Essential Medical Statistics: Author Professor Betty Kirkwood on her bestseller

Professor Betty Kirkwood is Professor of Epidemiology & International Health at the London School of Hygiene and Tropical Medicine.

She is an epidemiologist with a statistical background and has worked at the School since 1979. Her research has been driven by a desire to improve the health of young children in low and middle income countries, and of their mothers. Her substantive research is accompanied by a commitment to translating research findings into health policy and programme action and to research capacity strengthening. She has an extensive network of overseas collaborators and close links with the World Health Organization, and is also a fellow of the UK Faculty of Public Health and of the Academy of Medical Sciences.

Since its first publication in 1988, Essential Medical Statistics has been one of Wiley’s most successful titles and it’s second edition was ‘Highly Commended’ in the 2004 BMA Medical Book Competition. It is now an established classic amongst medical statisticians. An introductory textbook, it presents statistics with a clarity and logic that demystifies the subject, while providing a comprehensive coverage of advanced as well as basic methods.

Alison Oliver talks to Professor Kirkwood about the book, her current research interests, teaching (a former student is Dr Ben Goldacre) and her commitment to improve the health of young children in deprived areas of the world.

 

1. Congratulations on the success of Essential Medical Statistics. How did the book come about in the first place?

I come from a mathematics and statistics background and when I first came to the London School of Hygiene and Tropical Medicine to be interviewed in 1979, I was told that, whilst they were aware I wanted to teach statistics to the MSc students which I would still do, would I mind starting by teaching statistics to the community health and developing countries MSc students instead? I of course said yes and it turned out that I absolutely loved it. It surprised me. The course was almost entirely taken up by medics and most of them were scared stiff by statistics. All were high functioning people with lots of experience, including one who went onto become the Minister of Health for Mozambique the following year. Somehow I managed to calm them down. I was then asked by Robert Winston to teach medical statistics in Hammersmith and whilst I was there, one of the students on his course said that he was going to go back to Singapore and try and publish my lecture notes!

Winston and students kept on telling me that I really ought to write a book as I taught very clearly. Eventually I went to Zambia for a few months to strengthen statistics in epidemiology at the Tropical Disease Research Centre in Eidola and my head of unit, Peter Smith checked whether I wanted to go. I was going by myself and it was a time when Zambia had curfews, there was some unrest and it could be very boring in the evenings. That’s when it occurred to me that I could start writing, so I spent my evenings and weekends in Zambia working on a draft of most of the first edition. This was in the early 1980s and it was all by long hand. My twins were born not long after returning to the UK, but when they were about 2 years old, I went back to spending the odd evening on the manuscript and almost rewrote the first edition which was then published in 1988. I approached Wiley Blackwell because they published Peter Armitage’s Statistical Methods in Medical Research and they said yes after sending a sample chapter.

2. What were your primary objectives when originally writing the book?

The aim of writing the book is to make statistics accessible to non-statisticians and non-mathematicians because what I observed with the way most people were teaching was that “There, there. I’ll help you with statistics, you needn’t worry about the details.” I grew up with Peter Armitage’s book and as a mathematician, it is brilliant but it is structured in a way a mathematician thinks, not in the way someone working in the real world thinks – someone who needs to use data but does not have a maths background. Somehow, I appear to have that ability to translate between the two. I learnt a lot writing the book as it challenged me to understand my subject more. I tried to show what methods you can use at various points as students will like the practical after the lecture but then they come up with data and are not sure what method to use so that’s where I focussed attention in my writing.

3. Did you anticipate the success the book would have? Had you seen that there was a gap in the market for this book?

All my students were telling me there was a gap. I knew there were other good books around at the time but if you looked at the structure of them, they were aimed at statisticians and that is what I was aiming to do differently. I was doing it to capture in writing what I had been doing in practice and to help the students I was teaching. I didn’t even think that it would do as well as it did. It still takes me by surprise when people come up to me at conferences from all over the world and ask, “Are you the Betty Kirkwood?” They all use the same phrase which has happened about 50, if not more, times which is, “I have to tell you that you saved my life.” But that shows you how scared people are of statistics and they explain how that when they were at medical school, this was the one book they could understand. I was hoping to be helpful but I didn’t anticipate such kind responses and that it appears to make much more difference to people than I would have thought it was possible to do.

4. The book is now in its second edition. If there was one piece of information or advice that you would want your readers to take away and remember after reading, what would that be?

It’s much more what I hope they will find from it and I hope it’s going to take the fear out of statistics and give them courage. The one piece of advice I have always given to students is “Don’t let common sense go out of the window.” If you are not familiar with a subject, you’ve got to look and we try and do that in the book. Statistics isn’t a black box, you need to understand what you are doing and any statistical analysis that you do needs to be guided by what questions you have to answer and what your objectives are. You then need to lay everything out in a series of steps that make common sense because everything boils down to asking the right questions and obtaining the answers. People panic over what test they should do, so I hope that it comes out very clearly in the book which I believe to be the most important of all – you should look at your data. You don’t start by doing sophisticated data, you start by producing nice simple tables with the appropriate percentages and consider what is it that the data appears to show you.

5. Who should read the book and why?

Anyone who needs to use statistics. It’s written as medical statistics and does have a very strong epidemiological slant, largely because my work has been for a long time in that area. Epidemiology and evidence-based medicine do go side by side, so it’s for anyone who’s doing anything in medical, healthcare or epidemiological statistics but it can be used in other fields. Jonathan and I tried very hard not to have our examples confined to a certain type of reader, so it also focuses on infectious diseases and lower and middle income countries. We tried to balance it so that wherever you were in the world, it would work.

6. Why is the book still of particular interest now?

Statistics is the science behind assessing evidence. I can’t see it ever not being relevant.

7. What inspired you to study mathematics and statistics originally before moving onto become a child health epidemiologist and researcher?

I went to university at the end of the 1960s until the early 1970s. Many of us wanted to change the world. It was that sort of time. I wanted to do something helpful but wasn’t sure what. I had enjoyed Spanish and geography at school but my maths teacher persuaded me to pursue maths. By nature, I would have probably gone onto anthropology. I find numbers easy and think like a mathematician but I am fascinated by people. I studied maths at Cambridge and we had this probability course. In those days, you didn’t encounter statistics until university and I just thought it was absolutely fascinating. Statistics at Cambridge was all about the theory behind the methods. We didn’t look at a single set of data. I didn’t pursue a career in medicine as the very idea of cutting up a frog repelled me. Statistics was a discipline in which I could offer skills and a contribution.

8. You continue to teach here at the LSHTM. What is it that you love about the School?

I’ve been here 36 years and it is a place that has suited me very well. We’re lucky as academics. As we work with postgraduates, there are no quiet periods, even in the summer, because students are always working on projects. The School has always been engaged with the real world. One of the concerns I had when I first came here is that there were hardly any women here, teaching or studying. It’s now completely switched the other way round which is fantastic. I do wonder why men aren’t going into epidemiology. Yesterday I was teaching a class of about 70 which was followed by group work. There were eight presenters in total, seven of whom were women. It’s probably 70-80% women in my class.

9. One of your students was Dr Ben Goldacre and he is now an epidemiologist as well as a writer. What do you think of his work on All Trials?

I did teach him and ask him to lecture here sometimes when he’s available. The work that he does is fantastic and he writes fascinating stories. The students love it when he comes. I agree entirely with his work on All Trials. It is still difficult to get studies published if your intervention shows no impact. I would like to see that change. Even though journals pay lip service so they don’t have positive selection bias, you’ve only got to look at The Lancet, and I know not just for myself but for colleagues as well that The Lancet likes issues that have a big impact – “This saved 20% of lives” for example.

If you work on something that is as scientifically rigorous as anything else, testing something that people thought would really make a difference but in actual fact, you learn that it does not make that big a difference in practice as was originally thought, The Lancet are less interested. The Lancet did publish a ten year study I worked on which looked at the impact of weekly Vitamin A supplements on maternal mortality. There was a suggestion from an earlier study in Nepal that it could have a dramatic impact, although we never published the entire findings. 120,000 pregnancies were studied during those ten years. Therefore, we had a tremendous amount of data and we wanted to add to the current data on vitamin supplements. However, we were told it was not that interesting because it did not have an overall impact on mortality. But this was really important evidence! The first vitamin A trial I was involved on was on 4 month large dose supplements to young children and we reduced mortality – that was magical, the way you feel when you have a really positive result – we had a party that night to celebrate.

10. Your research, you say is ‘driven by a desire to improve the health of young children in low and middle income countries, and of their mothers.’ Over the years, what has been the work that you have been proudest of?

The moment when we realised that we had reduced hospitalisations in children. The work that we did in northern Ghana in the early 80s is the work I am proudest of. I love Ghana and it started 25 years of research that I only just finished last year. We are still writing up the last bits.

Following on from the Vitamin A study, we worked on what Vitamin A rich foods were available and we did some trials of improved practices e.g. could families add a teaspoon of red palm oil, which is rich in Vitamin A, to children’s’ food before serving. I presented these results with an interpreter to a meeting of village chiefs in Kintampo in the centre of Ghana. At the end, one of the chiefs came up to me and said he loved my presentation and announced to everyone there that thanks to my work, their job now was to go back to their villages and get people to do it. Feeling part of the community was wonderful.

The work I enjoyed the most was a trial we did in Ghana where we trained community based surveyance volunteers as part of the Gini worm eradication. Having set up that work force, the government has been using them to publicise immunisation days, giving out mosquito nets, etc. and we were testing a strategy to reduce newborn deaths by making home visits. We worked with these volunteers in seven districts and we trained 450 volunteers to make home visits to give advice during pregnancy and the first week of life, and how to correctly assess any children who needed immediate care. We had an 11% impact on mortality which if you put together with the other studies, over a large area, our findings were consistent with other trials. I got into Vitamin A because I think it needed doing and I am really proud of all my work on it with one definitive trial after another. I’m really lucky and very privileged to do what I do.

11. What are you working on at the moment?

I have a Wellcome Trust funded program for work in Pakistan. Monthly home visits are being made throughout pregnancy and the first two years of life with a focus on promoting child development through play and stimulation by encouraging maternal child responsiveness and introducing appropriate weaning food at the right time. Pakistan and particularly India have the largest number of undernourished children in the world and where we are working, 46% of 18-30 months olds are stunted. It’s absolutely awful and we have helped to reduce mortality but there is still an enormous amount to do. There is a very late introduction of weaning foods at 11-12 months. There are also quite strong beliefs that children feed themselves whereas many mothers know that sometimes you have to play aeroplanes or distract your baby with a toy in order to get food into his or her mouth! The mothers don’t even try and we are trying to improve the weaning period messages and ask them to respond to their children to encourage eating. Breastfeeding is getting out there but not weaning. We have trained how to council, how to problem-solve and help mothers overcome constraints. We hope this will make a difference and reduce under nutrition and improve cognitive development. I am now 65 and this may be my last major piece of work. But I would love to support people who wish to tackle the gap from the age of 2 years to preschool as in much of the world, this area is totally neglected. In general, children do really well being at home for the first two years and they tend to show increased development if they interact with others from the age of 2. This is where the gap is and I would very much like to work on how can we bridge that gap and make it sustainable and cost-effective.

12. Who are the people who have been influential in your career?

Helen Bromby was my maths teacher. She was very old-fashioned and she retired when I finished sixth form. She had this wonderful way of behaving. If anybody was misbehaving, she would say, “Oh, you are being so silly.” The person would then just shrivel. She had the best disciplinary record of anyone in the school. She never raised her voice to anybody. She believed in me and gave me a lot of confidence. I stayed in touch with her, she came to my wedding and we met up once a year until she passed away.

My very first job was working for the Central Public Health Laboratory Service, computerising the reporting of old communicable diseases within the UK. I was so lucky because I hadn’t done any biology. Thanks to summer jobs, I happened to have done a lot of programming. So I did all the programming, came up with all the codes for communicable diseases and it was the most amazing introduction to epidemiology and public health. It took 18 months in total and there was a wonderful guy there called Bill Fletcher who was a very old-fashioned records officer. Everything was done on pieces of paper and card. I learnt so much from him, especially the practical side of data analysis, which has influenced how I have approached data collection to this day.

Here at LSTHM, I shared a room with Richard Hayes for seven years and it was a wonderful experience. Hopefully he learnt from me as well! We got on very well and discussed many issues. If one of us ever got stuck, we would talk it through together and it made an enormous difference. Peter Smith, who appointed me here, was also an inspiration.