I've read almost all of his books in the original Polish and in English. Both good and bad translations. For example, the only English translation of Solaris was translated from French, making it a dull read in English, though it's much better in Polish.
The majority of his books, esp. those translated by Kandel, are very faithful to the original. Not the details, of course, but in the sense of wordplay and humor. For example, in the Polish version of the story "How the World was Saved" from the Cyberiad, the machine can make anything that starts with the letter "N". The joke is, the machine is asked to make Natrium (Latin for Sodium), but it refuses, saying that it's not the machine that can make anything starting with N in any conceivable language, since every word has some equivalent that starts with N in another language, and so therefore it would be equivalent to a machine that can make anything starting with any letter, but that's not what it is. Anyhow, you can see the English version for yourself:
In Polish, the made-up words are different, but the spirit, and the joke are very much the same in this instance.
In other stories, which rely too much on wordplay and puns, the translation has to be different to work, but Michael Kandel does an excellent job of that. It's certainly an art--though I'm fluent in both Polish and English, I couldn't come close to making such faithful translations.
Here's how an admission would go in the middle of a typical call night: I'd get called at, say, midnight to admit a patient from the ER. I'd go down there to examine the patient and admit them, which means find out what's wrong, formulate a plan of action, and stabilize them for the night.
We actually did have a primitive EMR, which held any recently (within a year or so) dictated discharge summaries -- those are a lengthy summary of what brought the patient in last time, how it was handled, what meds the patient was sent home with. Those were available to us about 1/4 of the time, and were a goldmine of information.
The remaining 3/4 of the time, we had nothing except the patient's memory (they're ill, it's the middle of the night, majority of patients don't keep track of their long lists of meds and dosages). So I'd request the patient's chart to be found. Usually, I'd hear the following from medical records:
A) The chart will be here in the morning: they're understaffed right now (they'd have 1 clerk in there at night)
B) The chart is off to some doctor's clinic from a recent visit, and hasn't come back yet. It'll be a couple of days
C) We have no idea where the chart is.
So I'd have to rely on the patient's recollection of what meds they are taking, what their medical history is, what their allergies are, etc, etc. If you've ever had to go to the ER in the middle of the night, you know how hard it is to remember that stuff about yourself, and how annoying it is to be asked the same questions by the clueless medical staff over and over again.
When I saw patients in my own clinic, it was just as bad. The records were often gone -- to the hospital for a recent admission and still being processed, to another doc or clinic, etc.
I bought a Vaio subnotebook and as an intern kept my own notes on my patients, and carried the notebook with me everywhere. I was ridiculed a lot, but I always had critical info about my patients at my fingertips.
Then I went to another hospital system for residency, and spent some time at the VA, which had an early EMR called VISTA. It was just fantastic! It had usability problems, and required a lot of typing, but it was amazing to see a patient's current medications, list of major problems, past history, etc, all instantly, integrated over hospital and clinic visits, and even across different VA systems across the country if the patient recently moved. It revolutionized care, in my opinion.
So no, it's not a panacea, but a damn sight better than what we have now in many instances!