Can someone more knowledgeable than me about epidemiology please help debunk this?

[TinyGarden]

https://scitechdaily.com/are-dogs-spreading-sars-cov-2-study-finds-living-with-a-dog-increases-risk-of-contracting-covid-19/

So far this study has only been reported in such illustrious tomes as The Daily Mail and the Metro.

Actual study here:

https://www.sciencedirect.com/science/article/pii/S0013935120311208?via%3Dihu

I know correlation is not causation and it looks like it's an online self report study. But would very much appreciate further scepticism from people who (unlike me) really know what they're talking about.

Not only cos I don't want to have to worry about my parents interacting with my dog (when it's allowed again). But also because it tries to paint an association between supermarket home deliveries and Covid.

Took me months to persuade my parents to go for online deliveries rather than pop to supermarkets!

[IrmaFayLear]

What I immediately thought upon reading this, is that who knows if they are telling the truth ?

You are asked whereabouts you think you contracted covid. You could have been in someone’s house, had a relative to visit, stopped to chat to a person when out and about and been quite close, you could have had a sneaky assignation.... but, no, you cast around and say, “Ah, it was the supermarket delivery” or “It must have been the dog!”

I know someone who caught covid. She was convinced it was an Amazon parcel. Convinced. It certainly wasn’t her partner who works in a factory and had a bit of a cough a week before she fell ill .

[orangenasturtium]

There is online grocery delivery available in Spain @PowerslidePanda. The question in the survey is:

Do you buy the essential commodities online (home delivery) or in-store?

So presumably they are referring to online grocery shopping rather than home delivery of goods bought in a supermarket, although the quote in the article contradicts that

I think everyone is pretty unanimous that it is a poorly designed survey with unreliable, self reported data from a selective group (mostly graduate students) that has been poorly analysed and makes some unsubstantiated claims.

[orangenasturtium]

Although I appreciate the irony that I haven't bothered to give examples to substantiate that!

[MedSchoolRat]

I was too harsh... I don't think it's a 'bad' study at all, not 'badly designed', but it is simply a survey of a sample of convenience without verification. What it can tell us with certainty needs to be put in context. There are buckets of things It can't say for certain. It works best as potential supporting evidence.

X-sectional is not actually at bottom of the hierarchy of evidence, just at the bottom of group observation studies, primary research.

Independent SAGE, btw, what they produce is at the very bottom. 'Expert opinion'. Remember that next time you wonder why Indie SAGE views don't carry much weight with a lot of scientists. All the atmospheric scientists going on about covid being airborne? That qualifies as mechanistic research most of what they do, computer & physics modelling etc. Case reports are basically anecdotes.