Live data from Hacker News

Doctors who use Google Translate to talk to patients want a better option

statnews.com

81–90 of 161 posts

Re: Doctors who use Google Translate to talk to patients want a better option

#81

Earlier quoted context omitted.

The proportion of patients who become addicted from properly prescribed opioids is extremely small. The majority of addicts created were provided with opioids that were not appropriate for their particular medical situation, in doses that were too high, and on a regimen that exceeded in length what was supported by the medical evidence. The proportion of new cancer patients who go on to receive palliative care that i…

A meta study of 38 studies from 7 years ago on patients using opioids for chronic pain found that 21-29% of the patients misuse the drugs and 8-12% of the patients are addicted to them. That doesn't seem "extremely small" like you said, though maybe it's better now. I think there are reasons why opioids are no longer recommended for the treatment of most patients with chronic pain that you aren't fairly acknowledging…

https://www.nejm.org/doi/full/10.1056/NEJMra1507771 Volkow is a mainstream researcher, using data from the peak of the crisis, and cites the rate at just under 8%. It is not at all unreasonable to assume that with proper prescribing (unlike what happened in the 90s and 2000s) that chronic pain patients can have opioid misuse and addiction numbers near or matching that of palliative care patients, where addiction rates are closer to 1% than 10%.

Re: Doctors who use Google Translate to talk to patients want a better option

#82

Is no one going to address the elephant in the room? If you're in a country where the lingua franca is X, you need to be able to speak X or life will be difficult. I'm not sure why the onus is on every hospital to become a kind of mini united nations here.

I recently went to Poland for a quick 2 day holiday. Felt great before I left. Felt great while I was there. Tested positive for covid the morning I was supposed to fly home. Fast forward a bit and I'm in the back of an ambulance while my partner was on the plane back home.

They put me up in a hotel for 10 days. Fed me 3 times a day. Checked on me twice a day. And on day 10, opened the door and I was on my own.

How I could have managed that without Google translate, I don't know. None of the doctors or nurses spoke more than a few words of English. And I could barely say hello, goodbye or thank you. (I can now!)

As you say, life would have been difficult. It was difficult even with google translate. But it made the interactions much easier on both parties involved.

I would hope that if someone was visiting my country on holiday, or to visit their relatives, and happened to get sick, that the hospital would do all they could to communicate with them and not treat them like they shouldn't even be there.

Re: Doctors who use Google Translate to talk to patients want a better option

#85

I and some others in the healthcare/healthcare tech spaces volunteered to create a discharge templating system for many of the top non-English and non-Spanish languages used in the United States - specifically the top languages in the Northwest to start. Our offering utilized data from, and would have been integrated with, an EMR whose name means "particularly impressive" even if their response was anything but. Afte…

My partner has used an EMR called Epic in her work and it, like all EMRs it seems, is trash. Government protections and a moat of network effects, they’re untouchable.

It's hard to design software that's both flexible and reliable enough to meet the needs of these medical users. In an ideal world imo, hospitals could plug in various modules into an overarching health IT framework -- that way they could pick and choose what works the best for them. However, that would mean they're now working with dozens of systems/companies instead of just one.

Re: Doctors who use Google Translate to talk to patients want a better option

#86

Earlier quoted context omitted.

Taigu is a Taiwanese-specific dialect of Hokkien. Hokkien is spoken outside Taiwan (e.g. Fujian).

> Hokkien is spoken outside Taiwan (e.g. Fujian). It would be pretty shocking if that wasn't the case, since Hokkien is spelled 福建.

Right, but when people say "Hokkien" in English they generally mean 閩南語 - though I've also heard the term "Southern Min" used.

Re: Doctors who use Google Translate to talk to patients want a better option

#87
post #73

Earlier quoted context omitted.

My partner has used an EMR called Epic in her work and it, like all EMRs it seems, is trash. Government protections and a moat of network effects, they’re untouchable.

EMR software, like all medical software, suffers from a variation of the Dead Sea Effect. Those who are truly invested in EMR software are not developers and developers are not invested in EMR software. So, as a developer, it's a job you take because you need a job. And if you're competent, you can get a job elsewhere eventually. If you're not, well, it's not like medical software has a deep talent pool to draw from.…

Besides pay, is there any reason a dev would be more invested in any other B2B app? A few people I know work in medical software, and a lot of them think it's meaningful work.

Re: Doctors who use Google Translate to talk to patients want a better option

#88
post #82

Is no one going to address the elephant in the room? If you're in a country where the lingua franca is X, you need to be able to speak X or life will be difficult. I'm not sure why the onus is on every hospital to become a kind of mini united nations here.

I recently went to Poland for a quick 2 day holiday. Felt great before I left. Felt great while I was there. Tested positive for covid the morning I was supposed to fly home. Fast forward a bit and I'm in the back of an ambulance while my partner was on the plane back home. They put me up in a hotel for 10 days. Fed me 3 times a day. Checked on me twice a day. And on day 10, opened the door and I was on my own. How I…

Right, but if you decided to move to Poland - you'd realise you probably need to speak Polish, right?

Re: Doctors who use Google Translate to talk to patients want a better option

#89
post #59

Earlier quoted context omitted.

What are countries with more functional healthcare systems doing differently? When it comes to comparisons with other countries, much of the dialogue in the US is centered on insurance coverage. That said, I'm curious if other countries have found better solutions to similarly fundamental issues, such as doctors working long hours, doctors being forced to maximize the number of patients, continued medical education,…

Career healthtech worker; opinions are my own; focusing only on the provider-specific questions you're asking: Biggest problem (w.r.t. provider overwork and availability) is regulatory capture by AMA and hospital groups. The AMA has for decades induced an artificial shortage of MDs by limiting the number of available residency slots. In particular, there's an acute shortage of slots for primary care providers (gen pr…

I initially thought this was a fringe comment, but I was surprised to learn that your criticisms have basis, with adherents like Milton Friedman [1]. It certainly toppled my presumptive association of the AMA with putting patient care first.

P.S. I found this tidbit on tobacco particularly troublesome [2].

[1] - https://en.wikipedia.org/wiki/American_Medical_Association#C...

[2] - https://www.sourcewatch.org/index.php/American_Medical_Assoc...

Re: Doctors who use Google Translate to talk to patients want a better option

#90
post #12

Deepl is a better option. It feels several years ahead of Google Translate in quality. I know enough Japanese to hurt myself, and Deepl is definitely giving the superior translations, by far, which is great for my studies. As an analogy, it’s like going from 144p to 360p (a good human translation ranges from 720p to 4k+). Yeah 360p video is still blurry but it’s in a different league than 144p.

The problem with machine translation is not the obvious general mediocrity, but the occasional undetected catastrophic failure, especially in medical context. I don't think DeepL (or any other fully automated system) is any more reliable in this sense.

Can any automated solution be reliable enough? I don't think so unless you can prove it does less harm than a professional medical translator.
Post reply on HN