Dear Dr. Hanley, Thank you for your very nice note. I have attached a spreadsheet with the numerators and denominators from the JAMA article on folic acid fortification. I am very pleased that you plan to use this as an example for your course in the fall. Please let me know if there is anything in the spreadsheet that requires further explanation or clarification. Sincerely, Margaret Honein National Center on Birth Defects and Developmental Disabilities, CDC -----Original Message----- From: James Hanley [mailto:James.Hanley@McGill.CA] Sent: Saturday, June 23, 2001 12:48 PM To: MHonein@cdc.gov Cc: James.Hanley@McGill.CA Subject: jama article Margaret A. Honein, PhD, MPH, National Center on Birth Defects and Developmental Disabilities, Centers for Disease Control and Prevention, Atlanta, MHonein@cdc.gov Dear Dr Honein. I read with considerable interest your article in JAMA and I congratulate you for a very important and thoughtful data analysis. I am sure you are swamped with calls for interviews etc., and so I understand if you will be a bit late in answering. I will use the data in your paper in the course I teach to our graduate epidemiology students on the statistical analysis of rates and proportions. I will use them to illustrate [Poisson and extra-Poisson] variation in rates over time, and the use of regression to try to isolate changes in rates following implemenntation of policy changes. Some of the examples I have used up to now are rates of fatal collisions on highways before and after the switch back to 65 mph from 55mph, rates of motorcycle fatalities after introduction of helmets, and rates of accidents before and after the "change the clock" in april and october, but in all 3 examples, the series are very short. Thus, might I presume on you to supply me with the numerators and denominators that were used to create the data points in Figures 1 and 2? I could try to reconstruct them from the graph and from the numbers in Tables 1 and 2, but they will be imprecise. I would be happy to receive them in whatever form is fastest and easiest for you.. since I am sure you have many calls on your time. If, by chance, you have a file with more extensive data than you were able to show in the paper, I would be happy to receive that instead, and I can get our students to do some calculations on their {as you will see from my web page, I give them the intro to articles, and ask them to form the questions, and use the data kindly supplied to me by authors, to do their own analyses. Of course, there are also several other wonderful teaching points in you article -- i.e. epidemiologic inferences, the assumptions on which they are built, data quality etc. What I especially like are articles that use available data, even if not the perfect data, since I am always preaching that the best should not become the enemy of the good. If you are interested, the material and other data sets I use in the course can be seen on my course web site http://www.epi.mcgill.ca/hanley/c626/ our dept. page is http://www.epi.mcgill.ca and my own page is http://www.epi.mcgill.ca/hanley/ You enter the course page [and my other course pages as well] using the username mantel and the password haenszel, both lower case. Even though my course pages contain some articles that I have not obtained publisher permission for, I find that the password keeps out most prying eyes .. even my students, to whom I give the password, have trouble getting in! I will be putting your JAMA article on there for the upcoming course in the Fall. Again, congratulations on producing a valuable article, both substantively and pedagogically. And if you would be able to send the numbers before I teach in september, that would be great. Yours sincerely James A Hanley James.Hanley@McGill.CA James A Hanley, PhD Department of Epidemiology and Biostatistics tel: (514) 398-6270 McGill University fax: (514) 398-4503 1020 Pine Avenue West, Montreal, Quebec email: James.Hanley@McGill.CA H3A 1A2 Canada web page: www.epi.McGill.CA/hanley/ Attachment converted: Macintosh HD:data from JAMA 2001.xls (XLS4/XCEL) (00123D45)