Telemedicine

Let me say that I am not a technophobe - I cannot be since my income comes from writing about new computer technology. However, in my day job I tell people how best to use the new technology, rather than simply fawning over the latest offering from big software or hardware manufacturers. My approach, which is to tell my readers when a product is not worth using (or how to get the inferior product to work), has not always made life easy for me - a few years back I was told by a product manager at Microsoft that I would never again be invited to test a Microsoft product, but in spite of that, I have.

I do think that technology has a role to play in managing chronic conditions, (in particular, in my own case with type 1 diabetes) but we have to be careful about what we use technology for, and don't simply replace common sense with a new gadget. As a born cynic, I usually regard the introduction of a new gadget to be simply a way for a company to make profits from the sick, so the gadget manufacturer has to try hard to convince me otherwise. Here are a few thoughts about telemedicine in our new NHS.

The first thing we have to bear in mind is that any telemedicine solution should be available to all. There is no point in choosing an extremely expensive piece of kit - even if it has all of the bells and whistles on it - because few people will be able to own the kit and benefit from the technology. There is an argument that the price of the kit will come down eventually, but you have to make sure that the starting point is low enough to get enough people initially using the kit. For this reason I reject solutions like the iPhone, iPad (and other tablets), Windows phones and Android phones. Smartphones in the region of £300 or more are simply too expensive: people won't buy them, and if the NHS buys them then there is the problem of patients losing them (or worse) generating a black market in "mislaid" smartphones.

If the device is to be used all the time then it has to be one which you, personally, can afford to lose. I have an £80 Samsung touch screen phone, I can afford to lose this, so I am happy to have it in my pocket all the time; I cannot afford to lose a £400 HTC Windows phone or a £600 iPhone, so I do not carry one. If the device is so expensive that I have to leave it in a padded case at home, then it will not be used regularly. Even the bargain basement phones available now are significantly more powerful than the first generation PCs, and yet without the first PCs we would not be where we are now.

The application should only be regarded as useful if the patient finds it difficult to live without it, but this also means that the application (and the device) must be freely available so that if they need to get a replacement they can. For example, a few years back I forgot to take my blood testing kit on holiday. These days blood sugar testing is a vital part of controlling diabetes, so a week away (and one when I was expecting to do a lot of hiking, as well as eating out in country pubs) from my blood sugar meter would be catastrophic. So, as soon as I could, I went into a local chemist where there was a display of meters and bought one for about £25. Blood sugar meters are usually sold at a loss (in fact, most are given away) because the manufacturers make their money from the test strips (in my case, each strip costs about 30p), so in this situation I chose the meter that came with the most number of "free" strips. Similarly, if I lose my insulin pen, I can go into any chemist and purchase disposable insulin syringes. The extra cost on me is acceptable - I don't want to have to pay for syringes or blood test strips all the time, but it is important that the cost is low enough (disposable syringes, a meter with "free" strips) for a one-off purchase, if necessary. So a device used for telemedicine must be freely available, and simple to set up.

The next thing to consider is features. The device has to be usable by as many people as possible. We are talking about telemedicine here, so the people using the device will be patients and many will have poor eyesight, poor co-ordination or impaired touch. A diminutive phone may fit nicely in your shirt pocket, but it is useless if a patient consistently presses the wrong button because they are too small.

I'm in my mid-forties and I have had cataract replacement, so that means that I have less than perfect eyes. I get extremely irritated when I am asked to evaluate some software written by a 20somthing with perfect eyes who's chosen to use 6pt light grey text on a slightly darker grey background. I get even more irritated when I find that there is no way I can change the colours or text size. I am not unusual, and whenever I bring up such complaints with colleagues who do similar work, they always agree with me. Yet the software vendors still seem to insist on only employing 20somthings with perfect eyesight who write software with small text and low contrast.

As far as I am concerned LCD displays are useless, the contrast is too low, but in the past all blood sugar meters have had LCD displays so I have always chosen the ones with the largest display to give me some change of reading what it says. Recently I bought a Bayer Contour USB, not because I can plug it into my computer, but because it has an OLED display. It is the most readable of any meter I have ever used, but even it has its faults. When the meter first displays the results it uses the full height of the display,  shown here (more or less life size):

This is readable, but this figure only shows for a few seconds before the device quickly zooms the figure down to half height because, well because some trendy 20somthing thought that it is "cool" to do that. Well fine, you be "cool" if you like, but please give me the option of only displaying results in the large font. The smaller display is shown here, and this is the size of the font when displaying stored results:

The extra data at the top of the display (and at the bottom when viewing stored results) may be useful, but surely I should be given the option to only display it when I press one of the buttons? So my perfect device is only perfect for the few seconds that it shows the large figure. This is an example of a device with a great feature that has been spoiled by over-egging that feature (adding the extra info on the display) or by not allowing me to choose how I want the display to work.

The final thing I should mention is development. Developers should recognise that although they may have one good idea, they do not have a monopoly on them and that others may have better ideas. Windows became an important platform for business not just because of the way that Microsoft sold the operating system, but because it was so easy to write applications with Visual Basic (and to a lesser extent, with C++ and MFC). Not only could you write applications in a matter of minutes, but you could also access the features of other applications (like Word or Excel) and use them in your application. The result was a large number of hobbyists developing applications, which resulted in a large number of free applications and a large pool of people from which the commercial software houses could employ, at a relatively low rate.

It is important, then to target a platform which has the largest number of users, and also to make your application and its data as open as possible. So if you develop some innovative way of presenting data, or using the hardware of the device, provide that through an open library so that other developers can use the library and provide additional features that you do not.

In terms of smartphones the most closed is the iPhone/iPad. Apps for these are written in Objective C which is not the most widely used language, it also uses an expensive SDK (software development kit) which will only run on a Mac (another expensive piece of kit) and can only be distributed through the Apple App Store. All of this reduces the available number of developers and makes the development expensive. Microsoft Windows Phones are relatively open compared to iPhones. The latest phones use .NET and the development tools, which run on Windows machines, are free. The .NET languages are C# and VB.NET and there are millions of such developers globally. Windows phones also run Javascript widgets, which brings in millions more Javascript developers. Android phones are also relatively open, with applications written using Java and there are millions of Java developers globally (although probably fewer in number than .NET developers). Java development tools run on any machine that runs Java: Windows, Macs, Linux etc.

However, I have already discounted smartphones because they are too expensive. These days most phones will run Java in some form or another. Thus it makes sense to develop a telemedicine application using Java. Java allows you to develop libraries that can be distributed compiled (so you can allow others to use your functionality without disclosing how you do it).

So in summary. A telemedicine application should target a cheap, freely available device, one that can be used by as many patients as possible. The application should be configurable, and allow patients to change the screen fonts, colours and even the buttons that are used. It should also be as open as possible, allowing other applications to use the data your application generates, and use your functionality through an open library.