More iffy science reporting: "Babies born at weekend more likely to die"

[Simon]

I guess this isn't so surprising when a front page news story backs up the government-of-the-day's controversial policy:
(The UK government want to bring in a fully-staffed 7 day service against much resistance: many people fear that since the current 5 + 2 service can't cope, the money would be better spent on raising standards for that. Junior doctors are striking over the govt imposing a new contract they say will reduce their earning by removing the premium for weekend working)

Babies born at weekends 'have higher death risk' - BBC News
Babies more likely to die if born in NHS hospitals at weekend - Telegraph

It's not really true. Babies born following admission of the mother to hospital on a weekend had a higher mortality rate than for admissions on a Tuesday. But so too did babies born to mothers admitted on Weds, Thurs and Fri - as the graph below shows (days from left to right are Monday-Sunday). Which looks like blatant reporting cherry-picking to me.

Crucially for the current debate:

No consistent association between outcomes and staffing was identified,

Study: Association between day of delivery and obstetric outcomes: observational study | The BMJ

 

[MeSci]

I guess this isn't so surprising when a front page news story backs up the government-of-the-day's controversial policy:
(The UK government want to bring in a fully-staffed 7 day service against much resistance: many people fear that since the current 5 + 2 service can't cope, the money would be better spent on raising standards for that. Junior doctors are striking over the govt imposing a new contract they say will reduce their earning by removing the premium for weekend working)

Babies born at weekends 'have higher death risk' - BBC News
Babies more likely to die if born in NHS hospitals at weekend - Telegraph

It's not really true. Babies born following admission of the mother to hospital on a weekend had a higher mortality rate than for admissions on a Tuesday. But so too did babies born to mothers admitted on Weds, Thurs and Fri - as the graph below shows (days from left to right are Monday-Sunday). Which looks like blatant reporting cherry-picking to me.
View attachment 13813

Crucially for the current debate:

Study: Association between day of delivery and obstetric outcomes: observational study | The BMJ

Was the misinformation supplied by the Science Media Centre?

 

[MeSci]

Was the misinformation supplied by the Science Media Centre?

Replying to my own message - just checked out the SMC and was surprised to see that they had actually picked up the same issue as you:

http://www.sciencemediacentre.org/e...nd-and-weekday-births-and-obstetric-outcomes/

So whose idea was it to distort/misreport it?

 

[msf]

I think all babies are equally likely to die.

 

[duncan]

YES!

Thank you, @msf.

Whatever has become of journalism???

 

[Simon]

I think all babies are equally likely to die.

!

(Though graph is for in-hospital perinatal mortality)

 

[TiredSam]

Replying to my own message - just checked out the SMC and was surprised to see that they had actually picked up the same issue as you:

http://www.sciencemediacentre.org/e...nd-and-weekday-births-and-obstetric-outcomes/

So whose idea was it to distort/misreport it?

They'll be worried about the competition if someone else is distorting results.

 

[Esther12]

I guess this isn't so surprising when a front page news story backs up the government-of-the-day's controversial policy:

But it's from the BMJ - rather more connected to the uk medical establishment I'd have thought.

I guess admission Wed-Fri might mean greater chance of receiving care over the weekend? Was that looked at?

I'm surprised by the politics of this if it is a paper spun in favour of the Govt., and against the doctors.

 

[Simon]

I guess admission Wed-Fri might mean greater chance of receiving care over the weekend? Was that looked at?

No, an admitted weakness of the study was that they analysed admission date, not treatment date. But babies get admitted all the time so weekend-admitted mothers would show up in Mon/Tues data too. So we aare none the wiser

Replying to my own message - just checked out the SMC and was surprised to see that they had actually picked up the same issue as you:

Not exactly - edit: yes, and also a comment there said the study had been done on the cheap, using low-cost database data with unacceptably high rates of missing data for both birth weight and gestation age (and whether single, twin, triplets etc). More expensive clinical record data wouldn't have had that problem.

So basically the study explains little, and the evidence it does have argues against weekend admissions being a cause of problems.

But best to get born on a Monday or Tues, just to be on the safe side.

 

[MeSci]

MeSci said: ↑
Replying to my own message - just checked out the SMC and was surprised to see that they had actually picked up the same issue as you

Not exactly - the one comment there said the study had been done on the cheap, using low-cost database data with unacceptably high rates of missing data for both birth weight and gestation age (and whether single, twin, triplets etc).

Did you read the same thing that I did? It said, among other things:

While a comparison of weekends versus weekdays does show a statistically significant difference, the figure shows the highest perinatal mortality for an individual day was on Thursday not Saturday or Sunday.