Friday, 14 August 2015

Smoking ban miracle strikes again

I should be bored of writing about these stupid smoking ban miracles by now but the sheer audacity of the deception never loses its power to shock. After looking at heart attacks, asthma and lung infections, they've now resorted to stillbirths. The 'they' in question include two past masters of the genre, Jill Pell (see here) and Christopher Millet (see here).

The Office for National Statistics has data on stillbirths (postneonatal mortality) as well as another category studied by Pell et al., neonatal mortality. You can clearly see the big drop after 2007 when the smoking ban was introduced...

What, you can't see it? What do you mean you can't see the emperor's beautiful new clothes?

Okay, the rates have been going down at a gradual and pretty constant rate for years, but that doesn't mean the finest minds in tobakko kontrol can't claim that 'almost 1,500 stillbirths and newborn deaths were averted in the first four years after the law to prohibit smoking in public places was enacted'.

The 'study' responsible for this delusion is published in Scientific Reports and it uses the now-familiar trick of predicting how many deaths are going to take place if the smoking ban is not introduced and comparing the prediction to the actual number of deaths that took place with a smoking ban in place. Abracadaba! There are always fewer deaths than were predicted. Hello BBC, I'm ready for my close up!

This would be pretty neat if the prediction was made before the activist-researchers had seen the data, but they have seen the data and it's not a prediction, it's a counterfactual. That means the quackademics plot the data on a graph and then draw on another line which supposedly represents how many deaths there would have been if the smoking ban didn't take place. The line can go anywhere so long as it is higher than the actual line, so that's where they put it. Bingo! The ban saved lives.

The authors of the latest piece of junk don't bother to explain how they came up with their counterfactual, but here's the key graph showing rates of stillbirth...

As you can see, the actual rate (in grey) was falling steadily before the ban and continued to fall steadily after the ban. Despite the decline between 2003 and 2007, the authors decided that the rate after 2007 would have risen and stayed high if the smoking ban hadn't been introduced (red line). As it turned out, the ban was introduced and the rate continued to fall at the same unexceptional rate that anybody else in the world would have predicted based on the secular trend. Thank goodness for the ban, eh? What a close shave!

It's disgraceful that this utter bilge gets published and reported time after time. It is unlikely to be coincidence that it came out on the same day as the Royal Society for Public Health called for the smoking ban to be extended. Either way, it's so obviously policy-based junk science that everyone involved should hang their head in shame. They won't, of course. They'll be popping champagne and congratulating each other for getting away with it again.

No comments: