Hacker News Reader

Comment by flave | Hacker News Reader

I’m sure the individual writer is smart educated and thoughtful, but the system of science journalism (science communication is different but equally flawed) is so bent-out-of-shape as to be effectively worthless.

Like, take this exact article as a great example. I’m sure Mr Biswas is genuinely very intelligent and thoughtful and a great journalist but having him write a science article is unfair on him and on readers.

Doesn’t even have an undergraduate in a science subject, has never worked as a scientist, and his job is as a national correspondent.

Perhaps my wording prioritised humour over fairness - I’ll take the criticism on that. But I don’t think my core point was wrong. How can you “communicate” something you yourself don’t understand?

Finally, I want to stress again - it’s not his fault. The system is broken.

View on HN

Can you point out the issues with the article?

The core issue is that the BBC report inflates what the study actually shows. The paper is a small, single-centre RCT of one specific surgery (laparoscopic cholecystectomy). Its primary outcome is a modest reduction in propofol and fentanyl dose under a very specific anaesthetic protocol. It does not demonstrate broadly faster recovery or an across-the-board clinical benefit. The authors themselves are cautious and explicitly list limitations.

The article strips out that narrow context and generalises. Phrases like “music eases surgery and speeds recovery” and “strongest evidence yet” extrapolate from a sample of 56 people undergoing one procedure to “surgery” in general. The paper doesn’t measure global recovery outcomes, discharge times, or longer-term effects. Satisfaction and pain scores are even reported as comparable between groups (P=0.361 and P=0.07).

There’s also mechanistic speculation in the article (implicit memory, psychological responses, “humanising the operating room”) that isn’t in the study’s data. The paper reports dose differences and perioperative physiological measures—not neuropsychological mechanisms.

> Its primary outcome is a modest reduction in propofol and fentanyl dose under a very specific anaesthetic protocol.

Ooh, that sounds like p-hacking. How many other protocols, and other potential outcomes in general, did they look at before picking the one to publish? If it's on the order of 20, then we can expect they'd encounter such a result by pure chance.

Sure.

The headlines says that music “speeds recovery” but the paper specifically says that patients had similar recovery profiles.

The media article overall overstates the findings of the study. It’s a very specific study on a specific cohort and a specific surgery (minimally invasion) but the article implies strongly that music helps with all surgery.

Also the paper specifically doesn’t touch on medial outcomes from the music - that’s fine as science since it’s granular, but it’s a pretty big thing to miss in the article.

The article misses a bunch of further questions that need more research. How does the patient playing music affect the surgeon? Is it music in general or specific music that helps? Is the patient choosing the music relevant?

“Reshape how hospitals think about surgery”? Not really, hospitals already use music in surgery so it’s not going to “reshape” anything. Over dramatisation.

It’s also just very shallow. Makes no mention of existing science/practice for example. Didn’t speak to any other researchers.

Look, the article is fine-ish but it’s just a regurgitation of the paper with more dramatisation and no analysis. Just post the paper especially on HN.

Good points - that's why I follow & support https://theconversation.com/ for news since it's Science Journalism is done by actual scientists working in the field.

> How can you “communicate” something you yourself don’t understand?

This goes both ways: how can you (as a scientist) communicate something when you don’t understand communication?

The answer to both is to let the person who understands it and the person who is good at communication collaborate.

Reply on news.ycombinator.com