data sasuser.crohns;
input cases month births;
case=1;   number=cases; rate=100000*cases/births;output;
case = 0; number=births-cases; output;
lines;
51  1 197430
46  2 191657
42  3 222636
50  4 220069
63  5 219246
55  6 209040
55  7 211222
56  8 206660
62  9 205330
58 10 197303
50 11 185786
39 12 188225
;
options ls=65 ps=55; run;

proc plot data=sasuser.crohns;
plot rate*month;

proc freq data=sasuser.crohns;
tables month*case / chisq trend;
weight number;
run;

/*
Date: Mon, 29 Oct 2001 14:43:26 +0100
From: Lars Pedersen <lap@SOCI.AU.DK>
Subject: Crohns disease
To: 'James Hanley ' <james.hanley@McGill.CA>

Dear Mr Hanley

Thanks for your interest in our paper.
Here are the numbers used in the analysis.
I hope it's ok.

Best regards

Lars Pedersen
Biostatistician
Department of Clinical Epidemiology
Vennelyst Boulevard 6
8000 Aarhus C - DK


-----Original Message-----
From: James Hanley
To: hts@soci.au.dk
Sent: 20-10-2001 16:02

Hi Professor Sørensen

For use in my teaching, could you let me have the numerators and
denominators for the 12 observed rates in your article in the bmj today?

I could ask students to assume that the denominators were
proportional to the number of days each month, but I
know they will object that monthly birth rates are not exactly uniform..

Thank you.

Dear J Hanley
Thanks for mail. We will provide you the data early next week.
Best regards,
Henrik Toft Sorensen

*/