Question regarding diabetes apps

[trvmd]

Hello everyone.

I'm not diabetic, but I am a Masters student of health informatics, coder, and doctor. As potentially my thesis, I'm building a system for physicians to remotely monitor all those self measured blood sugar readings you guys take daily. However, it's also going to work as a standalone app for logging blood sugar. This part of the system is nearly complete. I've sampled a lot of the options for blood sugar logs on the iPhone. Most are ugly, have tiny fonts and are a pain to use. Mine won't be.. I promise!

I wanted to ask you guys some questions about this type of app.. blood glucose recording apps.

My questions are:
1. Do you use smart phone apps to log your blood sugar?
2. What annoys you most about them?
3. Are the bluetooth BG monitors a pain in the butt to work with or do they make life easy?
4. How much of the statistical analysis would you like to see?


I'm going to be doing a whole lot of calculations for the physician side of things. But I know if I were testing my blood sugar that often, I would want to see all those numbers too. Things like multiple measures of glucose variability, risks of hypo/severe hyper glycemia, averages, week to week linear regression trends, trends in change around a meal, tables relating postprandial(after eating) glucose to insulin dose (for those using insulin) and lots more. I am curious if most diabetic individuals would like to see those numbers and get explanations on what they mean. I intend to have printable sheets so you can talk about them with your doctor, with references to all the research from which these calculations come.

It's not going to be some massive, confusing suite of diabetes related stuff. Recipes, carb counters, step counters and all that stuff is for other apps. What I've built has one feature, and does that one feature better than everyone else. I just fear walking a fine line between too confusing, or leaving out some interesting data that many individuals testing blood sugar daily would want to know.

Thanks for reading my post. I really, really appreciate any feedback. I'm not a company here to sell you something, just a guy who loves health care and programming and has done a lottt of research into diabetes.

 

[andcol]

Hi and welcome.

I will suggest one thing to you. Asking this group of people that take an active interest we will all say yes we want all the figures. If you ask a group of disinterested individuals who think pop a pill and carry on then they will not be interested nor will they test. In fact for T2s in the UK meters and testing is actively discouraged due to cost savings

Have a look at the app that I wrote (link in my signature) as it produced lots of data already

 

[tim2000s]

I'm building a system for physicians to remotely monitor all those self measured blood sugar readings you guys take daily.

Hi @trvmd, I think you've jumped straight to a conclusion that misses the most fundamental philosophical point that I regularly bring up in these discussions. The whole idea of patient centric care and not taking responsibility for your condition.

Fundamentally, diabetes is a condition that requires self care and self management. Far too frequently, we see people coming in to the forum who delegate the responsibility for this to their health care professionals, and this is encouraged by those HCPs. What you are proposing takes it a step further and suggests that someone doesn't need to think about what they are doing as their HCP will do it for them.

It is not possible to live with diabetes successfully without taking responsibility for your own diabetes, indeed those who are most successful cave done exactly this and I personally wouldn't want my data being sent real time to a physician to tell me what to do.

You see, each individual diabetic is slightly different, and what works for one absolutely doesn't work for all. And the only way to determine what works for you is to go through a process of recording and reviewing everything, and then making changes. The patient needs to be given the power and responsibility to do this themselves, and I'm afraid that HCPs are not very good at this. Introducing an invasive app will just make this worse.

The standard level of expectation is already that "HCPs will tell me what to do". In day to day life this needs not to be exacerbated, and what you propose does this. Please take this into consideration and read the quote in my signature whilst you consider how to proceed.

 

[azure]

I don't use any app. I test and respond if necessary. As @tim2000s says, monitoring diabetes is something the person with diabetes has to do. I cant imagine passively sending my test results to a doctor and waiting to be told what to do. How would that work? By the time they'd responded, I'd have sorted whatever issue I had. if I had an ongoing problem I couldn't sort, I'd phone my DSN and discuss the issue.

To my mind, my day to day test results are for me not my doctor. My consultant rarely asks to see them. They only ask general questions about whether I have any problems with control.

It might be more useful for people who are struggling, I suppose, but in that case I'd go for education over an app as it could only be to their benefit to be able to understand their body and be able to interpret and respond to blood tests

 

[trvmd]

I completely agree with you guys, and really appreciate the response. There is literally no disease that requires of the patient more self-knowledge and and understanding of pathology than diabetes, and I'm partly in awe and partly envious of the amount of knowledge and learning you guys do for it. It's why I've chosen diabetes to focus on.

And yes, I agree the patient should be empowered from these IT based interventions. I feel like many of the apps overlook how knowledgable a patient can be, particularly in type 1. Throwing warnings and basic **** at you contantly. But there is the other side of the coin - the patient who hardly tests their blood sugar, doesn't really want to learn and want their doctor to do the heavy lifting. I want to make it scalable for that reason. Empower the curious with everything I empower the physician with, and let those who don't want all that use the most basic aspects of the system.


Azure is correct in that SMBG is mostly for the patient to learn how to manage their disease - it's why the ADA/CDA in US/Canada can justify recommending it for NIDDM - even with mixed results from the literature (if you actually review it, you see that those studies not showing improvements tend to be lacking strong protocols and good use of the data).

But my problem is that this data isn't being used, as Azure said, your specialist rarely asks to see it. HbA1C is only available every three months. Your SMBG data is available very shortly after medication changes, and HbA1C does nothing to describe glucose variability, post meal excursion, or any risk of hypoglycemia (in type 1), among other things. What the data you guys are accumulating every day with SMBG, risk assessments can be made and targets can be adjusted accordingly. For instance, if a new type 1 had huge variability in fasting blood sugar, that is a risk for nocturnal hypoglycaemia, which can lead to hypoglycaemia unawareness. You and your doctor come up with your glycemic goals together. For some, that might be higher or lower than the ADA/CDA or UK equivalent recommend. Part of the story is missing when making those goals, and I want you and your care team to have that part of the story. But the individual I want to have that information the most is that patient who cares enough about their diabetes to learn as much as they can. Which is why I'm asking here. Because inherently posting on the internet, you guys are those patients of sufficient intelligence and curiosity and dedication. It's you whom I am targeting with the high end statistical evaluation.

Thanks again for the replies everyone.


Andrew, your app looks good. That's awesome you built it. When I'm done in a month or two, I'd love some feedback from you all.

 

[tim2000s]

But there is the other side of the coin - the patient who hardly tests their blood sugar, doesn't really want to learn and want their doctor to do the heavy lifting.

But therein lies the problem. An app that does this doesn't help them either. If they aren't motivated to record blood glucose, they really aren't going to use one app over another. What you have to do is work out how you get them to test blood glucose, and the fact the data goes to the physician isn't going to make that change. The reason they don't test is they aren't curious and they don't like it/don't want to accept it. The curious tend to end up in places like this.

If you really want to approach these guys, you need to capture the data from the glucose tests and transmit it directly, which means you need to work with someone like Dario, Inrange or Agamatrix, who have technology to get the reading straight through to the cloud. Even then, the amount of data you will have is sparse.

Have you also considered the HCP position for this? There is a huge amount of data for one person to try and review. Your Doctor's interface needs to really be a dashboard alerting to patients with problems and drilling down into what those problems are. If you have attended a diabetes clinic, you'd see that the routinely appointments over-run and that it is extremely difficult to make time for anything other than the patients on the day, so the person responsible for care needs to find this as easy and unintrusive to use as possible.

For instance, if a new type 1 had huge variability in fasting blood sugar, that is a risk for nocturnal hypoglycaemia, which can lead to hypoglycaemia unawareness. You and your doctor come up with your glycemic goals together. For some, that might be higher or lower than the ADA/CDA or UK equivalent recommend. Part of the story is missing when making those goals, and I want you and your care team to have that part of the story.

Nocturnal hypoglycaemia is only picked up if it is being recorded, and having experienced it, I can tell you that I was not likely to be recording anything - I would wake up the following morning with a raging hangover, and normalish blood glucose levels. Equally, the recommended blood testing regime, even on MDI, is not that likely to throw up enough data to showq you the variation that might lead to the idea that nocturnal hypos are occurring. Making a new type 1 dependent this early on in the relationship with the HCP is a very bad thing. I agree with Azure that at this stage they need appropriate education to keep going and understand their condition, then use the healthcare network in the way it was designed - as a pull mechanism, not a push. And education is definitely the key to that.

At every clinic appointment, the patient is expected to provide a meter to allow download of the data. If the meter is not there, that is usually a deliberate action (considering that all T1s would be expected to carry it with them).

Pump users are expected to upload to Diasend or Medilink on a regular basis as part of their usage contract, and dependent on the pump this includes either CGM or Blood Glucose data. Failure to comply results in risk of the pump going. I'm not clear which part of the story you see as missing in terms of what the HCP sees?

Don't get me wrong, I think the idea of being able to provide data back to the HCP makes sense, in certain very specific, circumstances. It needs to be patient rather than HCP driven as well. To pick up the kinds of things you are talking about, you really need to link it to something like Freestyle Libre or CGM, as that's the only way you will see the data in enough detail to make the call about the issues that may need to be picked up. If you hadn't noticed, I've spent a fair amount of time thinking about this one, prior to your post, so apologies for the essay. .

 

[azure]

"But my problem is that this data isn't being used, as Azure said, your specialist rarely asks to see it. HbA1C is only available every three months. Your SMBG data is available very shortly after medication changes, and HbA1C does nothing to describe glucose variability, post meal excursion, or any risk of hypoglycemia (in type 1), among other things. What the data you guys are accumulating every day with SMBG, risk assessments can be made and targets can be adjusted accordingly. For instance, if a new type 1 had huge variability in fasting blood sugar, that is a risk for nocturnal hypoglycaemia, which can lead to hypoglycaemia unawareness. You and your doctor comeup with your glycemic goals together. For some, that might be higher or lower than the ADA/CDA or UK equivalent recommend. Part of the story is missing when making those goals, and I want you and your care team to have that part of the story. But the individual I want to have that information the most is that patient who cares enough about their diabetes to learn as much as they can. Which is why I'm asking here. Because inherently posting on the internet, you guys are those patients of sufficient intelligence and curiosity and dedication. It's you whom I am targeting with the high end statistical evaluation."

Maybe I'm missing something but I still don't see the point. For most Type 1s, medication changes aren't prescribed, they are made by the person with diabetes in reaction to their blood test results. If I have an insulin to carb ratio of 1:10 for my evening meal, but always go low two hours after, I change my ratio. I don't need an app to tell me to do that, or to,point out I'm going low - yes, I know that and I'm going to sort it. Then intest to see if my chnage has worked. If I get good two hour results, then it's worked. Again, I don't need an app to tell me it's worked because it's obvious.

For the non-compliant, an app is unlikely to encourage them to,test more. If it sends results to their doctor, then that might actually DIScourage them from testing. If the problem is ignorance about how to manage their condition, then they need help and education somthat they can take control gradually. If someone was really nervous about doing that, I imagine they'd be on the phone to their DSN rather than 'trust' an app.

Finally, I can't imagine my consultant or DSN would have time to,plough through loads of blood test data. The day to day responsibility of diabetes rests with the person with diabetes.

I didn't mean any of that to sound at all rude I think it's good that you're thinking of ways to help. It's just don't see that any app like that would be that useful, in all honesty and in my opinion

 

[trvmd]

No apologies needed, I like this sort of discourse. In fact that why I came to ask, I'm still working on getting a clinical consult to give me some of the information you just did.

Yes those using CGM and already uplinked devices would not be my target audience, as they already have a mechanism for this. My target is those HCPs and patients who don't have this option available, I want a cheap, easy system that can be implemented on the fly with little participation from either party. The logging app I think accomplishes this, as actually recording a reading takes less than a second with how I have it set up. The physician interface isn't built yet, but I intend to make any review of the data require about 10-15 seconds per patient, and can be set up by the medical assistant. Only when it needs review or review is requested will it be presented. And yes, I have worked in diabetes clinics - exclusively type 2 however. The only mention of the SMBG data was "how are your readings at home".

Nocturnal hypoglycemia can be predicted, so can hypoglycemia unawareness. And extensive literature review has shown some of the factors in SMBG patterns that can be used for it. I'm digging up the references. While I agree that teaching a diabetic to rely entirely on their physician to monitor their blood sugar is the complete opposite of what diabetes care is all about, leaving this data out of the equation is not in anyones best interest. A big part of care is setting glycemic goals, and the recommendations from the guidelines are vague, with blanket recommendations and then suggestions to personalize based on things like life expectancy, risk of hypoglycemia and a few other things. I'm of the opinion that those risks can be quantified using the SMBG data, and thats what my goal is.

Thanks again for the feedback.

 

[tim2000s]

It's an interesting one. Most of the literature relating to predicting nocturnal hypoglycaemia is dependent on fasting SMBG. This is what we often refer to as basal testing, and I'd agree that it is something that shows the risk, but when you do it it also shows what you need to do to change. Likewise SMBG done post prandially following an evening meal and before bed also gives indicators. You are still reliant on a user doing that.

I also don't believe that it takes a second to enter the values. Opening the app takes longer than that.

Much of the current discussion in diabetology circles is about whether glycemic goals are the right approach. There is a view that simply "better" might be more beneficial, with an overall target being that which is known to reduce the relative risk of complications. Very few, if any, current diabetic specialists will advise an Hba1C above 7.5% except where there are exceptional circumstances. The guidelines are not vague in this respect. There is a very clear reason why the numbers selected are chosen.

Finally, as an anecdote, my Diabetes Specialist GP was working till past 10pm just catching up on paperwork. I suspect the clinical consultants are too. I'm not sure they also want this overhead. Diasend already provides a dashboard when the user uploads data from the clinic provided device in the clinic. Why do they need another way of doing this?