When I learned that Douglas Altman had left us forever, I became as sad as if he had been someone from my family. I went to my library and took my favorite book of biostatistics whose author is he: "Practical Statistics for Medical Research ". A classic, an unmistakable book, solid in every way (materially and educationally), hard covers, red color, impeccable edition, accessible to anyone who knows the 4 basic arithmetic operations. I bought it in 1994, in London, where I worked at that time and where I made my first weapons in the field of biomedical research working for the pharmaceutical industry. It was my first book of biostatistics in English, today I have more than 20, but this is still my reference book. A couple of years ago I read it again, from end to end. And a couple of weeks ago, without going any further, I consulted it again. I consult it because there are concepts that I forget, I learn and I forget again (I do not have mathematical training beyond the undergraduate degree in medicine, which is very little); I also consult it because, although statistics seem to be a discipline of the exact sciences, there is often debate and, worse, more often there are misconceptions, wrong use of statistical techniques. So, to substantiate some of my statements in front of a recalcitrant colleague, I often go and look out for the clarifying quote in Altman's book or in one of his articles that, being so many, I cannot go on naming them.

What I am most grateful to Altman is to have dedicated heart and soul and with such perseverance to bring statistics (as an inferential analytical method) to people like me, doctors or clinical researchers, for whom mathematical jargon and the complex abstract models used are an obstacle to cooperation and communication between clinicians and statisticians. The preface to his book begins with a consoling quote: "The difficulties that many intelligent people have with the 'sums' are endless - Greenwood (1948)."

The other essential virtue of his books and articles is that he always associated the statistical concepts that he tried to convey with the methodological aspects, the design and the practical interpretation of the results.

Finally, I cannot fail to mention the most notable article he wrote, notable for its frankness, its lucidity and boldness. In January of 1994, perhaps while I was buying his book, he wrote an editorial in the BMJ whose title was "The Scandal of Poor Medical Research." I imagine that, in the height of exasperation, he decided to unveil the dirty secret, known to many and hidden from everyone, that the quality of research in the area of medicine was bad, to the point that he proposed to do "less research, better research and for good reasons." Less research, because doing too much and bad research is not ethical and is a waste of resources. Better research, because it is the only way to find valid answers to health problems. And for good reasons, because the enormous amount of research done by researchers to advance an academic career or for reasons unrelated to an eventual clinical application is devastating.

We are going to miss him.