Percentages and the pastrami panic…

the hot dog horror, and the salami scare. This story in the NYT quotes a source:

 “We see a 4 percent increase in the risk of cancer even at 15 grams a day, which is a single slice of ham on a sandwich,” said Dr. Nigel Brockton, director of research for the American Institute for Cancer Research.  
Eating a more typical serving of 50 grams of processed meat a day would increase the risk of colorectal cancer by 18 percent, a 
2011 review of studies found.

What does this really mean? Lifetime risk of colorectal cancer is about 1 in 23, or a little over 4%.  Now, does that slice of ham double your risk (4% to 8%), or merely increase it from 4.3% to (1.04*.043 = .045), 4.5%? Do a full fifth (18 + 4 = 22) of the 50-gram noshers get these specific cancers? Of course not. The quote, and the story, are completely ambiguous, but if you follow the link, you find that the data are relative risk values, which is the second interpretation. 50 grams a day entails about a 1% extra risk, and that’s not even counting all the people already in the 4.3% who eat deli meat and get cancer. If you do, and you stop, your risk of these cancers goes down from about 4% to…a little more than 3%. Perhaps Zabar’s should sue the Times over this alarmism.

Eating a reasonable amount of these exceptionally yummy foods seems to me a good deal, at the price of being 1% more likely to get this type of cancer before I get one of the other kinds or a heart attack. YMMV, of course. Everyone dies of something, so a much more useful statistic would be the average number of [quality adjusted ?] life years I’m putting at risk from a ham habit, and from an occasional indulgence.

The lesson here is that any statistics involving percentages have to be stated carefully to make it clear whether an increase adds to an existing rate or multiplies it, and “X% added risk” simply doesn’t cut it. Dr. Brockton and the reporter are equally at fault here, along with the Times copy editor. Students and colleagues: don’t make this mistake, especially when you’re explaining science to the public. What Dr. Brockton meant to say is that “the 15g pigout habit raises your lifetime risk from 4 to 5%”. There’s no escaping the additional words. Or reporting base rates: something that “quadruples your risk of contracting the gleeps” is not a big deal if the incidence of gleeps is a fraction of a percent.

Evidence-based catfighting

Lessons of the great row in the Cochrane collaboration.

Want a change from watching the turds circle the drain in the Kavanaugh confirmation circus? Let me bring you a nasty academic spat between high-minded medical researchers. This is how learned gentlemen stab each other in the back! With a couple of serious morals. Everybody named below is a highly credentialed professional; I leave the titles out to avoid repetition.

The milieu is the Cochrane collaboration. Inspired by and named after the  epidemiologist Archibald Cochrane  (d. 1988), the Cochrane people promote evidence-based medicine through meta-analyses of randomised clinical research trials using methods as rigorous and objective as they can make them. (Our own Keith Humphries has been a Cochrane reviewer.) [Update] The very solid proposition is that if you can analyse correctly a handful of properly conducted trials, you are in effect adding the sample sizes, so you can draw much more statistically reliable conclusions than by cherry-picking one. There is of course a lot of art here behind “correctly”, “properly”, and “in effect’. [/update] They are not the only researchers carrying out meta-analyses, but a Cochrane review is widely regarded as the gold standard. Depressingly often, the answer is “we don’t know”.

One recent Cochrane review (lead author Marc Arbyn) was on vaccines against human papilloma virus (HPV), which causes much cervical cancer among women and lesser numbers of anal and penile cancers in men. This is not a trivial health issue. Fortunately there are vaccines sold by Merck (Gardasil) and GSK (Cervarix). Do these work? Short answer: yes. Are they dangerous? Short answer: no. (Please DO NOT quote me, read and cite the report, they do provide a summary for dummies.)

So far so routine. But then an article was published in the journal BMJ – Evidence-Based Medicine by Lars Jørgensen, Peter Gøtzsche, and Tom Jefferson, alleging that the vaccine review was sloppy on several counts and hinting that it was influenced by pressures from the Big Pharma vaccine vendors. (Note that while they argue that the side-effects are greater than the review says, the critique does not recommend stopping or curtailing vaccination programmes.) This naturally provoked a rebuttal from the Cochrane management (David Tovey and Karla Soares-Weiser), saying the criticism is wrong on all counts.

It did not stop there. Gøtzsche is, or rather was, a member of the Cochrane board, indeed a founder member of the organisation. He could presumably have raised his concerns there first rather than publicly. After a presumably furious board meeting, Gøtzsche was expelled and four other board members quit. The great collaboration is now in existential danger. Will donors, including the Gates Foundation, keep the funding flowing? Will Gøtzsche set up a breakaway fitzCochrane, applying his own higher standards? Will anti-vaxxers and misogynists exploit the row to attack the vaccination campaign? Only 27 % of American men under 26 are vaccinated.

It’s important that the crisis be resolved quickly and the collaboration continue. There’s not much outsiders can do to help this in the short term, and I am quite unqualified to take sides. I have though one reflection and one suggestion for the future. Continue reading “Evidence-based catfighting”

Hurricane season again

There are more Florences to come.

Hurricane Florence, from the International Space Station

Hurricane Florence, downgraded to a tropical storm, continues to dump massive quantities of rain on South Carolina, with more to come. She looks like a rerun of Harvey, which flooded Houston last year, cost $125bn. Are these “Acts of God or of the Queen’s enemies”, in the picturesque language of old British insurance contracts?


A bit of both. IPCC 4th Assessment Report, 2007, WG1:

A synthesis of the model results to date indicates that, for a future warmer climate, coarse-resolution models show few consistent changes in tropical cyclones, with results dependent on the model, although those models do show a consistent increase in precipitation intensity in future storms. Higher-resolution models that more credibly simulate tropical cyclones project some consistent increase in peak wind intensities, but a more consistent projected increase in mean and peak precipitation intensities in future tropical cyclones.

We’ve known for at least a decade, for the subset of “we” capable of wading through IPCC prose or reading more popular transcriptions of the science, which should include the press, TV weathermen and policymakers. In this case, the science is extremely simple in outline:

Warmer tropical seas → warmer and wetter air above them → conversion of extra heat energy into rotational energy by the cyclone mechanism → bigger and wetter hurricanes.

Sandy, Harvey, Irma and Florence have been hurricanes modulated by the modest global warming of 0.8 degrees C since 1880, the period with a full and accurate instrumental record. To be generous with the earlier uncertainties, let’s say at most 1 degree C above pre-industrial (say 1750). There is quite certainly more warming to come. Jerry Brown’s recent executive order, aiming at zero net emissions in California in 2045, was rightly hailed as brave political leadership (grandstanding to opponents). Sweden was there first, with the same date.  These are the cutting edge of real policy commitments; most countries have done nothing to translate into action their vague Paris Agreement commitment to zero carbon “in the second half of this century” (Article 4(1)).

Suppose by a miracle everybody else joined Jerry Brown tomorrow. We would, it seems, be on track to the more ambitious 1.5 degrees aspirational goal of the Paris Agreement. Meeting the main 2 degree cap only calls for moderate optimism, not a miracle. The range of good outcomes – never mind the bad ones – lies between doubling global warming from the pre-industrial level, and only increasing it by half.

More storms like Harvey, Irma and Florence are certainly on the way.

Continue reading “Hurricane season again”

Racist genetics

What is the “scientific” racist hypothesis?

I can’t get my head round genetic racism. I don’t mean why people believe comforting rubbish, that’s easy to understand. I mean: what exactly is the proposition? “Blacks are naturally dumber than whites” doesn’t hack it. “Dumber” is usually defined as “scoring lower on IQ tests”. There’s a whole argument about the relevance of IQ, and its malleability as shown by the Flynn effect, but it is a reproducible test with a fair correlation with other sorts of mental skills, so the meaning is clear and I’ll let it ride for the sake of the argument. But just who are “blacks” and “whites”, in multi-ethnic populations like that of the USA with a long history of miscegenation? My genetic knowledge is pretty thin, but it may be interesting to see how far you can get on the genuine science with Wikipedia articles, until someone with real expertise shows up.

There is no black race. Continue reading “Racist genetics”

Does cannabis availability help prevent opioid overdoses?

There’s been lots of chatter about the cannabis-opioid substitution question.

Newsweek headlines, “Can Legal Marijuana Solve the Opioid Crisis?”  while Dr. Jeff Sessions opines that cannabis is “only slightly less awful” than heroin.

People whose background is medical research tend to distrust anything that’s not a randomized controlled trial. They point to the positive correlation between cannabis use and opioid use at the individual level, and the fact that opioid deaths continue to rise even where cannabis is most freely available. Their position is, “We don’t know anything about this. Let’s due the clinical studies before taking action.”

But “not taking action” now means continuing to criminalize even the possession of cannabis. If cannabis substitutes for opioids, those laws cost lives: lives that can’t be regained ten years from now, after the clinical-trial results are in.

Moreover, the relevant clinical trials can’t actually be done in the U.S. Continue reading “Does cannabis availability help prevent opioid overdoses?”

Popper and Kahneman visit an Indian coal mine

Indian coal offers a nice moment of Popperian falsfication.

Daniel Kahneman has a simple explanation why we don’t think things through: laziness. It’s no work to rely on the sloppy, but fast and efficient, Hare mental system, using short cuts and stereotypes to get a response that is, under the current US President, good enough for government work. Rigorous thought is hard.

Karl Popper offered a short cut through the hard part that is still rigorous: falsification of hypotheses. One false prediction and you’re out. A nice idea, but it rarely works. You can save almost any hypothesis with tweaks, including Ptolemaic astronomy. So it’s back to comparing the best shots of the competing hypotheses, hard work again.

Just occasionally, life presents us with a simple Popperian test. Here is one I spotted, on the recondite but important subject of Indian coal burning. There are two entrants. Goliath is the IEA, a stuffy but reputed intergovernmental policy and data shop in Paris. David is IEEFA, a small energy policy think tank in Cleveland.

IEA: India’s coal consumption will more than double by 2040. (IEEFA pdf, page 1.) The source is presumably the IEA World Energy Outlook 2017, paywalled; it’s not in the free summary. See also this IEA FAQ:

The positive IEA outlook for coal demand through 2020 is based in part on growth in India and Southeast Asia that will more than offset structural declines in Europe and the United States.


The headline to the chart understates the predicted change: growth will be trivial after next year. This means that India’s overall carbon emissions may stabilise in less than a decade, assuming the electric transition goes as fast in transport as the government plans.

Who’s right? Continue reading “Popper and Kahneman visit an Indian coal mine”


A monument in Moscow to peer review.

Students at the Higher School of Economics in Moscow have unveiled what must be the first monument to peer reviewers.

The sculpture takes the form of a die displaying on its five visible sides the possible results of review — ‘Accept’, ‘Minor Changes’, ‘Major Changes’, ‘Revise and Resubmit’ and ‘Reject’.

The main academic sponsor, sociologist Igor Chirikov, spends part of his time at Berkeley.

It’s a better idea than a sculpture.  The block of concrete was just sitting there, and the challenge was to reuse it, which rather limited the artistic options. It’s a pity that the key words are all in English: a realistic reflection of the current state of scientific publication, but Russia has a great scientific tradition that could have been recalled at the same time. Do commenters have better suggestions for honouring the Unknown Peer Reviewer?  Disappointed authors might for instance go for a hangar full of vampire bats.