Greg Fegan has a PhD in Epidemiology from Tulane University, New Orleans, LA, and a MSc in Information Systems Engineering from the Polytechnic of the South Bank, London, UK. He is Head of Statistics, KEMRI-Wellcome Trust Research Programme, in Kilifi, Kenya.
Work profile
LindedIn profile
Constantine: Greg, what has been your trajectory so far? It seems you went from the UK to the US and back, and for the past few years you have been living in Kenya.
Greg: After a small stint of work in London doing research on people with disabilities, I went to the London School of Hygiene and Tropical Medicine (LSHTM) which exposed me to sub-Saharan Africa which I first visited in 1989 and led to my first publication (Lancet 1991;337(8756):1499-502). Pedro Alonso, the first author of that paper, is now the Director of the Global Malaria Programme at WHO. After failing to get the Head of Computing and Statistics job at UK Medical Research Councils' (MRC) The Gambia, I was contacted by the USAID-funded DHS program with an offer to work for them, which led me to Maryland where I stayed for about a year and a half. MRC The Gambia, came calling again, which led to me working at the MRC's labs in the Gambia as the Head of Computing and Statistics for a little over two years. The period in the Gambia was directly prior to my starting the PhD at Tulane.
I've lived in the US three times: once during a year off from college in 1982/3 (when I hung out in Southern California, near Venice Beach, as a tourist), then again in 1989/90 (on an H-1 visa, when I worked for DHS, living in Baltimore), and then most recently from 1994-2003 (during which time I attended Tulane School of Public Health and Tropical Medicine, living in New Orleans). Until about 2008 we used to take our annual home leave in the US, as both my wife and I were then permanent residents.
Constantine: You were originally trained in Information Systems, right? How did you get interested in Statistics?
Greg: Yes, I worked in computing for about 7-8 years, mainly with data entry and data management systems, rather than with the hardware. At some point in the early 1990s I was asked to closely manage the data for a malaria vaccine safety trial, and when playing around with a simple cross-tab on blinded data, I discovered something that was important for a related ongoing high-profile trial that we were also conducting at the time. Back then, I was primarily a computer/database guy, but stumbled upon a real finding by a simple, but appropriate, analysis.
Constantine: What is your job and what does your 'typical' workday involve?
Greg: I am Head of Statistics for the KEMRI-Wellcome Trust Research Programme, in Kilifi, Kenya. Kenya Medical Research Institute (KEMRI) is a parastatal organization that is charged by the Kenyan Ministry of Health to produce research of relevance to the health situation in the country. Wellcome Trust is an independent charitable foundation based in the UK. The KEMRI-Wellcome Trust Programme (KWTRP) is an independent entity formed as a partnership between KEMRI and Wellcome Trust, with a main site in Kilifi and a secondary site in Nairobi. The KWTRP has just celebrated its 25th anniversary, and is supported by a core grant from the Wellcome Trust, funneled through Oxford University. So, technically, I am employed by Oxford University, but I'm actually working full-time at the KWTRP at Kilifi. Our staff here number about 650, including about 20 expatriates (all British bar two Dutch and one New Zealander). By the way, there are a number of other KEMRI centers like ours, where there is a largely expat-directed and funded research program embedded within a KEMRI center, including KEMRI-CDC, KEMRI-Walter Reed, and KEMRI-JICA (which is the Japanese equivalent of USAID, I believe).
Regarding my day-to-day activities, to paraphrase and Anglicise Tukey a little, I mess around in a lot of other peoples' gardens. A typical kind of day will usually involve activities such as reading proposals, commenting on manuscripts for publication, consulting with scientists about study designs, maybe reviewing a paper for PLoS One, where I am one of their many Statistical Advisors. Once in a while I get to open a stats package but far from daily. Each day, I usually will have a chat with junior colleagues, interns, or attachment students whom we host. I also attend protocol development meetings, data management meetings, and scientific review meetings. On a less frequent basis, I meet with PhD students whom I co-supervise or whom I assist with specific technical aspects of their work.
Constantine: You have been a member of TSHS for a few years now. Do you consider TSHS your primary section or is it secondary?
Greg: I would say primary.
Constantine: What is your reason for participating in TSHS?
Greg: I'm keen to know how to do my job better and to learn from others.
Constantine: Do you go to JSM? If not, why not? If yes, what do you get out of it?
Greg: I have never attended JSM. My post is funded by the core grant for the KEMRI-Wellcome Trust Programme and there really isn't any money set aside for me to attend conferences just to attend. However, should I get a paper or poster accepted to a conference, then the Programme will usually fund such a trip. For example, I attended the ASA's Conference on Statistical Practice in New Orleans in 2013, where I had an e-poster on display.
Constantine: Do you consider yourself an "isolated statistician"? If you even think of yourself primarily as a statistician (many of us seemed to have a lot broader scientific consultation and training roles, rather than just traditional statistical functions).
Greg: I don’t really see myself isolated as a statistician, as we have at least 6 other people who have master-level training in statistics. I also have a good buddy and colleague in Tanzania, Jim Todd, with whom I regularly consult and whom I’ve known for over twenty years (going back to the time we were in The Gambia, where he was employed as the senior statistician at the MRC). I do see myself as a statistician, but maybe more as a statistical epidemiologist. However, it was partly my slight insecurity about not having a "stats" degree that I decided to get the PStat accreditation. Kenyans are very particular with, and arguably over influenced by, qualifications and I was concerned that my not having a formal qualification in Statistics could be a problem in the future.
Constantine: What are your thoughts about statistics in the future? About statistics education?
Greg: In my specific setup in Kenya I work for a research institute that happens to sponsor and host PhD level training, but we don't do classical teaching/training per se, as most of our PhDs have no coursework requirements. However, Kenya is undergoing a massive expansion in tertiary-level education, with the number of degree granting institutions having expanded almost tenfold within the past few years. We have a number of arrangements, both formal and informal, to try and assist local universities with improving their training of their students. I believe that the only way that Kenya can meet this new demand is by reusing a lot of the very useful MOOC material that is coming out of the West. The young people here are becoming more and more tech-savvy and they are thirsty to show off what they can do. They are also ambitious, hard-working, entrepreneurial, and determined to learn and contribute.
I am also a big believer and fan of Open Science. Nowadays, I always put my name to the stats reviews I do for PLoS One because I feel that is the right thing to do. I always cyberstalk the authors of papers I review and I think it is only fair that they can do likewise.
I think that journals should start to also ask for data and code as supporting material to publications, and that we should be training the next generation to make this the norm. Tools such as R with Rstudio and GitHub make the ability to do that global and I am keen to promote such things. As I have often said to my younger colleagues, "I am old enough and ugly enough not to have to do reproducible research, but you aren’t, and I suspect that in your careers it is almost inevitable that the game will require this."
Constantine: Tell us a little bit about where you live and your life in Kenya.
Greg: I have been in Kenya since 2003, coming here just after finishing my PhD. Kilifi is a beautiful spot on the Indian Ocean. The town itself has about 50,000 people and is located 65 km north of Mombasa, the second-largest city in Kenya. It is growing rapidly, especially since the inauguration of the local Pwani University, which has more than 5,000 students. A large and growing population of Kenyans who
originate from other regions of the country have made Kilifi
their permanent residence, investing both in the real estate and
business markets. There is a long-established white Kenyan population
that has been here since the mid-1970s, and there are also various other
expatriates who have settled here. For example there are quite a few
Italians who are either permanently here or own holiday homes.
My non-work time is taken up with my family for the most part. The kids like to swim, bike, snorkel, and mess around on boats—when they’re not playing games on tablets. They go to a local school that is completely run on solar and is about 5 km from our house, which we built in 2012. Basically, we drew the plans and my wife project managed the whole thing—she is very organized. The school which the kids attend grew out of a bunch of us here at KWTRP who had kids of a similar age. It started as a small nursery near our work and has now grown to a school with over 100 kids in it.
[This interview was conducted through emails, in November 2014. The text is based on Greg Fegan's own words, with minimal editing by Edna Ogada and Constantine Daskalakis.]