Book review - Risk Savvy by Gerd Gigerenzer

In my efforts to understand the new mathematics cirriculum, I came across this book at the library (What you don't go to the library?). Gerd Gigerenzer is  a psychologist and (from his bio page) "the Director at the Max Planck Institute for Human Development, Director of the Harding Center for Risk Literacy in Berlin, and partner of Simply Rational - The Institute for Decisions, which he founded in 2015." He likes to write book that give people the tools to make better decisions - which nicely he keeps quite separate from telling people what to do.

So what kinds of decisions? Well big ones. Should I travel by car after the September 11th 2001 events? Should I get a mammogram or a PSA screening? Should I worry about nuclear power? Where should I invest my money? Why do people gamble? 

The theme of the book is to use example decisions like the ones above to Illustrate the decision making processes that work best in different situations, and how advertising or just communication works to subvert our decision making. Most interesting to me is the distinction between decisions where you know a lot of the detail, and others where there are just too many unknowns. In the known-unknowns case (to paraphrase Mr Rumsfeld) detailed calculations make a lot of sense and checklists and rules can really help us to not be lazy and conform to the correct decisions after they have been corrected. The example given in this case was the use of checklists in hospital operating rooms, and the change in safety culture to be more like that of the airline industry. On the other side of the spectrum, where there is just too much information, decisions are better made with simple rules of thumb, or in the case of the person who has put a lot of analysis time in, gut instinct. (one nice idea to find ones gut instinct was the idea of flipping a coin, and as you are doing so finding yourself happy or sad at the coin outcome and changing it if it is not the answer you want!).

The medical side was interesting to me - I should know this, but I had never looked at the screening statistics this way before. The fact that screening for possible cancer does not distinguish between cancers that will kill you and those that will hang around for a long time while you live your life, means that the improvement in relative survival rate of 5 years after diagnosis is much higher than the decline in the mortality rate of different cancers, and that many people are treated in cases where they would have survived without often traumatic cancer treatments.  The example walked through in the book (in a section called 'How Rudy Giullani was misled' although I think he would have found another reason to complain about the affordable care act and the evils of socialism keeping rich peoples taxes above 0% - see my republican rule of thumb) is the PSA antigen test for prostate cancer, and a comparison between screened and non-screened populations. Where there was no screening and males with progressive prostate cancer (say 1000 males to help explain the math) that noticed a problem went to get treatment about 44% survived 5 years. In places with screening 3,000 would be diagnosed with prostate cancer - the same 1,000 with progressive cancer that is a threat to mortality as before and an additional  2,000 with non-progressive prostate cancer that will be survived with or without treatment. All of these extra survivors make it look like we are much better at treating cancer than before, but the actual number of deaths due to prostate cancer has changed little. Gigerenzer also notes that 80% of men in their 80s will have prostate cancer, and very likely die of something else, as the cancers grow slowly not causing much harm.

The idea that screening is good and saving more lives than ever is just as likely to come from doctors as anyone else. It is an easy mistake to make, especially if you see people every day, sell them tests that say they have a problem, sell them treatments, and see them live more than 5 years so you can declare victory and send them home - hopefully without too many side effects. But the success changes very dramatically if you look at the problem from the view of the medical examiner signing the death certificates. Everyone dies, but he or she sees the statistics of how old they were and what they died off, and thus is immune to over-diagnosis bias, or other bias's such as the need to make money. The Harding Center for Risk literacy produces fact boxes - here is the for Prostate Specific Antigen Screening in german, as well as others, including the mammogram question which recently affected my wife, who knew damn well that the biopsy was going to be negative as it was the time before, but eventually we gave into the letters from MD Anderson, and I think would have made a different decision knowing these fact boxes.

Do I recommend reading it? - Yes. It is a reasonably light read, - although I think it is best read with a happy german professor accent in your head, as it can sound a little direct and formal in the beginning but soon mellows out. I think it suffers from a little too much jumping around between many different topics - or perhaps just a little more care in string them together for the reader is needed - although as always his English is far better than my non-existant German.  Overall though, a focus on the tools you need to make your own decisions, and where perhaps other professionals don't know any more than you will likely help you make one or two more good decisions in your life than perhaps if you didn't read the book. So get it from the library, or if you insist on buying or need instant gratification it is available on iBooks and Amazon.

Prostate Specific Antigen Cancer Screening Fact Sheet from the NCI