Time to think again about computers
o you know the following brand names: Panhard-Levassor, De Dion-Bouton, Isotta-Fraschini, Hispano-Suiza, Minerva, Excelsior, Hammel, Horch, Maybach, or Lloyd? For those of you who do not recognize them, they are all automobile manufacturers or, in fact, they were. In 1898, there were 50 automobile manufacturing companies in the United States of America, a number that rose to 241 by 1908. The names above are European and tell a similar story for Europe. Dozens of companies started building motorcars around the turn of the century.
Cars built before 1914 looked very different from those built today, whereas those built before the Second World War more closely resemble those of today. In the last thirty years, cars have also become simpler and more comfortable. They are now very easy to operate and because their handling does not differ greatly it is not a major problem to change from one to another.
The point is that it took more than half a century for motorcars to evolve into a stable product that can be used by anyone with a minimum amount of training. The automobile industry responds to new developments in safety, comfort, environmental protection, or pure fashion, but in essence, the product remains the same. Along the way, however, many manufacturers faltered and disappeared.
This example should be kept in mind when updating your views on computers.
My first contact with computers in medicine was in nuclear medicine some twenty years ago. We had a DEC station connected to a gamma camera for cardiac examinations with thallium-201. It created beautiful color pictures of the myocardium, and I could easily manipulate the size of an ischemia or infarction by playing with this computer. This made a lasting impression on me which has stayed with me ever since: it made me realize that computers are only as intelligent as those who use them. As with cars, they are tools (or toys). As tools, they can make life a lot easier.
Fifteen years ago, in 1982, I bought my first personal computer for writing articles, making calculations, creating patient databases, and evaluating data collected in studies. It was a Kaypro II and I bought it in a gun shop on Long Island; at that time, there were no specialized computer shops.
The manufacturer of this computer has long disappeared from the market – as has this computer from my life. If you were to see this computer today, you would think you were watching a bad science fiction movie. It was a “portable” computer: a heavy, clumsy, noisy box, the cover of which would become the keyboard; the small, black-and-green screen was next to the two floppy disk drives; and, of course, there was no hard disk. Today you would laugh at its performance, yet it was only fifteen years ago.
Since then I have changed computers nearly every other year. At present I have four personal computers – more than enough you may think, but two of them are too slow for today's software applications, and I use one of them solely to read floppy disks.
In hospitals, and radiological departments in particular, the situation has changed accordingly. You find computers everywhere, not only connected to run equipment, but also freestanding or network-connected personal computers, but I wonder if someone has ever made a cost/benefit calculation for personal computers in radiology.
Of course, reports processed on computers usually contain fewer mistakes and look nicer than those done on an old-fashioned typewriter. With PCs is easier to create forms and calculate the hourly, daily, weekly, and monthly patient throughput. We can also use our PCs for image processing, archiving, communication, and even teleradiology.
But do we really need all this computer-based administrative bureaucracy, or are we just creating more work and data output in a manner worse than predicted by Parkinson’s law? Are we not only creating new employees in healthcare administration but also becoming serfs to our computers because we are unable to restrain ourselves from trying out new software tools and doing work other people could do?
Increasingly, radiologists are doing the work of secretaries; after all, the PC is sitting there on their desks and they do not have to bother dictating, explaining, or begging that the reports should be typed up immediately. Yet, the hours accumulate and you end up filling your workday with such jobs, not to mention the fact that you leave secretaries out of work.
When you try to find out which computer, which paraphernalia, and which software are most suitable and best for your purposes, you have three possibilities: Ask somebody you know, try to get advice at a shop, or read computer magazines.
But even if you ask a well-trained software engineer which computer you should buy and what software you should install, you will not get straightforward answers. At best, the answers will be ambiguous: It depends on what you want the computer for, how much you are willing to spend on it, whether you will use it only at your workplace, and certainly how good you are at programming.
Salespeople in computer shops are worse. Of course they want you to buy the latest, greatest, fastest, lightest, and most wonderful piece of equipment on the market. It does not matter than in three months there will be a new machine that meets these criteria. They want to sell, so you cannot blame them, but it is doubtful that you trust them.
As for reading computer magazines, there is no need to waste your time. Most of these publications are full of advertisements anyway – you are better off going to the shop, where at least you get to touch and see the equipment. The few articles you do find in these journals are written for experts, so you are left confused and with no easy answer to a simple question.
When you finally buy our computer, our troubles are just beginning. You must partly assemble it, install the software, learn how to use it, etc. Often you also have to change the electric plug because it does not fit into the socket. Endless hours are spent reading manuals, consulting friends and colleagues who have similar PCs, and finally trying to reach the telephone “hotline” of the manufacturer.
The lucky few will reach a person and not an answering machine, but the advice is seldom useful. To continue the installation, you are told you have to straighten out a paper clip and push it into the hole in the bottom of your computer, being careful not to drop the computer during this procedure. You must them continue with step 126 on page 74 of the manual.
Sometimes, reading the small print in the documents that came with the computer you can find sentences starting like the following: “If you experience erratic function with your computer, …“
In other words, the manufacturers knew from the beginning that the device they sell functions only randomly.
If you are persistent, you may end up with an operational computer and can begin catching up all the work that has accumulated since you thought about buying a new computer.
What is it about computers that makes us deviate from the normal procedures when buying equipment?
Usually, when your department buys a new refrigerator or angiography suite, somebody will deliver and install it. The same should hold true when buying computers: Have them delivered and installed by the people from the computer shop. If they tell you that they do not have such a service, then move on to the next shop. There are five per block.
Interestingly enough, a German poll has revealed that 50% of all new PCs break down within the first six months. Could it be that all kinds of products are being dumped in the market for us, the users, to test? After all, there will be plenty of good excuses if things do not work out as they should. Technology is advancing and growing rapidly, and users are so hooked on it that they are more than willing to try the latest fads and frills.
Much of the fantastic savings this technology is supposed to bring about are questionable. You can save time when it works properly, but has anybody kept an account of the time spent purchasing, setting up, learning, repairing, reinstalling, and cursing? We save space when information can be stored on a PC or on disks and accessed with the push of a button, but just in case, we keep mountains of diskettes with backups and all our hard copies on paper, x-ray films and slides.
There is only one thing that becomes obsolete faster than a personal computer: the popular color in women's fashion, which changes every six months.
Information technology (IT) booms and networking is the name of the game. Massive public relations and marketing efforts by the manufacturers have led to the conviction that radiology departments and private offices cannot survive without it. Often, decisions to install computers and computer networks are made prematurely, and they are made by technology freaks who call themselves “system consultants”, rather than by persons involved in daily routine, or healthcare administrators who could make cost/benefit calculations.
Many commercial enterprises have invested heavily in IT, but there is no proof of a proportional link between the amount invested in computers, accessories, and personnel and the success of the enterprise. Some companies spend 0.5% of their annual turnover on IT, others 3%. Market research shows no difference in profitability between them. It can be concluded that there is a certain level of necessary investment in data technology, and anything beyond this investment creates no added value. The same holds for medical enterprises.
We are becoming increasingly dependent on this technology, and there is no going back. I am the first one to acknowledge I do not want to go back to my typewriter and dictaphone. We are on a roller coaster, supporting a market in which we are both users and used but we must not forget we are still in a developmental period. Hopefully, in no more than 30 years we will reach the same point of stability that the automobile industry has reached today.
It must be possible to create hardware and software which is made and written for users who are not amateur computer-repairmen and which perform the tasks they are promised to have been developed for without erasing irretrievably all data created during the last week.
PS. Some years later. I am using Linux now, in its Ubuntu versions. It's not perfect for playing, but it's the perfect workhorse, exactly what I need. And if something goes wrong, it's easily solvable. If the computer crashes, I just reinstall Linux. It's pleasant and lets you forget the terror of Microsoft.