I read yesterday with grief that one of our profession’s greatest scientists, Dr. Anne D. Haffajee, has deceased in August, just two years after we lost her collaborator for so many decades, another giant of periodontal research, Dr. Sigmund S. Socransky, both at Forsyth in Boston. I have to admit that I had never met Dr. Haffajee personally. In the early 1980s we had a brief exchange of letters about an anaerobic gas-flushed syringe for obtaining a sample of the incredible oxygen-sensitive subgingival microflora in deep periodontal pockets. The rather complicated device had been invented at Forsyth (as so many other things), and Anne suggested that we should just manufacture our own syringe. It was not for sale.
Especially Anne Haffajee’s way of looking at data profoundly inspired me since my own early studies in the late 1980s. Due to her innovative data analyses throughout her immense successful career I really became fascinated of medical statistics as an invaluable tool for our complicated (i.e., in particular, site-specific) observations. And, the widespread criticism, by biostatisticians, of her early approach, namely erroneously considering the plentiful observations which can be done in just one oral cavity as independent, taught me some important lessons. However, while the Forsyth group seems to have boggled and soon unfortunately started to apply the most conventional methods after having summarized their most valuable site-specific data at the subject level, I went beyond. After some cumbersome exercises with Mantel-Haenszel methods, I managed to get a GEE macro by courtesy of Dr. L. Mancl at the University of Washington in Seattle. At that time, GEE had not been implemented in the major statistical software packages (and my colleagues at the Institute of Biometrics at the University of Münster in Germany are grateful to this day for my floppy disc containing Dr. Mancl’s macro). But I had to meticulously collect new data to apply the more sophisticated methods, in a way the other way round as usual. Much later, I learned about the numerous revealing advantages of multilevel modeling.
Thank you, Anne, for decades-long inspiration.
13 November 2013 @ 12:36 pm.
Last modified November 15, 2013.