

The Paradigm REAL-Time system really is the Guardian RT all grown up. My experience on switch day notwithstanding, I'd be lying through my teeth if I said it didn't have big advantages. A whole lot of advantages.

But there is one major drawback:

I can't hear it!

The alarms on the Guardian RT are sirens, without understatement. Low or high alarms going off at night have woken people sleeping in the next room to me! Added to the fact that you can have 'beep' and 'vibrate' alerts simultaneously (and this is a hearty up-'n-at-'em shake-itself-onto-the-floor kind of vibrate) and you'd be hard pushed to ever miss an alarm or alert.

In fact, I didn't.

By contrast, the 722 alarms are very quiet. Even the vibrate is pretty mild, and there is no option to have both alert types simultaneously. Whereas the Guardian RT awoke me as soon as an alarm went off, the 722 has taken between 10 and 18 minutes to wake me up at night. Yes EIGHTEEN minutes. A fast moving low could do a lot of damage in that time.

I'm still working on solutions to this, but it is early days yet, since I'm (fortunately) not getting many overnight alarms right now.

It has led me to think about the names of the devices and I couldn't help but wonder whether the Paradigm REAL-Time System is simply more of a Real-Time monitor providing data, whereas the Guardian is just that: A Guardian. Something extra to watch over you and wake you from your sleep should the need arise. It is daft that there should be any difference in the use of the systems. But then, it strikes me as pretty daft that one should be so LOUD and the other so quiet.

So, what are the other major differences?

To begin with the obvious:

Carrying one device vs carrying two

If you are using a pump and a continuous monitor it makes sense to try and combine them. Even leaving aside the bulk an unattractiveness of the Guardian RT monitor, to me it is a great advantage to only have one unit to clip to my clothing and carry around, one unit to worry about damaging or losing, one unit to explain to airport security personnel. And one unit that is the exact size of the pump I would wear anyway, no less.

Trend Graphs and Trend Arrows

The graphs showing three hour and twenty-four hour trends are great. I'd got used to drawing a mental graph with the last dozen or so numbers from the Guardian RT, but to have it there on the screen without having to think about it is great. A word of warning though: the screen size is very small - the same as any other Paradigm pump. I think the samll size of the graph can give some false impressions about exactly how stable the glucose level is since a reading of 6 mmol is not that far removed from one of 9 mmol giving the impression of a much flatter line than may be there in reality.

This in itself could be a reason for the inclusion of the trend arrows, but to be honest, I don't actually find them much use. Perhaps this is because I got used to mentally calculating the trend whilst using the Guardian RT, or perhaps it is because my levels have been pretty stable anyway over the last week (I have yet to see a double arrow, indicating a change of more than 2mmol in 20 minutes, in either direction). However, it strikes me that the trend arrows seem in themselves to be very much a response to single values, which is obviously ridiculous. They are basically a comparison of a pair of values taken twenty minutes apart. I've so far seen a down trend arrow when the current reading is actually higher that the previous one, and vice versa. Take the following as an example:

15:40  7.1
15:45  7.5
15:50  7.2
15:55  6.8
16:00  6.2  *No arrow*
16:05  6.4  *One down trend arrow appears*

The arrow appears at 16:05 since this value is >1mmol less than the value recorded at 15:45, despite the fact that the reading has gone up since 16:00. No arrow appears for the lower reading at 16:00 since this is not greater than 1mmol different to the reading at 16:40. The biggest drop in this data series, between 15:45 and 16:00 is not considered as these two reading do not make a twenty minute pair!

I'm not knocking trend arrows per se, I just think this feature could use a little more work to maximise its value!

Calibration Warning

Both the Guardain RT and the Paradigm REAL-Time System need to be calibrated by entering a fingerstick result a minimum of every 12 hours. On the Guardian RT, the time the next calibration is due can be accessed by pressing the up arrow once from the home screen, and the time and value of the last calibration by pressing 'SEL' twice from the home screen to get to the 'Meter BG' screen. The pump would also emit a small beep and/or vibrate at six hours since the last calibration/until the next one amd the word CAL would appear on the home screen. To be honest though, I got a lot of 'Enter BG' alarms where I had reached the 12 hour mark without remembering, and these were accompnied by a 15 minute skip in data.

The 722 allows you to set a 'CAL Reminder Alarm' to alert you anywhere between 5 minutes and 4 hours before the calibration is due. It also records the date, time and value of at least the 28 most recent calibrations (my 722 currently has 28 values, but it may store more.) Remembering to calibrate is also easier since each time a blood glucose value is entered into the Bolus Wizard, to calculate a bolus dose, (so at least every time I eat) it asks you whether you want to use that value to update the sensor. Before anyone asks, yes, this is better than the pump automatically using it to update since calibrations performed when the blood glucose is changing rapidly tend to throw things out of whack.

Sensor Restart

If you haven't heard that extending the three day life of the sensors is possible, then this must be the first thing you've ever read about about CGM technology, bar the official company info!

With the Guardian RT re-using an existing sensor requires treating the old sensor exactly like a new one. So you have to disconnect the sensor and transmitter, then search for the transmitter and then reconnect the two. Bingo! The monitor thinks your old sensor is a brand new one. Drawback? You have to wait for a complete two hour initialisation. Not only does this mean a gap of two hours in the data, but it also has to be timed so that you won't get woken up to input a meter calibration at the end of the initialisation, and ideally so you don't end up needing to calibrate soon after a meal when glucose levels might be unstable.

Timing it would be easier if there was better warning of the sensor end time. If you don't make a mental or written note of the day and time of your last "new sensor" the Guardian RT will only warn you after you have entered a meter calibration less than 12 hours before the end of the sensor life. The screen which shows when the next calibration is due (see above) changes to 'Replace' with a time, when the 72 hour life will be reached before the next calibration is due. But if a Calibration isn't required or entered until 71 hours of sensor life has already passed, you could find yourself with only a one hour warning. And even then, it is a warning you have to seek out and check for yourself!

With Paradigm REAL-Time System, you can restart the old sensor as a new sensor, and provided you don't disconnect the transmitter from the sensor first, it will be ready to begin giving data again in just 15 minutes. The fact that there is only a 15 minute gap in data makes it more acceptable to delay the restart if necessary, to wait until glucose levels are stable enough for a calibration. In addition, the exact life of the sensor is recorded in days and hours on the status screen, so you know exactly when it will need to be replaced.

Synchronisation and Missed Data

The Guardian RT accepts missing a single data value. This value is then retrieved along with the next value. If two successive points are missed, the Guardian RT will begin to search for the sensor. Provided it is found within five minutes, when the next data point is due, all three values are retrieved and recorded. Three data points seems to be the maximum the transmitter can store without passing them to the monitor, so a loss of synchronisation between transmitter and monitor of greater than 15 minutes causes a break in the data data. If the transmitter still is not found after 8 minutes, the unit will alarm with  'No Sync' alarm. It is then necessary to start another search manually, ensuring the transmitter and monitor are well within range.

The Paradigm System, on the other hand, can accommodate much more missing data. You can select how long you want the unit to wait before alerting you to the loss of Sync, referred to now as 'Weak Signal' to between 5 and 40 minutes. I'm told, although I've yet to confirm this personally, that the unit can retrieve the full 40 minutes of data. This would obviously be a big plus in situations such as going swimming, as it would be possible to leave the unit well away from the water, but still look back at a full set of data later to see the impact of the swimming session.

My experience over the lat week suggests however that the 722 tends to lose the signal more readily than the Guardian RT did. This is difficult to gauge exactly, since the 722 is for the vast majority of time attached to me by my infusion set and I'm unlikely to leave it very far away from me. The Guardian RT would quite happily continue to pick up data when I was actually in the next room, whereas the 722 will sometimes show 'Weak Signal' when I happen to have it on the opposite side of my body to the transmitter. It really isn't a great problem though. As I say, I'm unlikely ever to get that far away from the 722, and the fact that it picks up so much missed data is a definite plus. So far the only days on which I have not had a full set of 288 data points recorded at the end of the day are those when the sensor has been restarted. The Guardian RT, on the other hand, would often have gaps - including, of course, those where I forgot to calibrate on time!

Snooze Features

As I explained to Bernard here it is possible to turn both the high and low alerts on the Guardian RT off, as well as select the interval at which the alerts repeat to allow for the fact that both high and low levels take time to correct, and you don't want to be bugged by alarms every five minutes until you are back in range. That would be enough to test the patience of a saint! The Guardian RT also has the rather curious 'Snooze' feature for high alerts. This can be set independently of the repeat interval to prevent you being alerted to another high value for between one and twelve hours and applies only to high levels, not lows. I did use this feature a couple of times overnight, where I really wanted to get sleep and chose to chance staying on the high side for a little longer without being woken up if my first correction didn't fix things. Setting it was easier than going and changing the interval in the setup screen.

On the Paradigm System this feature is gone. You can still choose to turn either or both the low and high alerts off and each one has what is referred to as a 'snooze', which is the same as the repeat interval on the Guardian RT. The 'High snooze' can be between 5 minutes (??! - Anyone discovers an insulin that works that fast to correct highs, send some my way will 'ya?!) and three hours. The 'Low snooze' can be set to between 5 minutes and 1 hour. If you want longer than that, you'd have to turn the alert off. I guess the advantage of the 'snooze' on the Guardian RT is that the alerts will resume themselves after the set time, rather like the resumption of basal delivery after a temp basal of 0.0 u/h vs remembering to restart the pump after a suspend.

Transmitter Size

Tiffany has written here about the problems of concealing the transmitter under fitted clothing and this isn't helped by the fact that the 522/722 transmitter is actually bigger than the Guardian RT transmitter. It is slightly bigger all round and, significantly, around 30% thicker. This may be extra waterproofing. It may reflect that extra data storage the Paradigm transmitter seems to be able to handle. It may also be explained by the fact that the transmitter now emits beeps when connected to or disconnected from a sensor, and so must contain some sort of speaker. Overall I think that if you're prepared to accept the size of the Guardian RT transmitter, the difference is not that big. I sometimes find the thicker profile more uncomfortable if I end up lying on it, but to be fair, the clothes issue is not a biggie for me as it doesn't seem to show, unless you're really looking, under any of my favourite outfits!

Software

The Guardian RT comes with 'Guardian Solutions' software, which enables you to download all the data from the Guardian RT onto your computer. There is nothing flashy about the software. It gives you a graph for each day, as well as a 21 day overlay (but no options for different numbers of days, or to restrict it to weekend or weekdays only, for example). Beyond the daily high, low and average reading, there are no other stats either. However, the graphs are extremely useful for spotting trends (See here for an example) even without any other statistical analysis.

And I really miss those graphs right now.

I don't currently have any way to download data from my 722 pump. This is partly because I still have it on loan, and was not given a download cable. However, even if I had the cable, I'm not convinced I could use it. The Medtronic Carelink web-based Therapy Management System is not available to users outside the US or Canada. But aside from Carelink, there doesn't seem to b any other software to use with the 722 - or at least Medtronic haven't been able to tell me that there is.

I think this is a serious drawback. The on-screen graphs are great, but their small size limits their value, as does the fact that you can only see back to the last 24 hours, so you can't easily compare consecutive days, or any days for that matter, to spot patterns. Without the ability to manage all the information that continuous monitoring gives you, it loses so much of its value.

---

So there you have it. I recognise that much of this information is probably somewhat redundant to people making a choice of monitoring system, as I understand that the new improved version of the Guardian RT - the Guardian REAL-Time System is about to become available. Since I still have decisions to make about my long term choice of system, I've been looking into this one with interest. So for the next instalment: Upcoming features of the Guardian REAL-Time System!

And after that, an end to the obsession with CGM technology??!

Continuous Monitoring: In a Sentence

This post comes from the category of "posts that never quite made it..." since the information in it and the intention to knock it into shape, has been hanging around for quite a while, but somehow it never quite happened. The original title would have been "Guardian RT: In a Single Sentence".

I guess, it has never been my intention to discuss the Guardian RT or continuous monitoring in great depth since so many others have already made an admirable job of this task: Life After Dx (of course!) discusses use of the RT, Tiffany over at Candid Diabetes has done a fantastic job of describing her experiences with the Medtronic 522 integrated pump and sensor, and Amy at Diabetes Mine  is recounting her experiences with DexCom, following in the footsteps of the Insulin Factor's Matt Vogel. In short, there are plenty of capable voices out there.

I still wanted to take the opportunity though to try and sum up my perspective on the Guardian RT, which I used for just short of two months before my switch on Monday to the 722. This really is my perspective and my opinions and experiences. Continuous monitoring is definitely an area in which your mileage may vary. But, if I had to use just one sentence to describe this thing it would be:

The Guardian RT is a device that gives you prompts to test your blood sugar.

There is obviously more to it, but ultimately that, for me, is what it all boils down to. The Guardian RT has been able to replace some of the prompts that my body has no longer been giving me since my experiences with severe recurrent hypoglycaemia.

Low and High Alarms, or a series of readings that indicate I am trending sharply up or down are things that prompt me to do a traditional finger stick and treat as appropriate. The Guardian RT doesn't have the on screen grpahs and trend arrowns that both the 522/722 and Dexcom have, but scrolling back through the numbers and drawing a mental picture of where my blood sugar was heading was a frequent prompt to act. I couldn't tell you how many times in the last two months I've tested at times I would otherwise not have, particularly in the absence of prompts from my own body in the form of hypo symptoms, and I have found lows needing treatment and highs worthy of correction.

I've tended to find that the Guardian RT is spot on with my finger stick readings more than 90% of the time when the reading on the Guardian lies between 5 and 8 (90 - 145). Even at lower levels the correlation is great, provided that the lower levels have crept on slowly. Otherwise there will tend to be a lag of around 10-15 minutes.  When my blood sugar is above 9 (160), the Guardian RT tends to be almost permanently off the number. These things reinforce that fact that the Guardian is best as an indicator of when to perform a traditional test. A conservative low threshold - around 5 (90) - leads to a test before the Guardian can get too far behind, and hopefully before blood glucose can get too low. Even if the Guardian and a fingerstick are spot on at that time, looking back over the trend on the Guardian and checking the next couple of readings, along with knowledge about active insulin on board and recent food and activity can help decide how to proceed. And knowing that a number of 9 or more is unlikely to be spot on is a definite prompt to test.

But, for anyone who finds that somewhat disheartening, there is more to it: The discrepancy with high numbers appears to be consistent. That is, the discrepancy is almost always the same. Between 9.5 and 11 (170 - 200), it is around 2mmol (36mg/dL) too low, above 11 (200) around 3-4 mmol (55 - 72mg/dL) too low [Bear in mind I haven't has that many readings in this range] This is really significant, because it means I can predict what my finger stick reading is likely to be based on the Guardian reading. The finger stick is needed to check, but the numbers aren't quite as useless as they may appear.

In fact, a good way to view the RT, or any interstital fluid continuous data, is removed from the numbers entiriely. The scale it gives out could be in pink elephants for all it matters, as long as you know what the numbers mean for you, and you know how you need to act and when you need to test based on them. Also, in my experience, even if the numbers are a little out of whack, the trend is almost always absolutely spot on.

I'm happy with that approach right now.

I totally agree with many of the statements out there - that continuous monitoring is not a magic wand to transform control. It is simply another tool, another layer of information that we can use in our quest to stay healthy.

To me, it certainly gives peace of mind. It is another pair of eyes to watch over me... prompt me to act. And it clearly shows patterns in a way the snapshot finger sticks can never do.

And the drawbacks...

Of course there are plenty.

In the case of the Guardian RT and the DexCom there is the inconvenience of carrying the bulky monitor around, and the risk of losing that monitor if you set it down somewhere away from home.

Continuous monitoring gives an awful lot of data and as Allison  has pointed out, there is a tendency to get information overload. Some of the data may even be very surprising. I was totally floored by how little my blood glucsoe spikes after certain foods, having always assumed a spike was there. It takes time to shift the ingrained mindset for dealing with data from finger sticks, even the large number of fingersticks that many people seeking good control perform. There is a very real risk of becoming somewhat obsessional about the numbers and continually checking them out and following them. This in turn can easily lead to frustration when the flat line you aspire to in your mind's eye fails to become a reality. (Get real here: even people without diabetes don't get completely flat lines!) Not all the actions made based on continuous data will turn out to be the right ones and sometimes it is easy to regret intervening in a downward trend when it subsequently spikes up, or vice versa. But then, not every decision made based on any data will be correct.

On the other hand, there is a risk of becoming over dependent simply on the alarms, instead of listening to your body and remembering to use common sense and judgement. Oh, and checking out those trends!

Plus... the biggie... despite my discussion of accuracy issues above, things can go awry from time to time and the readings will go out of whack. This may be temporary - related to rapid changes in blood glucose levels not matched by changes in the interstitial fluid, or related to performing calibrations at a time when things are not stable. (I had a tendency in the beginning to want to tell the Guardian that it had got things wrong. "Damn you, stupid monitor. Look, my blood sugar is actually thiiiiis high." This is a mistake. It only confuses the unit more and sends things further out of whack.) Sometimes the sensor in use just seems to be a bit screwy. Either way, you know the solution: test,test, test.

There is no hiding from it.

Continuous monitoring won't reduce the number of finger sticks needed.

Hell, it may even increase the number performed!

(Coming up: Watch out for my High School Science essay standard 'Compare/Contrast'  on my experience of switching from the Guardian RT to the 722)

Is the Grass Not So Green?

Yesterday I swapped the Guardian RT for a Medtronic 722 pump. This is a temporary change as I currently only have the 722 on a finite trial arranged between my hospital and Medtronic. Even so, I've been looking forward to this. Since discovering at first hand the value of the Guardian RT and continuous monitoring I've become a total addict. And I was looking forward to a bunch of features that the 722 offers over the Guardian RT.

Like 3 hour and 24 hour trend graphs right there on the status screens.

Like the ability to restart an old sensor after the 72 hour cut off with only a 15 minute skip in data, rather than a full two hour initialisation.

Like only having one unit to carry around, which apart from obviously being easier will also surely simplify matters on flights I am scheduled to take.

But had I not already been a convert to continuous monitoring, had the 722 today been my first introduction, I think I might have thrown it out of the window and given up. And even as it was, I certainly began to wonder if things weren't better back with the Guardian. If the grass wasn't greener over here with the new pump.

After starting up the new sensor on the pump, I patiently waited for the two hour initialisation to end, and the pump to ask me for a calibration fingerstick result.

'Beep beep beep CAL ERROR'

Ok... let's try again. 15 minutes pass...

'Beep beep beep CAL ERROR'

And then: 'BAD SENSOR'

Gaaaah... Admittedly there was a likely reason for this: Unfortunately the timing of picking up and starting on the new pump, plus the timing of an important meeting I had on in the afternoon that meant I couldn't delay or skip lunch, the first calibration request fell just an hour after I finished eating. Cal Errors are the pump's way of telling you that it is confused because the data it is receiving from the sensor doesn't tie up with the data you've given it. Bad sensors can cause Cal Errors, but far more likely is that the blood glucose level is changing rapidly and so does not correlate well with the interstitial fluid glucose measured by the sensor.

No matter. Without disconnecting the transmitter, I selected 'New Sensor' to restart, and just 15 minutes later (i.e without the full two hour initialisation) it asked me for another fingerstick.

'Beep beep beep CAL ERROR'

Ok, I was annoyed now, because this fingerstick was within 0.5mmol of the previous two. So over the previous half an hour or so my blood glucose hadn't been changing that rapidly. I angrily punched in the value again.

Finally, a few minutes later I started getting readings and seeing the trend graph appear.

Great.

The values on that graph? Not so great.

Stress of my meeting notwithstanding, my blood sugars yesterday afternoon were horrible! I consistently hovered around the 12-13 (215-235) range. The pump painted an even bleaker picture though, shooting up to the graph limit of 17mmol (306) and staying there. So much for seeing trends. All I was seeing was a thick black bar dragging across the top of the screen.

When my meeting was over I took what I guess with hindsight was a bit of a rage bolus, and fell from 12.5 (225) to 4 (72) in under an hour. Whether it was the rapid change or not, something was irking the 722...

'Beep beep beep BAD SENSOR'

"Ok... chill, I can deal with this" I said to myself. Restarting as I perused the menu at the restaurant we'd settled on for dinner.

Right around the time my first margherita arrived 'Beep beep beep BAD SENSOR'. And to make matters worse I kept hearing little alarm beeps but with no accompanying alarm message.

I guess having already used the Guardian RT was something of a benefit. If I hadn't experienced a couple of bad sensor alarms with that, and realised that the sensor is usually anything but bad, I would almost certainly have pulled the sensor out at this stage, if not long before.

And probably followed that up by throwing the whole assembly against the nearest wall, or through the nearest window.

As it was, I switched the sensor feature off, but left the sensor in place, enjoyed my dinner and a hot chocolate fudge brownie sundae for dessert.

It was only when I got home and examined the sensor carefully that I realised that somehow the transmitter has come partially disconnected from the sensor. The "alarm beeps" I'd kept hearing were actually coming from the transmitter itself as the end engaged and disengaged slightly with pressure. I've never had this experience before, and the connection was tightly covered with IV300 so I still don't understand how it happened. The transmitter beeps are a new feature not present on the RT transmitter, so I wonder if this has been a problem for others and those beeps have been added as a warning system.

Later, after reconnecting and initialising, and allowing enough time for that sundae to quit messing up my readings, I was good to go. Calibrated first time and normal service resumed. Since then I've had great accuracy and I'm very much enjoying not having two units to carry around.

Oh yeah... and checking out those graphs. This is what I like to see...

Perhaps I'll reserve judgement on the colour of the grass for a little while longer...