5 ms·

My mother works in the healthcare field, and when they announced the iPad I immediately knew that doctors would be clamoring to use them, and it's good to know

by kylec 16y ago

My mother works in the healthcare field, and when they announced the iPad I immediately knew that doctors would be clamoring to use them, and it's good to know that, at least with the webapp-based systems, that the iPad 'just works'. However, I'd also be interested in hearing from doctors that might not happen to be "chief information officer and dean for technology" and see how easily and effectively they can use the device.

niels_olson 16y ago
I'm a surgery intern. I think tablets will have to become as ubiquitous and uninteresting as cell phones before residents would start carrying them. If there's a code and you have to respond, you can't just drop a tablet in your pocket and run. If you leave it at the nurse's station at the county hospital, there's every reason to assume some gangbanger visiting his gangbanger buddy will take it. If you're going to the OR, there is precious little horizontal space where you can safely lay something down, or even lean it. Everything in an OR is on wheels, and is likely to get sprayed with some fluid, before, during, or after the case. I also think authentication will need to get lighter and more responsive. I think thumbprint detectors in the margin would be ideal. Pick it up and login. Logout if it sits without input for 2 minutes. If someone else picks it up, immediately log me out and log them in. Along with authentication, most hospitals use Citrix clients, but not all Citrix servers seem to be configured for various clients. I have never been able to access our Citrix server via iPhone, but the BlackBerry folks can. As for electronic notes, the "eCare" system at Sentara hospitals here in Virginia is overall faster than the paper notes at the naval hospital, but not as coherent. Designers need to come in, look at hand-written notes, and figure out how to lay that out in electronic notes. Tables upon tables of (value, high/low marker, units, normal range) for all 30 or so elements of a CBC is unnecessary when all I want is \ 13 / 7.2 ---- 335 84 N / 35 \ and that should fit on two lines, and the BMP should fit next to it, and the calcium-mag-phos should fit next to that, and the LFTs, and the coags . . . it should take all of 2-3 lines on the note, because that way it all stays in your field of view and you can think about it without having to scroll back and forth past gobs of data you're not interested in. Same with radiology: echo the impression, not the whole report. Vitals: always, always, always Tmax, Tcurrent, HR, BP, RR, SpO2, unless you're in a critical care setting, where you want the same order, but with ranges. It's all really fairly straight forward, but someone who knows CSS needs to sit down and figure it out.
- exit 16y ago
  i feel like i see a lot of clipboards with paper in medical dramas, etc. - in terms of form-factor and what you were saying about horizontal surfaces, how do they work out? or maybe clipboards aren't really that common?
  - niels_olson 16y ago
    The most common form factor in the wild is the 8.5 x 11, usually folded once to make it 8.5 x 5.5. The most common clipboard is actually one that folds exactly like that: http://www.whitecoatclipboard.com/ http://www.whitecoatclipboard.com/
- phren0logy 16y ago
  I'm a physician also, and I want to offer a counter-point. Handwritten notes are done the way they are for speed, not because it's the best way to present data. For example, in the notes I read the CBC "fishbones" often have little arrows or annotations next to them indicating trend or prior values. They do a very respectable job, but I disagree that emulating them is the best approach moving forward. Also, our hospital also uses Citrix. It sucks. I cannot think of any reason why this approach would be preferable, from a user's point of view. But then, nobody who buys this stuff actually has to use it, which is the crux of the problem.
  - ashearer 16y ago
    It's true that handwritten notes are optimized for speed of writing, but they're also optimized--perhaps secondarily--for the speed of reading, which happens at least as often as writing (we can certainly hope). There's a third optimization at work, for small size, both for its own sake and as a byproduct of optimizing for writing speed. Less ink tends to be faster. Tufte frequently mentions the importance of saving ink when drawing charts, going as far as the rule of thumb that if any line can be erased without changing the meaning of the chart, the chart is better after doing so. But of course, it's not really about saving ink itself, it's just that total ink is a good proxy for the presence of distractions that impair clarity (chartjunk). Similarly, fast-to-write notes like the fishbone also tend to be fast for trained practitioners to read at a glance. Though they may not be exactly what we'd come up with if starting from scratch with reading speed as the sole criterion, they get much closer than the typical lab results report. Current electronic alternatives tend to be optimized for something else entirely: ease of printing on a fixed-pitch line printer. We can improve readability greatly by broadening the formatting options beyond that, to include options previously only possible with freeform pen and paper. An ideal layout optimized for readers and free from obsolete constraints (such as writing speed and legacy print engines) would probably abbreviate fewer words compared to handwritten notes, because unusual abbreviations slow down the reader. At the same time, it would get rid of common words that ended up being frequently repeated on computer reports (medjunk). It could include better graphical and formatting cues (selective boldfacing, symbols, sparklines, spatial layouts like fishbones, etc.) We can also take advantage of handheld devices like the iPad to allow trends and ranges to be brought up by tapping on a result, instead of trying to display everything all at once.
angstrom 16y ago
I worked for a medical imaging company last year developing the communication channel for a web portal to allow referring physicians to access study images remotely. The tablet isn't the hurdle, it's the lame network infrastructure and ridiculously slow internet connections. Most the imaging companies have viewers that are nothing more than glorified jpeg viewers. The pain points are accessing the data and how it's stored. Doctors often have extra image studies performed simply because getting the images from previous studies performed at another imaging center is such a hassle.

by kylec 16y ago

niels_olson 16y ago

I'm a surgery intern. I think tablets will have to become as ubiquitous and uninteresting as cell phones before residents would start carrying them. If there's a code and you have to respond, you can't just drop a tablet in your pocket and run. If you leave it at the nurse's station at the county hospital, there's every reason to assume some gangbanger visiting his gangbanger buddy will take it. If you're going to the OR, there is precious little horizontal space where you can safely lay something down, or even lean it. Everything in an OR is on wheels, and is likely to get sprayed with some fluid, before, during, or after the case. I also think authentication will need to get lighter and more responsive. I think thumbprint detectors in the margin would be ideal. Pick it up and login. Logout if it sits without input for 2 minutes. If someone else picks it up, immediately log me out and log them in. Along with authentication, most hospitals use Citrix clients, but not all Citrix servers seem to be configured for various clients. I have never been able to access our Citrix server via iPhone, but the BlackBerry folks can. As for electronic notes, the "eCare" system at Sentara hospitals here in Virginia is overall faster than the paper notes at the naval hospital, but not as coherent. Designers need to come in, look at hand-written notes, and figure out how to lay that out in electronic notes. Tables upon tables of (value, high/low marker, units, normal range) for all 30 or so elements of a CBC is unnecessary when all I want is \ 13 / 7.2 ---- 335 84 N / 35 \ and that should fit on two lines, and the BMP should fit next to it, and the calcium-mag-phos should fit next to that, and the LFTs, and the coags . . . it should take all of 2-3 lines on the note, because that way it all stays in your field of view and you can think about it without having to scroll back and forth past gobs of data you're not interested in. Same with radiology: echo the impression, not the whole report. Vitals: always, always, always Tmax, Tcurrent, HR, BP, RR, SpO2, unless you're in a critical care setting, where you want the same order, but with ranges. It's all really fairly straight forward, but someone who knows CSS needs to sit down and figure it out.

exit 16y ago
i feel like i see a lot of clipboards with paper in medical dramas, etc. - in terms of form-factor and what you were saying about horizontal surfaces, how do they work out? or maybe clipboards aren't really that common?
- niels_olson 16y ago
  The most common form factor in the wild is the 8.5 x 11, usually folded once to make it 8.5 x 5.5. The most common clipboard is actually one that folds exactly like that: http://www.whitecoatclipboard.com/ http://www.whitecoatclipboard.com/
phren0logy 16y ago
I'm a physician also, and I want to offer a counter-point. Handwritten notes are done the way they are for speed, not because it's the best way to present data. For example, in the notes I read the CBC "fishbones" often have little arrows or annotations next to them indicating trend or prior values. They do a very respectable job, but I disagree that emulating them is the best approach moving forward. Also, our hospital also uses Citrix. It sucks. I cannot think of any reason why this approach would be preferable, from a user's point of view. But then, nobody who buys this stuff actually has to use it, which is the crux of the problem.
- ashearer 16y ago
  It's true that handwritten notes are optimized for speed of writing, but they're also optimized--perhaps secondarily--for the speed of reading, which happens at least as often as writing (we can certainly hope). There's a third optimization at work, for small size, both for its own sake and as a byproduct of optimizing for writing speed. Less ink tends to be faster. Tufte frequently mentions the importance of saving ink when drawing charts, going as far as the rule of thumb that if any line can be erased without changing the meaning of the chart, the chart is better after doing so. But of course, it's not really about saving ink itself, it's just that total ink is a good proxy for the presence of distractions that impair clarity (chartjunk). Similarly, fast-to-write notes like the fishbone also tend to be fast for trained practitioners to read at a glance. Though they may not be exactly what we'd come up with if starting from scratch with reading speed as the sole criterion, they get much closer than the typical lab results report. Current electronic alternatives tend to be optimized for something else entirely: ease of printing on a fixed-pitch line printer. We can improve readability greatly by broadening the formatting options beyond that, to include options previously only possible with freeform pen and paper. An ideal layout optimized for readers and free from obsolete constraints (such as writing speed and legacy print engines) would probably abbreviate fewer words compared to handwritten notes, because unusual abbreviations slow down the reader. At the same time, it would get rid of common words that ended up being frequently repeated on computer reports (medjunk). It could include better graphical and formatting cues (selective boldfacing, symbols, sparklines, spatial layouts like fishbones, etc.) We can also take advantage of handheld devices like the iPad to allow trends and ranges to be brought up by tapping on a result, instead of trying to display everything all at once.

angstrom 16y ago

I worked for a medical imaging company last year developing the communication channel for a web portal to allow referring physicians to access study images remotely. The tablet isn't the hurdle, it's the lame network infrastructure and ridiculously slow internet connections. Most the imaging companies have viewers that are nothing more than glorified jpeg viewers. The pain points are accessing the data and how it's stored. Doctors often have extra image studies performed simply because getting the images from previous studies performed at another imaging center is such a hassle.