Hacker News new | past | comments | ask | show | jobs | submit login
The Sad State of Diabetes Technology in 2012 (hanselman.com)
161 points by dennisgorelik on June 17, 2012 | hide | past | favorite | 63 comments



The CPAP industry is as bad or worse. Here are the drivers I see:

1. FDA -> Medicare -> Private Insurance. They set the reimbursement codes and what each code will pay. On the internet, you throw up a website and the customer decides if your approach has value. With a pay per code system, anything outside of the sanctioned approach is either puts you at legal risk ("you aren't allowed to do this, it is not sanctioned") or financial risk ("thanks for the better way, but it disrupts money flow X in the system, so we're not giving you a reimbursement code").

2. The "Insurance is better" customer bias. The average consumer of healthcare expects to show up, slap their insurance card down and receive treatment. This abstracts them from being the direct payment entity of the services they provide. If you piss off the patient but provide the service to code, then the insurance company will pay. If you please the customer but go outside the payment structure, the insurance will not pay. Businesses get paid.

3. Regulatory and bureaucratic haze. Go to a CPAP trade show some time, it is a different world. No one is focused on improving technology and services for the CPAP user. They want to get together and lobby to prevent reimbursement cuts. They want to make sure the business they run stays good with the new regulations and can survive the next wave of audits. They do not shop machines to find what works for the patients, they have too many other stakeholders to please: referring physicians, insurance payors, what their front lines people have experience setting up. They are an extension of a giant that has far more control and say about their own business than they do. They fight back or buck the trend, they are literally audited out of business or worse.

4. Talent Deficit. There are very few places to innovate and if you find a line and do it, you have to fight the system hating you for doing it on top of the normal startup pains. It takes very special people to want to walk that road with you. For the bigs in the industry, mfgs and large insurance based sellers of equipment, their HR departments are hiring for technical positions based on resume check boxes and history with microsoft. The products they put out are heavily windows, heavily compliant, fearful of making waves with any stake holder. You can't build a better one and win because you won't get paid and don't have a network to leverage. People who could come in and really make bold moves get on the ground, see the reality and leave to do something easier.


"2. The "Insurance is better" customer bias. The average consumer of healthcare expects to show up, slap their insurance card down and receive treatment. This abstracts them from being the direct payment entity of the services they provide. If you piss off the patient but provide the service to code, then the insurance company will pay. If you please the customer but go outside the payment structure, the insurance will not pay. Businesses get paid."

THIS. Combine this with the fact that most consumers receive insurance through their employer (and expect this too) further abstracts them from being the payment entity.


I've had Diabetes since I was eleven years old. For the last seventeen years. To be in good health, every day I have to:

* Measure my blood sugar at least six times a day

* Take at least six shots of insulin

* Always think what I eat. How many grams of carbohydrates is in the stuff I put into my mouth.

* I have to be very careful with alcohol. I've once woken up by an emergency crew while having a hangover. Never again. (Weed is much better though.)

* I have lots of small annoyances which require medical care every now and then.

* Influenza is a catastrophe - it will take ages and my blood glucose is not going normal easily.

Try to forget any of these, and bam your general health is at risk. It's not so easy all the time to be fully with your body and take care of yourself.

What I would love is a bloodless way of measuring my glucose, a way to get the results to my iPhone and a way to calculate the amount of carbohydrates I have in front of me. No, a database of different foods is not enough. I have to do it 4-5 times a day, remember.

When you have to do something many times a day for the rest of your life, it should be as easy as possible.


Have you thought about a pump (no shots, more accuracy) and a DexCom CGM (only finger stick a few times a day, pretty accurate). I complain in the post but I will say this, of you're serious about staying alive, you MUST get an Insulin Pump.


MD here: (disclaimer - I'm A doctor, but not YOUR doctor, so ask YOUR doctor). Seriously consider a pump. Seriously.


I've always thought the pump is not for me. It seems to be pretty big and you always have to carry around it with you. My life is still pretty fast paced and sometimes that kind of machinery is just an annoyance to get along with. The shots are not the problem, the everyday need of full control of your life is.


It's smaller than the phone you already carry around. It's mathmatically and demonstrably superior to shots. They have integrated CGMs now. Ask around, the folks with the fastest paced lives (diabetic Olympians, marathoners, world travelers, etc) ALL have pumps.

Unless you have a 6% HA1C on shots, get a pump and live decades longer.


My wife is Type 1, and was on the pump but went off of it after about a year. Says she feels in more control without it, and that it gave her a false sense of "security." And she has had better results since, and more stable sugars.

I don't know, obviously, not being diabetic myself, but I will say she went on the pump really soon after being diagnosed and didn't really have a chance to establish a routine first, and that probably played a large part in her doing better on shots right now. I've been encouraging her to start thinking about using the pump again, as I do feel that, used correctly, as I believe it is a better method, but she has to do what she feels works best for her.

Being a developer myself, the idea that she has to use a custom cable and custom software to manage the data she has in her meter drives me a bit nuts. But the meter industry is, IMO, full of companies that don't really want to help users, they just want to get rich off of diabetics. I know thats a bit cynical, and I'm sure there are companies that sincerely care as well, but it's how I often feel about the meter industry.


Will back up Scott here -- while not an olympian, I rowed for 4 years on the pump and have biked x-country w/ it.

I seriously doubt if I would've been up to doing either while managing injections instead of button presses.


Thanks everybody, I'll talk with my doctor tomorrow when I have an appointment.


I used to think the exact same thing, somehow I think I was really just afraid that someone would see thing thing embedded in me if I ever had my shirt off. When I finally got a pump I realized that it simultaneously keeps itself out of the way while allowing me quick access to my (approximate) levels. I can give my self insulin mid sentence in a board room meeting. I can see my levels dipping and get a lifesaver before it sets in. The worst that ever happens is that someone asks what the hell I am doing with a pager.

In short, the pump CGM combo it is the best thing I have ever done.


I wrote them off as too experimental and bulky when I was first diagnosed just over 10 years ago. I guess technology has advanced since then. I'll make an appointment with my doctor to chat about it. Thanks for the reminder to re-evaluate.


Totally. I don't even care if someone sees it and I'm happy to explain if they do. It's a lifesaver.


A pump does not remove the need for shots and as for the accuracy, well that is up for debate. You still have to learn the bias of the readings it generates versus the manual test. It is a lifestyle change from shots, you will be moving the point of insertion quite a bit and some sports may be better served with a shot regimen.

There is one nice feature of having a pump, with an always attached device you can get warnings when your asleep which can save your life. Many of these devices will either beep or vibrate or both when you enter a danger range. Even if it does not alert you it could alert whomever your sleeping with; even pets will take notice and possibly wake ya.

Last note, most insurance plans expect the pump to be replaced after a couple of years so find out up front how many years before you should expect to replace it.


You DO have to constantly calibrate the pump though. It's not a perfect solution.

I am astounded there is not a close loop system yet.


Are the papers here [1] relevant? I sat in on this conference session as a student, and don't have access to the full papers to check my recollection. As I recall, most of the papers revolved about the theme of closed loop control of blood sugar.

At the time this was presented, the results were billed as the biggest advance in Intensive Care in recent history. The authors claimed to have hard numbers, showing that closed loop control of blood sugar for ICU patients was delivering (something like 40% if memory serves me???) fewer deaths for ICU patients. I'm not in the ICU area, so I don't know if the techniques outlined in the paper made it into clinical practice or real products. If so, you'd think it would trickle down into diabetes management. As a non-diabetic, am I completely misunderstanding the problem?

The International Federation of Automatic Control (IFAC) is the peak body for control research, so there's a good chance you will read about any new closed loop system there.

[1] http://www.sysid2006.org/Wednesday.html#web4

---

Edit: A couple more links from my poking around. Might provide some reading if closed loop control of diabetes is of interest.

http://www.medschl.cam.ac.uk/paediatrics/pages/wilinska.html

Closed Loop Insulin Infusion for Critically Ill Patients (with pictures of an actual device):

http://www.clinicip.org/

---

Edit: Found the actual numbers here: http://cordis.europa.eu/search/index.cfm?fuseaction=proj.doc...

42% reduction in mortality, compared to the old insulin control method, which I presume is open-loop?


You mean calibrate the CGM? There's no pump calibration needed.


My relative is definitely calibrating some instrument related to the pump multiple times per day.


Ah, then they are calibrating their CGM (likely integrated into the pump, as they are likely using either an Animas Vibe (outside the US) or a Metronic Minimed Paradigm). They aren't calibrating the pump, they are calibrating the pump's CGM vis a vi their meter via finger stick.


Not an expert, but maybe he's talking about the relative changing the basal rate (correct words? The 'give me this amount continuously' rate a pump will inject) or (this is something my type I ex-gf did a lot) he mistakes some 'correction shots' after meals for calibration?


Definitely involves blood and a small oval thing with a LCD screen.


Please don't consider this as a plug, just trying to help.

Our company, MyNetDiary, provides an awesome food diary app (the only 5-star paid diet app on the iPhone). It's highly polished and uses some very advanced tech under the hood. The most frequent word in user review is "easy".

For almost a year, we are working on a special diabetes tracking app built on top of it. It does not integrate with BG hardware (would need FDA Class 1 approval), but we are exploring options.

The app will help you keep track of foods, exercise, and - with manual entry - your BG readings and insulin.

We've been doing this for 5 years and know what we are talking about - this is the best app for tracking diabetes.

It's a couple of weeks from release, we are testing release candidate. If you are interested in trying it sooner, we can provide an Ad-Hoc build for iPhone. More info: http://www.mynetdiary.com/diabetes-tracker-for-iPhone.html

My contact info is in profile.


You, and everyone with diabetes, should seriously read the first chapter of this book [1]. The author discovered he could treat his wife and son's diabetes with a major diet change, by eliminating all foods that cause a blood sugar (and resulting insulin) spike - sugar, HFCS, grains, most processed food, etc.

His assertion is that he current method of treating diabetes is based on classical mechanics, 1-in-1-out, etc., where you measure your blood sugar and try to manually maintain it at a certain level with sugar and insulin.

But ...

1) the human body is a complex, non-linear, dynamic system. It is based on and responds to signalling, genetic triggers, etc., and ...

2) you don't need to micromanage your body, it will take care of itself as long as you don't essentially poison it.

3) the old method presumes there is something wrong with your body, when in actuality, the problem is with the way your body reacts to 'modern' (post-agricultural) foods. Diabetes is a symptom of that, so cut those foods out, and diabetes goes away.

It may not apply to all types of diabetes, IANAE, but seriously, stop by a bookstore and spend 20m reading just the first chapter (maybe 2, can't recall exactly), and figure out if it's something worth looking into further.

1. http://www.amazon.com/The-New-Evolution-Diet-Paleolithic/dp/...


First of all, the distinction needs to be made between type 1 and type 2, these are two different things.

If you respect Ray Kurzweil's ideas, which I think a fair number of people here do, then I would suggest anyone with diabetes concerns (at risk, got it, etc) read his book "Transcend."

Direct quote:

"Largely because of excessive consumption of simple carbohydrates and sugary foods, the number of people with type 2 diabetes in the United States has increased tenfold in the past 35 years, and as of 2008, type 2 diabetes affected over 21 million Americans. In just the 10 years between 1997 and 2007, the incidence of type 2 diabetes nearly doubled in the United States, from 4.8 to 9.1 per 1,000 people."

The reason why Paleo is recommended is because high glycemic index carbohydrates are virtually eliminated. That certainly is not the only diet that accomplishes this. In Transcend, a low calorie somewhat Japanese diet is encouraged. That is something that is extremely difficult, behaviorally, for Westerners to strictly adapt. In my own experience it took about 3-4 years of behavior modification to do that, and I ended up medically underweight. "Paleo", at least for males from what I've seen, is fairly easy.

If you disagree with Kurzweil, well then I can't say much else, he certainly knows more about this than I do.


I realise your intentions are good, but most diabetics get the paleo advice every other day.


...and they should take it to heart.

I went from >12% to <6% A1C by cutting out carbs, taking a low dose of metformin (750mg slow release per day) and throwing in a bit more light to moderate exercise.

Credit to Dr. Bernstein (http://www.diabetes-book.com/) who is a type I diabetic himself, but really we're now hearing the same things from Taubes, Lustig, the Paleo folks and many others.

Obviously low carb diets alone aren't the whole solution. Berstein's book contains a wealth of information on living with diabetes well into old age. It's a great read for diabetics or for anyone caring for a diabetic.


You're a Type 1 on metformin? That's highly unusual. Or are you Type 2?


I'm type 2.

Just to clarify, I'm not selling notions of curing type 1 diabetes through diet. If you're type 1 you will need insulin. Even type 2 may require insulin.

The point I'm trying to make is that in both diseases (type 1 and 2) control of blood sugar is key to reducing mortality. Reducing carb intake will result in lower blood sugar rises which will require less insulin to counter. Less insulin means less chance of hypoglycemic events. By minimizing these control inputs you can balance your blood sugar in the normal range and prevent and even reverse the damage to your body from diabetes.


Yeah, but I have the type 1 diabetes, which is btw. the national disease in Finland.

One of the things that started the disease was an influenza back when I was kid. I also have the genes from my family, so suddenly I didn't have insulin in my body anymore. It has nothing to do with my lifestyles.

I've been taking care of my disease since I was a kid and believe me, I need my daily shots of insulin.


Yep, totally. I do Paleo and while it but reduced my daily insulin in half, I still use a total 20U a day. It's definitely needed. Type 1 is NOT Type 2. It's a shame, frankly, that they are both called "Diabetes."


> ... the old method presumes there is something wrong with your body ...

The comment you replied to is presumably about type 1 diabetes, an inflammatory disease that destroys the body's ability to produce insulin. You are referring to type 2 diabetes.


>Now I wait 5 seconds but we still have blood sugar strips with +-20% accuracy

This is the part that still both sickens and baffles me. I had onset at about 20, despite not looking like a typical diabetic (5-foot-8, 140lbs). 11-years ago, my endocrinologist made sure to drill into my head this margin of error, and that the margin of error was highest in situations where I would be hypoglycemic or suffering from an insulin reaction. As a diabetic himself (and for the record, having a diabetic endocrinologist is the best thing you can hope for), he was all too aware and frustrated with the limitations.

Now, a few days ago it was pointed out that the margin of error for test strips in a medical facility is much lower. The one thing I've been waiting for is to get that level of accuracy. If there's one thing I want, it's that. More than anything. I couldn't give a rat's ass about anything else.

Getting that number consistent could've sped up the timeline in which it took to get my blood sugar levels to be consistent. I'm not terribly methodical, but at the end of the day my endocrinologist doesn't complain when my HA1C results are consistently between 6.2 and 6.6, and have been for the past 6 years, despite relying only on subcutaneous injections(1) and continued dietary changes and experimentation. However, I bet I could keep it at 6 on the nose -- while "cheating" a bit more -- if I knew that a reading was nearly dead-nuts on every time.

(1) Not a fan of pumps. My mother was and still is on TPN. After seeing the joys of dealing with infections, I prefer the very rare dermatitis.


A major reason why the accuracy on blood sugar strips is +/- 20% is because that is all that the standard (ISO 15197:2003) requires:

* Within ±0.83 mmol/L of lab results at concentrations of under 4.2 mmol/L

* Within ±20% of lab results at 4.2 mmol/L or more

Companies are juggling cost, reliability and performance/accuracy, with the first two generally winning in the marketplace as key selling points.

Some devices offer significantly better accuracy than 20%, depending on the market that you are in (not all are for sale in USA, IIRC).


The industry will never let you get off strips because the strips cost a penny or two to make but sell for $1 each.

So they make millions off insurance for the strips at the expense of society and the people who cannot get insurance.

If they ever invented a $100 device that doesn't need strips they would completely destroy the crazy profit they have.

The article didn't mention the Contour USB which is completely digital but of course still needs strips.


This is a very well articulated representation of the frustration I feel as a technically minded T1 diabetic. Thank you. I dream of the day when both my meter and insulin pen transmit data in an open format over bluetooth. It is clear that any steps that can make the monitoring and analysis of diabetic care more transparent to users will genuinely save lives.

For anyone who is disinclined to get a pump, I highly recommend the Bayer Contour USB. I've had it for a week and it is _leagues_ better than anything else I've used. I was so thrilled when I discovered that with it I could simply access my readings as an SQLite DB rather than the hell of trying to communicate with and parse the data from anything from Lifescan over a bloody 3.5mm serial port.


If sugar level meter and insulin pen start transmitting data in an open format - would non-technical diabetics be able to use it?


SQLIte? That's brilliant.


>"I've been diabetic for almost two decades"

I'd say that is more compelling evidence that the state of technology is good. Lack of bluetooth may be somewhat inconvenient, but it is not life threatening...and the lack of bluetooth is less inconvenient than dialysis several times a week.

I'm not unsympathetic toward the author's medical condition, but proprietary interfaces are even part and parcel of devices as successful as the iPhone, and it's hard to see a strong medical case for adding bluetooth to a medical device...sometimes hardwiring is just a better solution for life critical applications.


Sure, I hear what you saying but even a crappy diabetic on shots live two decades. The tech hasn't kept me alive, diet and excercise has. Saying the current state of tech must be good because I'm alive doesn't hold, I'm afraid.

I'm concerned with quality of life and convenience. As the other commenter said, if it's something you are doing 10 times a day, thousands of times a year, you want it to be convenient.

Diabetes wears on you. It's the incessant nature of it that hurts.


I look at the devices and think of the HP12B calculator or 37 Signals. Sometimes, a piece of hardware is just suitable for its purpose. Sometimes, less features is better design. Its a small step down a slippery slope between adding Bluetooth and requiring it to access device features.

There's a legitimate philosophy behind not treating medical devices as electronic gadgets.


Sometimes, life critical devices have to make themself desirable to the user.

It sounds stupid on the face of it, but for a 13 years old already stuffed to death with complexes, the difference between a bluetooth device no one sees and can be checked quickly, and something with tubes running through your clothes to your pants can be like night and day.


Ditto this. When I was 15/16 and considering a pump (I got one when I was 17), I was seriously concerned about social implications. That was an extra year or two spent injecting when I should've been on the pump.

I'm willing to take some of that blame for being shallow, but not all of it. I know I wasn't the last teen diabetic w/ similar thoughts.


Turns out most girls really don't care.


At least the good ones. I did have one get all grossed out. Good riddance.


Fair point: If you want to disrupt an industry which is entrenched by the legal/regulatory system - you have to find ways around it.

Think education, health, law ...

Any more?

In fact from the list of the oldest professions only the "oldest" is free from regulation (only, for the most part, suffering from prohibition).


I wonder if there was an Open Source hardware way to do this...


There's no really good way around it other than by changing the attitudes of the culture. At present, people generally like their nanny state and want more of it, not less. So we'll all suffer the consequences in the meantime.


You would also have to change the culture of business. As long as companies are out to make money instead of contribute value the force of the state will be employed to keep snake oil off the market.


As an entrepreneur (and engineer) who has built two healthcare companies, I believe we'll get to a point where the speed at which innovation moves in other sectors will start to take foot in healthcare. But like everything in the space, it takes time. Sometimes too much time.


What changes you think will happen to make healthcare innovation go faster? And how can we help push those changes ?


My doctor visit routine goes like this: Nurse asks for my password to login into mymedtronic.com to "download my data" this costs me 20 dollars. Nurse gets access to special healthcare only charts, you know ones that are actually somewhat useful and prints them. My doctor looks at the prints and stuffs them into my file, rubs my feet and tells me to keep up the exercise. That was a 50 dollar foot massage. I don't want to think what that would cost without insurance.


Your doctor is looking for lesions - diabetic foot kills a large number of people. The ones that don't die have a leg amputated.


I know that is what he is doing, what I don't know is if that was worth my time or money. I am nimble enough to see lesions for myself and smart enough to see a doc if I get one.

But that was not my main concern...

The worst thing is that instead of using the the data in combination with all their patients to form some sort of massive dataset that could help everyone... they print a chart for my file.


Keep an eye on the guys at mySugr. I've had a few interesting chats where they've told me about the issues you come up against in the medical industry. They're diabetics, they get it and they're working around (and through!) it to make a really usable and useful diabetes app. It's German only at the moment, but check it out at http://www.mysugr.com/.


I can't believe that no one has suggested the use of RFIDs to measure blood-sugar levels! These guys already have a product that's under testing: http://www.positiveidcorp.com/products_glucochip.html

The only trouble with things going inside the body is the need to get FDA approval and the vigorous testing that goes along with that.


I'd like to see the Pebble watch serve as a way to view blood sugar levels and send commands to an insulin pump ... at one point I thought about setting up a kick starter project for this, but all the issues with bluetooth seemed daunting. Thank you to Scott for making these issues heard.


there are almost no issues with Bluetooth LE. Pebble supports it, the new MetaWatch supports it, iPhone 4S and many new Android phones support it...


I would have hoped that we would have a gene therapy to treat diabetes at this point in time. But alas, the medical industry has stagnated under the weight of bureaucracy and an overall decline of investment as investors demand the same kinds of ROIs as those seen in the dot-com sector.


http://www.echotx.com/symphony-tcgm-system.shtml Wireless cgm and non-invasive. Passing every clinical trial with excellent accuracy too. "Coming soon" though.


The industry will never come up with a cure because there's more money in chronic treatments.

At the same time, why do we even need diabetes technology? Just go on a low-carb diet. My grandma was borderline diabetic a year ago. She was getting very worried, so I was finally able to convince her to try it and now her test results are great.

Even if you're a type I diabetic, this will manage your symptoms although it won't cure you. Check out "Good Calories Bad Calories" by Gary Taubes.


I'm not sure low-carb is necessarily effective for people with serious insulin dysregulation. I'm pretty sure for Type 1 in particular, avoiding carbs is insufficient (you also require insulin injections). Personal bias: Dad has diabetes, caused by pancreatitis, and although minimising carbs led to some weight loss (not that he is/was overweight), it has no impact on his blood sugar.


Yes, type 1 is a different beast.

The latests theories point to gluten to cause a leaky gut, that causes complex foreign molecules to enter the bloodstream, and as some molecules are similar to our own, this causes the immune system to attack some cells in our body. The cells in the pancreas that make insulin are one of these.

The bad news is that if you have Type 1, you can't be cured, and need insulin for life.


Treatment = profit while cure stops all income.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: